View ISL5416KIZ_4222304.PDF datasheet online --- IC-ON-LINE

Datasheet File OCR Text:

1 ? fn6006.3 caution: these devices are sensitive to electrostatic discharge; follow proper ic handling procedures. 1-888-intersil or 321-724-7143 | intersil (and design) is a registered trademark of intersil americas inc. copyright ? intersil americas inc. 2002-2004. all rights reserved all other trademarks mentioned are the property of their respective owners. isl5416 four-channel wideband programmable downconverter the isl5416 four-channel wideband programmable digital downconverter (wpdc) is designed for high dynamic range applications such as cellular basestations where the processing of multiple channels is required in a small physical space. the wpdc combines four channels in a single package, each including: an nco, a digital mixer, digital filters, an agc and a resampling filter. all channels are independently programmable and may be updated in real time. each of the four channels can select any of the four digital input buses. each of the tuners can process a w-cdma channel. channels may be cascaded or polyphased for increased bandwidth. selectable outputs include i samples, q samples, and agc gain. outputs from the part are available over the parallel, serial or up interfaces. features up to 95msps input four parallel 16-bit fixed or 17-bit floating point inputs programmable rf attenuator/vga control 32-bit programmable carrier nco with > 110db sfdr 20-bit internal data path filter functions - multi-stage cascaded-int egrator-comb (cic) filter - two programmable fir filters (first up to 32-taps, second up to 64-taps) - half band interpolation filter - resampling fir filter overall decimation from 1 to >4096 digital agc with up to 96db of gain range up to four independent 16-bit parallel outputs serial output option 16-bit parallel p interface 1.8v core, 3.3v i/o operation evaluation board and configuration software available pb-free available applications basestation receivers: gsm/edge, cdma2000, umts. block diagram ordering information part number temp range ( o c) package pkg. dwg. # isl5416ki -40 to 85 256 bga v256.17x17 ISL5416KIZ (see note) -40 to 85 256 bga (pb-free) v256.17x17 isl5416eval1 25 evaluation kit note: intersil pb-free products employ special pb-free material sets; molding compounds/die attach materials and 100% matte tin plate termination finish, which is compatible with both snpb and pb-free soldering operations. intersil pb-free products are msl classified at pb-free peak reflow temperatures that meet or exceed the pb-free requirements of ipc/jedec j std-020b. input select clock & format range control input channel routing nco mixer cic fir1 filter fir2 filter agc ihbf resampler output routing & formatting aout(15:8) aout(7:0) fsynca oea test register ain(16:0) enia clka input a channel o output i q i q i q i q i q i q input d input c input b channel 2 channel 1 channel 3 rf attenuator vga control eout(15:0) clko1 clko2 /intrpt jtag synchronization p interface synco syncin1 syncin2 p(15:0) add(2:0) ce up mode rd or rd/wr wr or dstrb reset data sheet august 2004
2 mux 16 / digital tuning mixer nco 32-bit control >110 db sfdr 24 / 24 / cic filter 24 / 24 / mux cascade inputs 1-5 stages r=2-64k bypass gain x1, 2, 4, 8 20 / 20 / 24 / 24 / round saturate fir 24 / 24 / 1 1-32 taps r=1-8 bypass gain x1, 2, 4, 8 20 / 20 / round saturate fir 24 / 24 / 2 1-64 taps r=1-8 bypass filter cascade output agc r o u n d 24 / 24 / 16 / 16 / f i f o 0 - 96 db bypass ihbf 16 / 16 / 16 / 16 / 16 / 16 / resampling filter mux 24 / 24 / select format to parallel sequencing and routing ain(16:0) bin(16:0) cin(16:0) din(16:0) test input to serial routing to up interface input format range control ain(16:0) enia clka mux digital tuner output format slot control ch 0, 1 muxing mux ext agc cntrl input format range control bin(16:0) enib clkb mux digital tuner output format slot control ch 0, 1 muxing agc gain input format range control cin(16:0) enic clkc mux digital tuner output format slot control ch 2, 3 muxing input format range control din(16:0) enid clkd mux digital tuner output format slot control ch 2, 3 muxing agc gain mux eout(15:0) test register cascade in cascade out o u t p u t m u l t i p l e x i n g serial outputs aout(15:0) fsynca oea bout(15:0) fsyncb oeb cout(15:0) fsyncc oec dout(15:0) fsyncd oed clko1 clko2/ intrpt up interface sequenced up read data p(15:0), upmode, rd (rd/wr ), wr (dstrb ), ce, add(2:0) synchronization jtag synco syncin1 syncin2 trst tms tclk tdi tdo channel 0 channel 1 channel 2 channel 3 input a input b input c input d typical channel reset isl5416
3 256-lead bga top view k j h g f e d c b a 123456789 11 10 l m n p1 oed dout0 din0 12 dout4 13 14 cin11 din10 din11 din4 din3 vcc p0 dout3 vcc cin12 rd add0 dout11 dout12 cin13 din12 din13 gnd gnd wr cin14 dout14 cin15 gnd dout15 cin16 clkc bout0 vccio bout1 bout2 ain0 ain2 bout3 ain1 ain3 bout4 vccio reset ain5 bout7 ain4 ain6 bout9 vccio oeb p15 clkb vccio fsyncb bin11 bin12 ain13 bin16 oea syncin2 gnd eout15 bout13 bin10 vcc ain15 vcc synco aout15 syncin1 clka aout14 aout13 gnd enia ain10 ain11 ain12 ain14 eout13 p dout2 vcc dout1 cin9 eout1 dout5 dout6 dout7 din9 vcc din6 enid din2 din1 power pin ground pin signal pin thermal ball nc (no connection) bin4 bin7 vcc bin0 p6 p5 p7 gnd ain9 bin9 ain8 ain16 vcc bout12 bin8 ain7 vccio bin6 bout8 bout14 bout6 gnd bout5 din16 p4 eout3 dout13 vcc cin10 gnd vccio bout15 gnd eout12 p14 bin3 bin2 vcc eout4 trst add2 p2 p9 p8 gnd vcc = +1.8v core supply voltage vccio = +3.3v i/o supply voltage cout12 cout11 cout13 clko1 cout15 aout0 aout1 aout2 vcc aout4 aout5 aout7 aout9 bout10 dout9 gnd vccio fsynca aout10 aout8 cout14 dout10 cout10 cout2 oec cout0 eout0 cout4 din8 din7 din5 enic cin2 ce tclk cout3 cout1 cin8 cin0 cout5 cout6 fsyncc cin7 cin6 cin5 cin4 cin3 cin1 gnd vccio fsyncd cout7 cout8 dout8 cout9 r t 15 16 aout12 aout11 bout11 enib bin14 bin15 eout14 p12 bin5 eout10 bin13 eout11 vcc p13 p11 tdo aout6 gnd gnd gnd upmode gnd p10 bin1 tms tdi aout3 eout8 eout9 clkd eout6 eout7 gnd gnd eout5 eout2 p3 din15 din14 gnd add1 gnd clko2/ intrpt note: thermal balls should be connected to the ground plane unused input balls should be connected to ground or v cc io as appropriate isl5416
4 pin descriptions name type internal pull-up/down description power supply vcc - positive power supply voltage (core), 1.8v 0.09 vccio - positive power supply voltage (i/o), 3.3v 0.165 gnd - ground, 0v. inputs ain(16:0) i pull down parallel data input bus a. sampled on the rising or falling edge (programmable) of clock when enia is active (low). the bus order can be programmed (see iwa = 0*00h, bit 4). bin(16:0) i pull down parallel data input bus b. sampled on the rising or falling edge (programmable) of clock when enib is active (low). the bus order can be programmed (see iwa = 0*00h, bit 4). cin(16:0) i pull down parallel data input bus c. sampled on the rising or falling edge (programmable) of clock when enic is active (low). the bus order can be programmed (see iwa = 0*00h, bit 4). din(16:0) i pull down parallel data input bus d. sampled on the rising or falling edge (programmable) of clock when enid is active (low). the bus order can be programmed (see iwa = 0*00h, bit 4). enia i pull down input enable for parallel data input bus a. active low. this pin enables the input to the part in one of two modes, gated or interpolated. in gated mode, one sample is taken per clk when enix is asserted. enib i pull down input enable for parallel data input bus b. active low. this pin enables the input to the part in one of two modes, gated or interpolated. in gated mode, one sample is taken per clk when enix is asserted. enic i pull down input enable for parallel data input bus c. active low. this pin enables the input to the part in one of two modes, gated or interpolated. in gated mode, one sample is taken per clk when enix is asserted. enid i pull down input enable for parallel data input bus d. active low. this pin enables the input to the part in one of two modes, gated or interpolated. in gated mode, one sample is taken per clk when enix is asserted. control clka i pull down input clock for data bus a. clka or clkc may be used for ain(16:0). clkb i pull down input clock for data bus b. clkb or clkc may be used for bin(16:0). clkc i input clock for data bus c. clkc is also the master clock for all channels of isl5416 clkd i pull down input clock for data bus d. clkd or clkc may be used for din(16:0). syncin1 i pull down global synchronizati on input signal 1. syncin1 can update the carrier ncos, reset decimation counters, restart the filter, and restart the output section among other functions. for most of the functional blocks, the response to syncin1 is programmable and can be enabled or disabled. syncin2 i pull down global synchronizati on input signal 2. syncin2 can update the carrier ncos, reset decimation counters, restart the filter, and restart the output section among other functions. for most of the functional blocks, the response to syncin2 is programmable and can be enabled or disabled. synco o synchronization output signal . the processing of multiple is l5416 devices can be synchronized by tying the synco from one isl5416 device (the mast er) to the syncin of all the isl5416 devices (the master and slaves). an optional inter nal synco to syncinx connection is provided. reset i pull up reset signal. active low. asserting reset will halt all processing and set certain registers to default values. isl5416
5 jtag tdo o test data out tdi i pull up test data in. tms i pull up test mode select. tclk i pull down test clock. trst i pull up test reset. active low. if jtag not used, tie this pin low. if there is a trace connected to the pin and there is enough board noise, the jtag port might get into an unexpected state and stop communications with the part outputs aout(15:0) o parallel data output bus a. a 16-bit parallel data output which can be programmed to consist of i, q, agc. data from channels 0, 1, 2 and 3 can be multiplexed into a common parallel output data bus. information can be sequenced in a programmable order. can be ones complemented. can be divided into two 8-bit busses. see data output formatter section and microprocessor interface section. see table 24. bout(15:0) o parallel data output bus b. a 16-bit paralle l data output which can be programmed to consist of i, q, agc. data from channels 0, 1, 2 and 3 can be multiplexed into a common parallel output data bus. information can be sequenced in a programmable order. can be ones complemented. can be divided into two 8-bit busses. see data output formatter section and microprocessor interface section. cout(15:0) o parallel data output bus c. a 16-bit parallel data output which can be programmed to consist of i, q, agc. data from channels 0, 1, 2 and 3 can be multiplexed into a common parallel output data bus. information can be sequenced in a programmable order. can be ones complemented. can be divided into two 8-bit busses. see data output formatter section and microprocessor interface section. dout(15:0) o parallel data output bus d. a 16-bit paralle l data output which can be programmed to consist of i, q, agc. data from channels 0, 1, 2 and 3 can be multiplexed into a common parallel output data bus. information can be sequenced in a programmable order. can be ones complemented. can be divided into two 8-bit busses. see data output formatter section and microprocessor interface section. below is the table of the serial output bits allocation for dout. eout(15:0) o a 16-bit parallel vga/attenuator control ou tput. partitionable into separate 4 or 8-bit busses. clko1 o output clock 1. can be programmed to be at clkc/n for n = 1 to 16. the polarity of clko1 is programmable. clko2/ intrpt o available only on rev b (final) version of the part. provides a complementary output or a second clock to simplify board routing. po larity is programmable. it can also be programmed as an interrupt from one or more channels for a sequenced read (fifo-like) mode. see register gwa = 0000h, bit 13. pin descriptions (continued) name type internal pull-up/down description serial output bits allocation ser. output a ser. output b ser. output c ser. output d sclkx * dout0 dout4 dout8 dout12 ssyncx * dout1 dout5 dout9 dout13 sd1x * dout2 dout6 dout10 dout14 sd2x * dout3 dout7 dout11 dout15 * x denotes a, b, c, d as appropriate isl5416
6 fsynca o frame synchronization out put signal for bus aout(15:0). fsyncb o frame synchronization out put signal for bus bout(15:0). fsyncc o frame synchronization out put signal for bus cout(15:0). fsyncd o frame synchronization out put signal for bus dout(15:0). oea i pull up output three-state enable for parallel data output bus a. active low. oeb i pull up output three-state enable for parallel data output bus b. active low. oec i pull up output three-state enable for parallel data output bus c. active low. oed i pull up output three-state enable for parallel data output bus d. active low. microprocessor interface p(15:0) i/o microprocessor interface data bus. see microprocessor interface section . p15 is the msb. add(2:0) i microprocessor interface address bus. add2 is the msb. see microprocessor interface section . wr or dstrb i microprocessor interface write or data str obe signal. when the micropr ocessor interface mode control ( p mode) is low, data transfers (from p(15:0) to the internal write holding register) occur on the low to high transition of wr when ce is asserted (low). when the p mode control is high this input functions as a data strobe dstrb control. in this mode with rd/wr low, data transfers (from p(15:0) to the internal write holding register) occur on the low to high transition of dstrb . with rd/wr high the data from the address specified is placed on p(15:0) when dstrb is low. see the microprocessor interface section . rd or rd/wr i microprocessor interface read or read/write signal. when the microprocessor interface mode control ( p mode) is low, the data from the addre ss specified is placed on p(15:0) when rd is asserted (low) and ce is asserted (low). when the p mode control is high this input functions as a read/write control input. data is read from p(15:0) when rd/wr high or written to the appropriate register when low. see the microprocessor interface section . p mode i pull down microprocessor interface mode control. this pin is used to select the read/write mode for the microprocessor interface. when 0, rd and wr , when 1, dstrobe and rd/wr . when p mode is 0, the microprocessor inte rface consists of separate rd and wr strobes; when p mode is 1, the interface consists of a rd/wr control and a single data strobe. see the microprocessor interface section . ce i microprocessor interface chip sele ct. active low. this pin has t he same timing requirements as the address pins. pin descriptions (continued) name type internal pull-up/down description isl5416
7 functional description the isl5416 is a four channel digital receiver integrated circuit offering exceptional dynamic range and flexibility. each of the four channels consists of a front-end nco, digital mixer, cic-filter, two fi r filters, agc, interpolation half band filter and re-samp ling filter. the parameters for the four channels are independently programmable. there are four 17-bit parallel data input busses (ain(16:0), bin(16:0), cin(16:0) and din(16:0)). the isl5416 supports both fixed and floating point parallel data input modes. the floating point modes support gain ranging a/d converters or a/d converter and rf/if atte nuators or vgas. gated or interpolated data input modes are supported. each input can be connected to any or all of the internal signal processing channels, channels 0, 1, 2 and 3. the four channels share a common processing clock (clkc). four input clocks are provided to allow for clock skew between input sources. each input has a range control circuits to monitor the signal level on the parallel data busses and to control the gain prior to the a/d converters. a 16-bit bus (eout(15:0)) is provided to control the external vga/rf attenuators. each front end nco/digital mixer/cic filter section includes a quadrature numerically controlled oscillator (nco), digital mixer, barrel shifter and a cascaded-integrator-comb filter (cic). the nco has a 32-bit frequency control word. the sfdr of the nco is >110db. the barrel shifter provides a gain of between 2 -45 and 4 to compensate for the gain in the cic. the cic filter order is programmable from 1 to 5 and the cic decimation factor can be programmed from 2 to 512 for 5 th order, 2048 for 4 th order, 32768 for 3 rd order, or 65536 for 1 st or 2 nd order filters. the cic filter can also be bypassed. each channel back end section includes two fir filters, an agc, interpolation half band fi lter and resampler. the first fir filter can have up to 32 taps and the second can have up to 64 taps. the 32-tap filter ca lculates 4 taps per clock, while the 64-tap filter calculates 8 t aps per clock. the coefficients for the programmable digital filters are 20 bits wide. each fir filter can be bypassed. the agc section can provide up to 96db of either fixed or automatic gain control. for automatic gain control, two se ttling modes and two sets of loop gains are provided. separate attack and decay slew rates are provided for each loop gain. programmable limits allow the user to specify a gain range less than 96db. a fixed coefficient interpolate-by-2 half band filter and a non-integer resampling filter follow the agc. coefficients for the resampling filter are provided in rom. four 16-bit parallel data out puts (aout(15:0), bout(15:0), cout(15:0) and dout(15:0)) are provided. the output of each channel can be routed to any of the output buses. outputs from more than one channel can be multiplexed through a common output if the channels are synchronized. dout(15:0)) can alternately be used as four serial output pairs. a common output clock (clko1) is used for the parallel output buses. a second clock output pin (clko2/intrpt ) is provided to simplify board routing or to allow a complementary output clocks. the isl5416 is programmed through a 16-bit microprocessor interface. the output data can also be read via the microprocessor interface. the isl5416 is specified to operate to a maximum clock rate of 95 msps over the industrial temperature range (-40 o c to 85 o c). the i/o power supply voltage range is 3.3v 0.165v while the core power supply voltage is 1.8v 0.09v. input select/ format block clocking the channel processing and output timing is clocked with the rising edge of clkc. each input bus can be clocked with the rising or falling edge of its own clock or with the rising or falling edge of clkc. the frequency of all the clocks must be the same, but providing separate clocks allows the inputs from multiple a/d converters to have a small amount of skew. input format the inputs can be fixed point or floating point with mantissa/exponents sizes of 14/3, 15/2, or 16/1. the exponent inputs are added to the exponent from the internal range control circuits, so if the range control circuits are used, the exponent pins ar e typically grounded and/or disabled via software in iwa = 0*10h, bit 3. the input format may be twos complement or offs et binary format in either fixed or floating point modes (iwa = 0*00h). gated/interpolated modes for input sample rates at sub- multiples of the clock rate, gated and interpolated input modes are provided. each input channel has an input enable (enix , x = a, b, c or d). in the gated mode, one input sample is processed per clock that the enix signal is asserted (low). processing is disabled when enix is high. the enix signal is pipelined through the part to minimize delay (latency). in the interpolated mode, the input is zeroed when the enix signal is high, but processing inside the part continues. this mode inserts zeros between the data samples, interpolating the input data stream up to the clock rate. the spacing between enix signals must be constant in the interpolated mode. multiplexed input mode each input section can select one channel from a multiplexed data stream of up to 8 channels. the input enable is delayed by 0 to 7 cl ock cycles to enab le a selection register. the register following the selection register is enabled by the non-delayed input enable to realign the processing of the channels. the one-clock-wide input enable must align with the data for the first channel. the desired channel is then selected by programming the delay. a delay of zero selects the first chan nel, a delay of 1 selects the second, etc. each input section selects only one channel of the multiplexed stream, so a separate input bus must be used for each channel of the multiplexed data stream. isl5416
8 syncinx use syncinx main purpose is as a processing start-up signal after a reset to align the start of processing of multiple channels or chips. this assures that the carrier phases have a known relationship and that the output timing aligns for multiplexing outputs. it can also be used after start-up as a system timing synchronization signal. two syncinx signals are provided so that one can be used as a regularly occurring signal (such as at time slot boundaries) and one as an infrequent signal (such as at start up or at 1 pps). if more than one air interface standard is processing in one part, one syncinx signal could be used for the slot timing for each standard. register updates from a processor write are synchronized to the clock, so that the register updates in multiple channels of the same part are time aligned. however, when synchronizing multiple parts the processor will need knowledge of the syncinx timing so that enabling the syncinx in multiple parts occurs between syncinx pulses. alternatively, syncin1 could be used as a regularly occurring synci signal and syncin2 could be a gated version. the channel processing control register might only be updated on syncin2 and the other synci functions would respond to syncin1. clkc clkx clkc clkx r1 r2 r3 r4 r1 r2 r3 r4 ^ ^ ^ interpolate slot# r1 r2 r3 r4 clk/clk m u x m u x clkx/clkc clkx clkc figure 1. input section f m t d m u x r m a p r r e g r e g r e g r e g ^ r e g ^ m u x processing enable to channels and range control to serial frequency offset clk mantissa exponent din (only) obin bus reverse float/fix (16:0) (0:16) enix xin(16:0) note: to simplify the board routing, each of the four input data busses can be reversed, msb for lsb (see iwa = 0*00h, bit 4) isl5416
9 isl5416 vga/rf attenuator (a/d range control) the range control section monito rs the output of the a/d and adjusts the rf/if gain to ma intain a desired a/d output range. the gain adjustments are in 6 db steps. the levels, adjustment rates, and gain to bit mapping are programmable. the range control section uses three programmable thresholds. two thresholds, an upper and a lower threshold, are compared against the average magnitude of the a/d output. the range control ad justs the gain to keep the average a/d output between the upper and lower thresholds. if the average is above the upper threshold, an internal attenuator control register is increased by a programmable amount. if the average is below the lower threshold, the gain attenuator control register is decreased by a separate programmed amount. the number of samples averaged for each decision is programmable. the adjustments to the attenuator control register can be less than 6 db to further filter the inputs. only the three msbs of the attenuator control register are used to control the rf/if gain, and these are weighted as 6, 12, 24 db steps. the third threshold, an immediate threshold, is compared against the magnitude of each a/d sample. if the magnitude of any a/d sample exceeds the threshold, the attenuator control register is immediately increased by the amount programmed for the immediate threshold. because there will be some time delay from a register change until the effect of the change is seen at the a/d, the immediate threshold is disabled for a programmable number of clock cycles after it has been triggered. to maximize the input sensitivity the range control also includes a programmable bias. if the average signal is between the upper and lower threshold, the bias value is added from the attenuator co ntrol register. this bias removes attenuation when it is no longer needed to avoid missing small signals due to high input noise figure. four counters control the amount of time that the input is averaged and align the adjustments to time slot boundaries. one counts out the time slot pe riod. if desired, this counter can be reset by a syncinx signal to align its count to the system timing. a second counter provides a programmable delay from the start of the first counter's period to the start of the integration peri od. this compensates for system delays or allows the adjustments to be made over a certain portion of the time slot. the third count er sets the integration period for averaging the input samples for the upper and lower threshold decisions. the fourth counter controls the number of integration periods per time slot. see figure 2 for a block diagram. note that the counters are ignored for the immediate threshold decisions. the user can program a separate code for output on the eout bus for each of the eight possible states of the three msbs of the attenuator control register. these codes can be up to 8 bits, but if four gain control sections are used, only four bits are available for each gain control section. the mapping of the gain control bits to eout bits is done in gwa = 0001h and the codes are programmed in iwa = 0*17h and 0*18h. the three msbs of the attenu ator control register can be routed internally to the channels to be used as the floating- point exponents. this adds gain in 6 db steps to compensate for the 6 db steps of rf attenuator. the msbs can be added to the input exponent bits if desired. there is a programmable delay from the attenuator control register to the channel input to compensate for rf/if filter group delay and a/d and isl5416 pipeline delays.
10 reg r e g mux ld down counter =0 ld down counter 0 ld down counter <0 ld down counter 0 =0 =0 sync interval (slot period) en en en delay (sync to start of integration) integration time integrations per slot - 1 enable counters enable updates ? ? down counter ld prld(7:0) a b th1 b th2 reg a>b a>b a 11 nco/mixer after the input select/format section, the samples are multiplied by quadrature sine wave samples from the carrier nco. the nco has a 32-bit fr equency control, providing sub-hertz resolution at the maximum clock rate. the quadrature sinusoids have exceptional purity. the purity of the nco should not be the determining factor for the receiver dynamic range performance (a typical spectrum plot is shown in figure 20). the phase quantization to the sine/cosine generator is 24 bits and the amplitude quantization is 19 bits. the carrier nco center frequency is loaded via the up bus. the center frequency control is double buffered -- the input is loaded into a holding register via the up interface. the data is then transferred from the holding regist er to the active register by a write to a special address or by a syncinx signal, if enabled in iwa = *000h. to synchronize multiple channels, the carrier nco phase accumulator feedback can be zeroed on loading to restart all of the ncos at the same phase (see iwa = *005h). the phase of the nco can be offset by programming iwa = *003h. the phase offset is not double buffered. mux mux r e g ena enb enc end en p expa expb expc expd exp p a mant bmant cmant dmant r e g r e g r e g r e g r e g reg reg reg reg reg fir filter r e g r e g r e g 0...-15 0...-23 r e g 0...-29 0...-23 69b 62b 53b 44b 34b barrel shifter integrate integrate integrate integrate integrate reg reg reg sin/cos gen reg reg reg reg serial freq offset cof cofsync encof fixed input gain = 2 -45 shift gain = 2 0 to 2 47 0...-23 0...-18 mux r e g r e g mux r e g r e g c l i p r e g r e g r e g max exp base shift value 0...-23 r e g r e g r e g decimation counter r e g r e g r e g r e g mux cic order fir coefficients 11-1 21-21 31-33-1 41-46-41 5 1 -5 10 -10 5 -1 center freq phase offset (comb) figure 3. nco, mixer and cic block diagram (from up) note: bus numbering such as 0...-23 indicates bit weights such as 2 0 ... 2 -23 (test input) isl5416
12 pn generator after the mixers, a pn (pseudonoise) signal can be added to the data. this feature is provided for test and to digitally reduce the input sensitivity and adjust the receiver range (sensitivity). the effect is the same as increasing the noise figure of the receiver, reducing its sensitivity and overall dynamic range. the one bit pn data is scaled by a 16 bit programmable scale factor. the overall range for the pn is 0 to 1/4 full scale. a gain of 0 disables the pn input. the bit weighting for the gain is shown in table 1. the minimum, non-zero, pn value is 1/2 18 of full scale (-108 dbfs) on each axis (-105 dbfs total). cic filter next, the signal is filtered by a cascaded integrator/comb (cic) filter. a cic filter is an efficient architecture for decimation filtering. the power or magnitude squared frequency response of the cic filter is given by: where m = number of delays (1 for the isl5416) n = number of stages and r = decimation factor. the passband frequency response for 1 st (n=1) though 5 th (n=5) order cic filters is plotted in figure 20. the frequency axis is normalized to f s /r, making f s /r = 1 the cic output sample rate. figure 19b shows the frequency response for a 5 th order filter but extends the frequency axis to f s /r = 3 (3 times the cic output sample rate) to show alias rejection for the out-of-band signals. figure 19a provides the amplitude of the first (strongest) alias as a function of the signal frequency or bandwidth from dc. for example, with a 5 th order cic and f s /r = 0.125 (signal frequency is 1/8 the cic output rate) figure 19a shows a first alias level of about -87 db . figure 19a is also listed in table form in table 84. the cic filter order is progra mmable from 0 to 5. the cic may be bypassed by setting the cic iwa = *001h bit 15. a barrel shifter precedes the cic filter to compensate for the large gain range of the cic. as the barrel shifter only adjusts in 6 db steps the total cic/barre l shifter gain ranges from 0.5 to 1.0. the barrel shifter is also used to convert floating point input data to fixed point for processing. the exponent bits from the input and/or range control ar e added to the shift code programmed by the user to expand the input range. the shift code that the user programs must take the expected exponent range into account i.e. the computed shift control must be reduced by the maximum exponent value. also note that since the exponent shifting reduces the effective size of the integrators, the maximum decimation factor is reduced (see tables 2-4). the integrator bit widths are 69, 62, 53, 44, and 34 for the 1 st through 5 th stages, respectively, while the comb bit widths are all 24. the integrators are sized for decimation factors of up to 512 with 5 stages, 2048 with 4 stages, 32768 with 3 stages, and 65536 with 1 or 2 st ages. higher decimations in the cic should be avoided as they will cause integrator overflow. in the isl5416, the integrators are slightly oversized to reduce the quantizatio n noise at each stage. a cic filter has a gain of r n , where r is the decimation factor and n is the number of stages. because the cic filter gain can become very large with dec imation, an attenuator is provided ahead of the cic to prevent overflow. the 24 bits of mixer output are placed on the low 24 bits of a 69 bit bus (width of the first cic integrator) for a gain of 2 -45 . a 48 bit barrel shifter then provides a gain of 2 0 to 2 47 inclusive before passing the data onto the cic. the overall gain in the pre-cic attenuator can therefore be programmed to be any one of 48 values from 2 -45 to 4, inclusive (see iwa = *005h, bits 25:20). this shift factor is adjusted to keep the total barrel shifter and cic filter between 0.5 and 1.0. the equation which should be used to compute the necessary shift factor is: base shift = max(0, 45 - ceil( log2( r n ) ) - maxexp) maxexp = sum of the maximum exponent range from a floating point input and the range control. cic barrel shifts of greater than 45 will cause msb bits to be lost. most of the floating point modes on the isl5416 make use of the cic barrel shifter for gain. this limits the maximum usable decimation. see floating point input mode section for details. if the cic is bypassed, base shift = 45 - maxexp. maxexp = sum of the maximum exponent range from a floating point input and the range control. table 1. pn generator bit weighting 2^ signal 0 -1 -2 -3 -4 -5 -6 -7 -8 -9 -10 -11 -12 -13 -14 -15 -16 -17 -18 -19 -20 -21 -22 -23 pn sssxxxxxxxxxxxxxxxx gain reg xxxxxxxxxxxxxxxx pf () mf () sin f r ---- - ?? ?? sin ----------------------- - ?? ?? ?? ?? ?? 2n = isl5416
13 table 2. maximum allowed cic decimation vs. number of stages and maximum exponent cic stages maximum floating point or range control exponent 01234567 5 512 445 388 337 294 256 222 194 4 2435 2048 1722 1448 1217 1024 861 724 3 32768 26007 20642 16384 13003 10321 8192 6501 2 65536 65536 65536 65536 65536 65536 65536 65536 1 65536 65536 65536 65536 65536 65536 65536 65536 table 3. maximum cic decimation versus number of stages and maximum exponent to maintain at least 24 bits of dynamic range at the cic output cic stages maximum floating point or range control exponent 01234567 5 512 445 256 128 64 32 16 8 4 2435 2048 1722 1448 1217 1024 861 724 3 32768 26007 20642 16384 13003 10321 8192 6501 2 65536 65536 65536 65536 65536 65536 65536 65536 1 65536 65536 65536 65536 65536 65536 65536 65536 table 4. maximum cic decimation versus number of stages and maximum exponent to maintain at least 20 bits of dynamic range at the cic output cic stages maximum floating point or range control exponent 01234567 5 512 445 388 337 294 256 222 128 4 2435 2048 1722 1448 1217 1024 861 724 3 32768 26007 20642 16384 13003 10321 8192 6501 2 65536 65536 65536 65536 65536 65536 65536 65536 1 65536 65536 65536 65536 65536 65536 65536 65536 isl5416
14 back-end routing fir filter blocks there are two programmable fi r filters in each channel. the main function of the first f ilter, fir1, is to reduce the cic output sample rate and maximize the efficiency of the second filter, fir2. fir2 provides the final filtering for the channel of interest. fir1 can compute up to 32 taps and has programmable 20-bit coefficients, 20-bit data inputs, and 24- bit outputs. fir2 can compute up to 64 taps and has programmable 20-bit coefficients, 20-bit input data, and 24- bit output data. fir1 can compute 4 filter taps per clock and fir2 can compute 8. all of the available taps can be utilized if the overall decimation through the cic and firs is 8 or more. the impulse response of each fir can be symmetric or asymmetric. the decimati on for the each fir is programmable from 1 to 8. to maximize dynamic range, the output bit width of the cic and each fir is 24 bits. a programmable gain stage is provided before each fir to compensate for losses in preceding stages and round to t he 20-bit fir input bit width. gains of 1, 2, 4, or 8 can be programmed. saturation logic is provided to prevent overflow. fir1 includes a half-band filt er mode where a fixed center coefficient of 0.5 is added and the zero valued half-band coefficients are skipped in the computation. this allows fir1 to compute a 15-tap half-band filter in two clock cycles or a 31-tap half-band filter in four clock cycles. note: when loading halfband coefficients, the coefficients must be centered around the fixed center coefficient, e.g. if there are 23 taps, three compute clocks are required, there are 11 on either side of the center and multiplier 1 computes c0, c2, c4, multiplier 2 c6, c8, c10, etc. if there are 19 coefficients, multiplier 2 computes c4, c6, c8 and multiplier 1 computes z, c0, and c2, i.e. an extra zero valued coefficient must be added at each end of the coefficient set to center t he coefficients at the fixed coefficient. the filters will have unity gain if the sum of all coefficients is equal to 1 for coefficient bit weighting 2 0 . . . 2 -19 mag agc loop filter fir2 mux agc mult m u x from cic gain i q fir1 gain gain external agc gain mag detect 1x 2x 4x 8x 1x 2x 4x 8x fifo ihbf hoif up to 32 taps 4 taps/clk hbf mode 8 taps/clk up to 64 taps cic b ypass c ascaded c hannels i q isl5416
15 r e g r e g r e g r e g r e g r e g r e g r e g r e g r e g r e g r e g r e g r e g r e g mux mux mux limit det rnd reg file 1 reg file 0 reg reg file 3 reg file 2 c o n t r o l 0...-23 0...-19 4...-30 x1, 2, 4, 8 fixed 0 / 0.5 a d d limit det rnd mux limit det rnd r e g r e g r e g r e g reg file 1 reg file 0 reg file 7 reg file 6 c o n t r o l 0...-19 4...-30 a d d limit det rnd mux 0...-23 x1, 2, 4, 8 r e g r e g r e g 0...-19 figure 4. fir1 and fir2 block diagrams enable mux mux bypass isl5416
16 agc the automatic gain control (agc) section adds gain to maintain the output signal level at a programmed level. the agc moderates signal level variation at the output of the part and reduces the number of bits that must be carried in any post processing. in the isl5416, the agc follows the channel filtering. the gains through the nco, mixer, and fir filter sections are fixed gains and do not induce am distortion before the large interfering signal s can be filtered out. if large interfering signals are not removed by the filtering prior to the agc, the gain adjustments by the agc can am modulate the large signals and cause am sidebands to fall inside the frequency band of interest. a block diagram of the agc is included in figure 6. the agc consists of a forward gain path and a loop filter path. in the forward gain path, the i/q samples are scaled by the agc forward gain value provided by the loop filter. the forward path gain is divided between a barrel shifter and a multiplier. the overall forward path gain range is 0 to 96.33 db. the barrel shifter provides 0 to 90 db of gain in steps of 6 db. the multiplier provides linear gain between 1.0 and 2.0. saturation is provided if there is overflow. the agc only adds gain. the loop filter path co mputes the gain error, filters it, compares it to gain limits, and provides it to the forward gain path, to the up interface, and to the output section. in the loop filter path the gain error is computed by first computing the magnitude of t he forward path output. the magnitude is then subtracted from a programmable threshold or set point. the re sulting error value is then scaled by a programmable loop gain and integrated and provided to the forward path. programmable limits on the forward gain allow the user to restrict the gain to a smaller range than the 96 db provided. the forward gain control word and programmable gain limits are floating-point numbers consisting of a four-bit exponent that controls the barrel shifte r and a mantissa portion that controls the multiplier. the barrel shifter gain is 2 exp . the multiplier gain is 16 bits, but the two msbs are fixed at ?01? and are not included in the gain control word. the mantissa msb is therefore weighted as 0.5 and the mantissa gain is 1.0 + mant. the total agc gain in db is then: 20*log 10 ( 2 exp * (1.0 + mant)). the agc range is then 0 to 96.33 db for the exp range of 0 to 15 and mant range from 0 to 1. plots of agc gain versus the control word are provided in figures 5a and 5b. the agc gain word is available through the up interface and as a real time output. the gain word is inversely proportional to the received signal strength in the channel. signal strength in db can be easily estimated by complementing the gain word and adding an offset equal to the fixed receive path gain in db. the agc includes a set of counters to synchronize the agc to system timing. the counters can be aligned to the syncinx signals if enabled in iwa *000h. one counter is programmed to count modulo n clocks where n is the length of the time slot. this counter can be restarted with syncinx to align/re-align it with the slot s. a second counter counts out a delay from the syncinx or counter-generated sync. this delays the agc timing from the syncinx signal to compensate for filter group delay or other system delays. a third counter counts out an interval. the interval can be used to divide the slot into fast and slow update periods (timed mode) or into measurement and update periods (sampled mode). the counters can also be disabled and the agc allowed to free run (continuos mode). a programmable data delay can be inserted in the forward data path. the loop f ilter uses the samples into the delay for computing the new forward gain. the forward gain is then applied to the samples coming out of the delay. the gain applied to the output can be cont inuously updated or can be updated under the control of the counters. when updated continuously, the delay causes the forward gain to be based on samples before and after the delayed sample. this moderates large signal variations and minimizes the amount of time that the forward path may be in saturation or be at a small level. the sampled mode is used for burst type signals where the gain adjustment is made during the first part of the burst and then held for the duration of the burst. the programmable delay can be set so that the first samples of the burst are exiting the delay when the gain is updated. in this mode, the gain may have large instantaneous changes, so proper timing alignment is very important. in the timed mode, loop filter continuously updates the forward gain but uses one set of loop gains during part of the burst and another set for the rest of the burst. this allows the time slot to be divided into adapt/hold or fast/slow intervals. the maximum throughput of the agc depends on the mode. in the continuous (counters disabled) and timed modes without delay, the minimum spacing between samples into the agc is 2 clocks. when the delay is enabled, this increases to 4. in the sampl ed mode, the delay is always enabled and the minimum spacing is 4. the minimum spacing is 1 when the agc is bypassed. the agc loop feedback path includes a magnitude computation, an error detector, error scaling (loop gain), and a loop filter. the magnitude computation in the loop filter is a multi-pass operation with one pass computed per clock cycle. the accuracy of the computation d epends on the number of passes. the minimum number of clocks between samples into the agc is 2. there is a gain in the magnitude computation that must be taken into account when programming the agc set point. this gain also depends on isl5416
17 the number of passes in the computation. a listing the accuracy and gain is provided below. with maximum gain and with full scale i and q inputs equal to ~+/-1.0, the maximum out put from the computation is 1.414 * 1.647 = 2.329. the e rror detector subtracts the magnitude from the programmable agc threshold value. the agc threshold value is set in iwa register *009h and should be programmed to k ti mes the desired magnitude of the i/q where k is the gain of the magnitude computation. two adjustment/settling modes are provided in the isl5416. in the mean settling mode, the loop adjusts the gain so that the average magnitude is equal to the programmed set point. in this mode, the error is scaled by the loop gain and integrated to compute the forward gain. the loop settles to the final value asymptotically because the size of the adjustment decreases as the error decreases. the initial settling from large errors is fast, but the final pull in is slower. after the loop has settled, the small adjustment size causes minimal am distortion of the signal. the other settling mode is the median mode. in this mode, the sign of the error is used increase or decrease the gain by a fixed amount. the amount of the adjustment is programmed by the loop gain. the loop settles to the point where there are an equal number of samples above and below the set point. the loop settling is roughly linear in db, but after the loop has settled, the step size remains the sa me, so the amount of am distortion may be objectionable. the isl5416 provides two programmable loop gains, each with a separate attack and decay settling. the micro-proc essor can control the loop gain, or the agc counters can select the loop gain, so a large loop gain can be used for initial settling and a smaller one for tracking. the counters can also select the settling mode, so the median mode can be used at the beginning of each time slot and the mean mode used after the initial settling. the agc loop filter is an a ccumulator (integrator). the output of the accumulator is the forward gain word that controls the barrel shifter and multiplier, closing the loop. there are programmable limits on the accumulator range to minimize settling time by restricting the agc to only that portion of the 96 db range that is needed. the accumulator can be loaded by the microprocessor. the gain load is double buffered-the gain is first loaded into a holding register by the up. the gain is then transferred from the holding register to the accumulator by a write to a special address location or by the syncinx if enabled in iwa *000. the agc can be set to a fixed gain either by setting the both upper and lower gain limits to the desired gain or by setting the loop gain to zero and programming the accumulator directly. the bit weighting for the agc loop is provided in table 86. i table 5. agc magnitude computation accuracy and gain passes error +/- (db) gain 2 0.48 1.581 3 0.13 1.630 4 0.03 1.642 8 0.0001 1.647 isl5416 agc forward gain response 0 12 24 36 48 60 72 84 96 0 32768 65536 98304 131072 163840 196608 229376 262144 code db figure 5a. isl5416 agc forward gain response isl5416 agc forward gain response ma gn i fi ed vi ew 0 1 2 3 4 5 6 0 2048 4096 6144 8192 10240 12288 14336 16384 code db figure 5b. isl5416 agc forward gain response magnified view (actual and ideal linear in db) isl5416
18 figure 6. agc block diagram r e g r e g delay memory 0 - 63 samples barrel shift s a t z -5 clocks mux overflow detect sampled mode update forward gain barrel shift s a t z -6 clocks overflow detect | x | z -4 clocks set point ? mux 0 1 1 0 -a +a 2 mux sign sign barrel shift mux r e g p data limit detector upper / lower limits mux loop gain select mean/med 1 mean/med 2 decay 1 decay 2 at tac k 1 at tac k 2 delay = max(1 sample, 17 clocks) isl5416
19 interpolation half band filt er / re-sampling filter a rate change section follows the agc. this section is used to resample the signal from fir2 to increase the sample rate for finer time resolution and/or to resample the data to another sample spacing. this section consists of an interpolation half-band filter, an interpolating resampling filter, a decimation counter/sampler, a fifo, a set of ncos, and a ?leap? counter. this processing stage allows the filtering in fir2 to be done at the lowest sample rate that meets the nyquist criteria and the data then resampled to the desired final sample rate. the output/input sample rate ratio can be almost any value from 0.125 to >4096. a block diagram is provided below in figure 8. the re-sampling filter (hoif) can accept inputs at any rate up to its maximum output rate of one half the clock rate. preceding the resampler is an interpolation halfband filter. this filter can be used to provide a fixed interpolation by 2 when the resampler is bypassed or, when used with the resampler, to increase the image-free dynamic range of the output. the ihbf can output at up to the clock rate if the resampling filter is bypassed and up to one half the clock rate if the resampling filter is enabled. frequency response plots are provided below for the half-band and resampling filters. an example frequency response for a fir2 response together with the half-band and resampling filters is also provided. the resampling process produces images of the signal at multiples of the input sample rate. large interfering signals must be removed from the spectrum with the cic, fir1, and fir2 filters or the images created from them in the resampling process may cause problems. the level of the images created by resampling process has a fixed dbc level for a given set of filters and sampling ratio. as the signal level in the channel increases and decrease, the images levels will increase and decrease by the same amount. as the ratio of the fir2 output sample rate to the band edge increases, the level, in dbc, of the images decreases. 0 5 10 15 x 10 6 - 10 0 -8 0 -6 0 -4 0 -2 0 0 figure 7a. interpolation half band response 0 0.5 1 1.5 2 2. 5 3 x 10 7 -8 0 -7 0 -6 0 -5 0 -4 0 -3 0 -2 0 -1 0 0 10 figure 7b. ihbf (interpolate by 2) and re- sampler (interpolate by 2) 0 5 10 15 x 10 6 -8 0 -7 0 -6 0 -5 0 -4 0 -3 0 -2 0 -1 0 0 10 figure 7c. individual and composite responses (fir2 output at 7.68 mhz with ihbf, interpolate by 2) isl5416
20 two ncos and two counters se t the sample rates through the rate change section. nco1 sets the output sample rate of the resampling filter. nco1 is 48 bits and is updated at the clock rate, so its output frequency is: fout 1 = fclk * n 1 / 2 48 , where n1 is the 48-bit programming word. the carry output of the phase accumulator is us ed as the output clock, so there can be one clock period of jitter. nco2 is programmed for the input sample rate to the resampler (equals the half- band filter output rate). nco2 is updated at the nco1 output rate. nco2 controls the phase of the resampling filter. this nco also has a 48-bit phase accumulator. the equation for programming the output frequency of nco2 is: fout 2 = fout 1 * n 2 / 2 48 = fclk * (n 1 / 2 48 ) * (n 2 / 2 48 ) when the resampling filter is enabled and fout 2 = fclk * n 2 / 2 48 when the resampling filter is bypassed. nco2 can have one output sample period (fout 1 period) of jitter (one clock period when the hoif is bypassed). a static phase offset can be programmed for nco2. the range of the phase offset is 0 to 2 nco2 output sample periods (0 - 2 resampling filter input sample periods). the programming resolution is 1/256 of a resampling filter input sample period. this programmable offset allows the user vary the group delay of one channel relative to another in very fine increments to compens ate for differences in system delays. if the resampler is not needed for rate change, it can be used for phase shifting by setting bit 22 in iwa *001h. while the 48-bit phase accumulators provide very good frequency programming resolution, at some input/output sample rate ratios, there will be a slow phase drift due to the finite word length. to correct for this, a ?leap? counter is provided to reset the phase of the ncos after a programmed interval to remove any accumulated error. the leap counter is 32 bits. if properly programmed, this phase correction will not be seen in the output of the part. the input rate to the ihbf/rs section must match the output sample rate of fir2, i.e. the ou tput rate of nco2 must equal the input sample rate of the part divided by the decimation factors in the cic, fir1, and fir2. the leap counter can guarantees this over the long term , but due to the jitter of the phase accumulator outputs, a fifo is provided to guarantee that there are no dropped samples. the fifo is filled at the output sample rate of the agc and is emptied by fout 2 (or fout 2 /2 if the ihbf is enabled). after reset, the fifo is filled to a depth of two before the ncos are enabled. this minimum fill depth guarantees that there are enough samples in the fifo that the fifo never empties or overflows due to nco jitter if the ncos and leap counter are properly programmed. fifo reads are enabled after an 0 0.5 1 1. 5 2 2. 5 3 x 10 7 -8 0 -7 0 -6 0 -5 0 -4 0 -3 0 -2 0 -1 0 0 10 figure 7d. individual and composite responses (fir2 output at 7.68 mhz with ihbf, interpolate by 2 and re-sampler, interpolate by 2 0 5 10 15 x 10 6 -8 0 -7 0 -6 0 -5 0 -4 0 -3 0 -2 0 -1 0 0 10 figure 7e. fir2 and ihbf composite response 0 0. 5 1 1.5 2 2.5 3 x 10 7 -8 0 -7 0 -6 0 -5 0 -4 0 -3 0 -2 0 -1 0 0 10 figure 7f. fir2, ihbf and resampler composite response isl5416
21 additional 0 to 3 input samples as programmed by the user. this additional depth provides for additional programmable group delay. the additional fifo depth can only be programmed at reset. because the ncos are enabled after a depth of 2 is reached, the data into the ihbf/resampler is zeroed until the programmed fill depth is reached. if both the half-band and resampling filters are enabled, the programmable fifo depth, t ogether with the nco2 phase offset, provides from 0 to 4 fir2 output sample periods of programmable group delay in 1/512 increments. because the ihbf and rs combination can only interpolate, for resampling ratios <1, the signal must first be interpolated to a multiple of the desired sample rate and then decimated to the final rate. a decimation counter is provided after the resampling filter to down sample to the desired rate. the nco1 and nco2 frequencies are programmed in iwa = *011h - *014h. these registers are double buffered. the up writes to a holding register. da ta is then transferred to the active registers by a write to iwa = *017h or by a syncinx if enabled in iwa = *000h. the gain data from the agc is not interpolated. the output of the agc is sampled with each i/q sample to the output section. r e g r e g r e g r e g r e g r e g r e g r e g r e g r e g r e g r e g r e g r e g mux round fifo ihbf mux fractional resampling interpolation filter 0...-15 wr rd depth = 2 + (0 to 3) samples divide by 1 or 2 reg leap counter reg phase offset 0 to 511/256 samples fclk nco2 co nco1 n x fout en divide by n fout fihbf_out fihbf_in fill depth before enabling reads 0, 1, 2, or 3 figure 8. ihbf and resampler block diagram isl5416
22 data output formatter section four 16-bit output data busse s are provided on the isl5416. all of the busses share a common output clock, clko1, which is derived from clkc. clko2 signal is provided for easier board routing or for the differential outputs. each bus has an output sync which is typically used as a frame sync. each bus can be divided into two 8-bit busses if desired. when a new data sample is available from a channel, it starts a time slot counter that sequences through up to 8 output time slots. the data type for each time slot is programmable as well as the fsyncx assertion. the data from more than one channel can be multiplexed through the same output bus if channels are synchronized. the data from channels 0 and 1 and from channels 2 and 3 can be multiplexed directly. multiplexing channels 0 and 1 with 2 and 3 is done by oring multiplexer outputs together. see figures 10 and 11. this means that related channels (such as diversity channels) should be grouped into channels 0 and 1, or into channels 2 and 3 for ease of data routing. the data type, sync assertion, and bus routing are programmed in registers 0*01h through 0*04h. two of the eight time slots are prog rammed in each location. the i/q data from each channel is rounded to 4, 6, 8, 12, 16, 20 or 24 bits at the output of the channel. the agc gain can be rounded to 8, 12, or 16 bits. a 24-bit output is provided to the output section for i and q data and a 16-bit output is provided for the agc data. the data is msb justified in the output bus and the lsbs below the programmed number are zeroed. 24 bits of i/q data is available from the agc if the ihbf/hoif is bypassed. i/q are 16 bits if the ihbf/hoif section is enabled. serial outputs are available. see gwa = 0000h, iwa = 0*06h, 0*07h, and 0*08h. 01ac upper 16 x 8-bit mux 01ac lower 16 x 8-bit mux (15:8) (7:0) i0 (23:16) (15:8) (7:0) q0 (23:16) (23:16) (23:16) (15:8) (7:0) (15:8) (7:0) (15:8) (7:0) i1 agc0 q1 (15:8) (7:0) agc1 (15:8) (7:0) i0 (23:16) (15:8) (7:0) q0 (23:16) (23:16) (23:16) (15:8) (7:0) (15:8) (7:0) (15:8) (7:0) i1 agc0 q1 (15:8) (7:0) agc1 (15:8) (7:0) 01 ac upper channels outputs output byte figure 9. multiplexing channels isl5416
23 aout(15:8) aout(7:0) dout(15:8) dout(7:0) oe_23d oe_01d oe_23d oe_01d chan 0 data chan 1 data chan 2 data chan 3 data 64 64 64 64 8 8 8 8 01ac 23bd oe_23a oe_01a oe_23a oe_01a high high low low bout(15:8) bout(7:0) oe_01b oe_23b chan 0 data chan 1 data 64 64 8 8 01bd oe_01b oe_23b high low cout(15:8) cout(7:0) oe_23c oe_01c oe_23c oe_01c chan 2 data chan 3 data 64 64 8 8 23ac high low mux #3 mux #2 mux #1 mux #4 figure 10. output routing isl5416
24 serial data output the serial data output control r egister contains sync position and polarity (ssynca, b, c or d), channel multiplexing, and scaling controls for the sd1x and sd2x (x = a, b, c or d) serial outputs (see iwa registers 0*06 - 0*08h). serial data output time slot content/format registers these two registers are used to program the content and format of the serial data output sequence time slots (see iwa registers 0*07h and 0*08h). there are four data time slots that make up a serial data output stream. the number of data bits and data format of each slot is programmable as well as whether there will be a sync generated with the time slot (the syncs are only associated with the sd1 serial outputs). any of three types of data or zeros can be chosen for each time slot. seven bits are used to specify the content and format of each slot. channel routing mask the multiplexing mask bits for each channel (see microprocessor interface section, iwa register 0*06h bits 19:16 for sd1x or bits 15:12 for sd2x) can be used to enable that channel?s output to any of the four serial outputs. these bits control and gates that mask off the channels, so a zero disables the channel?s c onnection to that output. to configure more than one channel's output onto a serial data output, the sd1 serial outputs and syncs from each channel (0,1, 2 and 3) are brought to each of the sd1 serial output sections and the sd2 se rial outputs are brought to each of the sd2 serial output sections (the syncs are only associated with the sd1 serial outputs). there, the four outputs are and-ed with the multiplexing mask programmed in the serial data output control registers of channels 0 thru 3 and or-ed together. by gating off the channels that are not wanted and delaying the data from each desired channel appropriately, the channels can be multiplexed into a common serial output stream. it should be noted that in order to multiplex multiple channels onto a single serial data stream the channels to be multiplexed must be synchronous. microprocessor interface the isl5416 microprocessor ( p) interface consists of a 16- bit bidirectional data bus, p(15:0), three address pins, add(2:0), a write strobe (wr ), a read strobe (rd ) and a chip enable (ce ). indirect addressing is used for control and configuration of the isl5416. the processor interface to the isl5416 is a mixture of direct and indirect addressing. to minimize the amount of processor address space and bus routing, there are only eight 16-bit direct address locations. two of these are used to access an internal 32-bit bu s. to write data to internal indirect locations, the data is first written to direct addresses 0 (bits 15:0) and 1 (bits 31:16) . the internal address is then written to direct address 2. wh en the address is written, a synchronization circuit generates an internal write strobe, synchronized to the clock, to clock the data into the target register. the synchronization process requires 4 clock cycles, so data should not be wri tten to direct addresses 0 or 1 for four clock cycles after a wr ite to address 2. to read data from internal locations, the internal address is first written to direct address 3. the data can then be read from direct addresses 0 (15:0) and 1 (31:16). the indirect address register is shared between direct addresses 2 and 3, with only writes to address 2 generating write strobes. because of this, the address does not have to be re-written to verify a write unless broadcasting data to more than one channel. direct address 2 is used for status when read. the status bits are defined in table of microprocessor direct read/write addresses. direct addresses 4 through 7 are used for fast read access. addresses 4 through 7 (one address per channel) are used for sequenced read (fifo-like) addresses. the user can program the order that the data would be read from the part. the user can select i, q, agc gain (real time or sampled), and two types of data from the range control circuit. when a new output is available, the data type pointer is reset to the first data type. after each read , the pointer is incremented to the next data type. to signal a new output on the channel, a signal can be routed to the clko2/intrpt pin. a channel is enabled to generate interrupt in address iwa = 0*0ah, bit 31. if separate interrupt signals are required for each channel, the fsyncx pins can be used. the indirect address space is divided into top level or global locations for parameters that are shared between channels or i/os, i/o control locations, and channel control parameters. the global locations are between addresses 0000h and 00ffh. the i/o control locations are between 0100h and 0fffh. bits 11, 10, 9, and 8 select i/o busses d, c, b, and a, respectively. what this means is that by setting a single address bit of 11:8, the control register is written for that i/o control section. by setting more than one bit, the same data is written to the corresponding registers of more than one i/o control section. reads must select only one i/o control section. channel control registers are located between addresses 1000h and ffffh. bits 15, 14, 13, and 12 select channels 3, 2, 1, and 0, respectively. the user can write to individual channels or to multiple channels at once by setting the appropriate channel select bit, 15:12. read addresses must specify a single channel. isl5416
25 to write to the internal registers: 1. load the indirect write holding registers at direct address add(2:0) = 0 and 1 with the data for the internal register (32 bits). 2. write the indirect write address of the internal register being addressed to direct address add(2:0) = 2 (note: a write strobe to transfer the co ntents of the indirect write holding register into the tar get register specified by the indirect address will be generated internally). 3. wait 4 clock cycles before performing the next write to the indirect write holding registers. to read internal registers: 1. write the indirect read addr ess of the internal register being read to direct address add(2:0) = 3. 2. perform a read of the indirect read holding registers at direct address add(2:0) = 0 and 1. note: after an indirect write to a single channel, the data can be read at direct addresses 0 and 1 after 4 clock periods. jtag jtag: the ieee 1149.1 joint test action group boundary scan standard operational codes shown in table 2 are supported. a separate applic ation note is available with implementation details jtag op codes supported bits 0, 2, 4, 6 are the or-ing together of the individual data path. saturation bits listed in *01fh (table 14). these bits are latched when set. awrite to *01f clears the bit(s). table 6. instruction op code extest 0000 idcode 0001 sample/preload 0010 intest 0011 bypass 1111 table 7. status bits read direct address 2 for (15:0) 15:8 revision code. 0x00h = prototype silicon. 0x01h = production silicon. 7:6 channel 3 status. bit 7 is always zero, bit 6 indicates data path saturation. 5:4 channel 2 status. bit 5 is always zero, bit 4 indicates data path saturation. 3:2 channel 1 status. bit 3 is always zero, bit 2 indicates data path saturation. 1:0 channel 0 status. bit 1 is always zero, bit 0 indicates data path saturation. isl5416
26 power-up sequencing the isl5416 core and i/o blocks are isolated by structures which may become forward bias ed if the supply voltages are not at specified levels. during the power-up and power-down operations, differences in the starting point and ramp rates of the two supplies may cause current to flow in the isolation structures which, when prolonged and excessive, can reduce the usable life of the device. in general, the most preferred case would be to power-up the core and i/o structures simultaneously. however, it is also safe to power-up the core prior to the i/o block if simultaneous application of the supplies is not possible. in this case, the i/o voltage should be applied in 10ms to 100ms nominally to preserve supply component reliability. bringing the core and i/o supplies to their respective regulation levels in a maximum time frame of a 100ms, moderates the stresses placed on both, the power supply and the isl5416. table of microprocessor direct read/write addresses add(2:0) pins register description 0 wr indirect write data bus (15:0), ?master (15:0)?. 1 wr indirect write data bu s (31:16), ?mas ter (31:16)?. 2 wr indirect write address register for in ternal target register (generates a wr ite strobe to transfer contents of the write holding register into the target register specified by the indirect address, see also tables of indirect address registers). 3 wr indirect read address register (used to select the r ead source of data - uses the same register as direct address 2 but does not generate a write strobe). 4 wr reserved 5 wr reserved 6 wr reserved 7 wr reserved 0 rd indirect read, bits 15:0. 1 rd indirect read, bits 31:16. 2 rd status (15:0). 3 rd reserved 4 rd channel 0 sequenced read address. 5 rd channel 1 sequenced read address. 6 rd channel 2 sequenced read address. 7 rd channel 3 sequenced read address. isl5416
27 table of indirect read an d write address registers note: these indirect read addresses are repeated for each chan nel. in the addresses below, the * fields are the channel and i/o select nibble. these bits of the in direct address select the target i/o or ch annel register for the data being read. table 8. top - global writes fedcba9876543210 description 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 top level, common output controls and channel-to-channel data routing 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 range control to eout mapping 0000000000000010up test input data, test output enables 0000000000000011up test data strobe 0000000000000100synco strobe generation 0000000000000101synco with internal syncin1 feedback 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 software hard reset 0007 - 00ff reserved table 9. top - global reads fedcba9876543210 description 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 control register 0000h 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 control register 0001h 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 control register 0002h 0000000000000011n/a (strobe) 0000000000000100n/a (strobe) 0000000000000101n/a (strobe) 0000000000000110n/a (strobe) 0007 - 00ff reserved table 10. xin, xout, xrng writes (x = a, b, c, d as specified by * nibble bits) fedcba9876543210 description 0000 ****0 0000000input format 0000 ****0 0000001parallel output routing 1 - time slots 0, 1 0000 ****0 0000010parallel output routing 2 - time slots 2, 3 0000 ****0 0000011parallel output routing 3 - time slots 4, 5 0000 ****0 0000100parallel output routing 4 - time slots 6, 7 0000 ****0 0000101range control syncinx controls 0000 ****0 0000110serial output control 0000 ****0 0000111serial output sd1x time slots 0000 ****0 0001000serial output sd2x time slots 0000 ****0 0001001range control dc blocking filter 0000 ****0 0001010sequenced read order **** 0*0b - 0*0f r eserved 0000 ****0 0010000range control control isl5416
28 x = these control registers are shared between inputs a and b and between inputs c and d. 00000*0*00010001range control slot period, delay x 00000*0*00010010range control integration time, number of integrations x 00000*0*00010011range control lower threshold, delta x 00000*0*00010100range control u pper threshold, delta x 00000*0*00010101range control immediate threshold, delta x 00000*0*00010110range control bias x 0000 ****0 0010111range control attenuation control codes 0 - 3 0000 ****0 0011000range control attenuation control codes 4 - 7 0000 ****0 0011001range control up attenuation accumulator load 0000 ****0 0011010range control a ccumulator load strobe 0000 ****0 0011011range control a ccumulator read strobe 0000 ****0 0011100range control integrator read sample strobe 0000 ****0 0011101range control dc offset read sample strobe 0*1e - 0*ff reserved table 10. xin, xout, xrng writes (x = a, b, c, d as specified by * nibble bits) fedcba9876543210 description table 11. xin, xout, xrng reads (x = a, b, c, d as specified by * nibble bits) fedcba9876543210 description 0000* * * *00000000control register 0*00 0000* * * *00000001control register 0*01 0000* * * *00000010control register 0*02 0000* * * *00000011control register 0*03 0000* * * *00000100control register 0*04 0000* * * *00000101control register 0*05 0000* * * *00000110control register 0*06 0000* * * *00000111control register 0*07 0000* * * *00001000control register 0*08 0000* * * *00001001control register 0*09 0000* * * *00001010control register 0*0a 0*0b - 0*0f reserved 0000* * * *00010000control register 0*10 0 0 0 0 0 * 0 * 0 0 0 1 0 0 0 1 control register 0*11, * = 1, 4 only 0 0 0 0 0 * 0 * 0 0 0 1 0 0 1 0 control register 0*12, * = 1, 4 only 0 0 0 0 0 * 0 * 0 0 0 1 0 0 1 1 control register 0*13, * = 1, 4 only 0 0 0 0 0 * 0 * 0 0 0 1 0 1 0 0 control register 0*14, * = 1, 4 only 0 0 0 0 0 * 0 * 0 0 0 1 0 1 0 1 control register 0*15, * = 1, 4 only 0 0 0 0 0 * 0 * 0 0 0 1 0 1 1 0 control register 0*16, * = 1, 4 only 0000* * * *00010111control register 0*17 0000* * * *00011000control register 0*18 0 0 0 0 * * * * 0 0 0 1 1 0 0 1 range control load value read (holding register) isl5416
29 0000* * * *00011010n/a (strobe) 0 0 0 0 * * * * 0 0 0 1 1 0 1 1 range control accumulator read after a write to this location to stabilize 0000* * * *00011100range select integrator read after a write to this location to stabilize 0 0 0 0 * * * * 0 0 0 1 1 1 0 1 range select dc offset read after a write to this location to stabilize 0*1e - 0*ff reserved table 11. xin, xout, xrng reads (x = a, b, c, d as specified by * nibble bits) (continued) fedcba9876543210 description table 12. channel control registers (writes) fedcba9876543210 description * * * *000000000000channel sync control * * * * 0 0 0 0 0 0 0 0 0 0 0 1 channel data path, control ? * * * * 0 0 0 0 0 0 0 0 0 0 1 0 carrier center frequency holding register ? * * * * 0 0 0 0 0 0 0 0 0 0 1 1 carrier phase offset * * * *000000000100carrier frequency load strobe * * * *000000000101nco/cic control * * * * 0 0 0 0 0 0 0 0 0 1 1 0 pn noise level * * * * 0 0 0 0 0 0 0 0 0 1 1 1 fir1, fir2 cintrol, re-sampling filter output decimation * * * *000000001000agc control * * * *000000001001agc set point * * * * 0 0 0 0 0 0 0 0 1 0 1 0 agc upper and lower gain limits * * * *000000001011agc loop gains * * * *000000001100agc counter preloads 1 * * * *000000001101agc counter preloads 2 * * * *000000001110agc up gain loading register ? * * * *000000001111agc up gain loading strobe * * * * 0 0 0 0 0 0 0 1 0 0 0 0 agc strobe to sample gain for read * * * *000000010001re-sampling filter nco1 - output rate 47:16 ? * * * *000000010010re-sampling filter nco1 - output rate 15:0 ? * * * *000000010011re-sampling filter nco2 - input rate 47:16 ? * * * *000000010100re-sampling filter nco2 - input rate 15:0 ? * * * * 0 0 0 0 0 0 0 1 0 1 0 1 re-sampling filter nco2 - phase offset/delay * * * *000000010110re-sampling filter leap counter period * * * *000000010111re-sampling filter frequency update strobe * * * *000000011000channel soft reset strobe * * * *000000011001up strobe to update data path register * * * *000000011010advance strobe * * * *000000011011retard strobe *01c -*01e reserved * * * * 0 0 0 0 0 0 0 1 1 1 1 1 channel status register clear strobe isl5416
30 *020 -*07f reserved * * * *000100000000fir1 coefficients thru * * * *000100011111 *120 - *1ff reserved * * * *001000000000fir2 coefficients thru * * * *001000111111 *240 - *fff reserved note: the controls marked as ? ? ? are double buffered, i.e. the data is loaded into a holding register by the up and then transferred to an active register by either a write to an indirect locati on that generates an update strobe or by a syncinx signal if enabl ed in iwa = *000h. table 12. channel control registers (writes) (continued) fedcba9876543210 description table 13. channel control registers (reads) fedcba9876543210 description * * * *000000000000control register *000 * * * *000000000001control register *001 * * * * 0 0 0 0 0 0 0 0 0 0 1 0 carrier active center frequency * * * *000000000011control register *003 * * * *000000000100n/a (strobe) * * * *000000000101control register *005 * * * *000000000110control register *006 * * * *000000000111control register *007 * * * *000000001000control register *008 * * * *000000001001control register *009 * * * *000000001010control register *00a * * * *000000001011control register *00b * * * *000000001100control register *00c * * * *000000001101control register *00d * * * * 0 0 0 0 0 0 0 0 1 1 1 0 agc gain is read at direct addresses 4 - 7 after sampling with a wtite to iwa = *010h * * * *000000001111n/a (strobe) * * * *000000010000magnitude (after sampling with a write to this location) * * * * 0 0 0 0 0 0 0 1 0 0 0 1 active resampler nco1 frequency 47:16 * * * * 0 0 0 0 0 0 0 1 0 0 1 0 active resampler nco1 frequency 15:0 * * * * 0 0 0 0 0 0 0 1 0 0 1 1 active resampler nco2 frequency 47:16 * * * * 0 0 0 0 0 0 0 1 0 1 0 0 active resampler nco2 frequency 15:0 * * * *000000010101control register *015 * * * *000000010110control register *016 * * * *000000010111n/a (strobe) * * * *000000011000n/a (strobe) isl5416
31 * * * *000000011001n/a (strobe) * * * * 1 1 0 1 0 n/a (strobe) * * * *000000011011n/a (strobe) * * * *000000011100n/a (strobe) *01d - *03e unused * * * *000000011111channel status *03f - *07f unused * * * *000100000000fir1 coefficients thru * * * *000100011111 *120 - *1ff unused * * * *001000000000fir2 coefficients thru * * * *001000111111 *240 - *fff unused table 13. channel control registers (reads) (continued) fedcba9876543210 description table 14. channel status bits (*01fh) once set, status bits stay set until cleared by a write to *01fh bits description 15 fifo crossed threshold 14 fifo empty 13 fifo full 12 saturation in ihbf (i or q) 11 saturation in q channel 24 -> 16 bit rounder (to fifo prior to ihbf) 10 saturation in i channel 24 -> 16 bit rounder (to fifo prior to ihbf) 9 unused. 8 saturation in agc forward gain path (i or q) 7 saturation in q channel fir2 6 saturation in i channel fir2 5 saturation in q channel fir1 to fir2 gain block 4 saturation in i channel fir1 to fir2 gain block 3 saturation in q channel fir1 2 saturation in i channel fir1 1 saturation in q channel cic to fir1 gain block 0 saturation in i channel cic to fir1 gain block note: fir saturation bits will indicate saturation if an inte rmediate accumulation exceeds 1.0, even though, the final result i s less than 1.0. the headroom for intermediate sums of products is up to 16 (2 4 ). isl5416
32 tables of top level registers in the tables below ?reset st ate? indicates the register contents after a hw reset or a sw hard reset. unless noted, a soft channel reset does not clear register contents. a soft channel reset does clear the slave registers of a master/slave pair (such as channel data path control) but does not clear the master. this register controls the out put sync signal polarity, the output clock rate and polarity, output data modes, and channel cascading. table 15. common output control functions (gwa = 0000h) reset state = 0x00000001h p(31:0) function 31 enable serial output. 1 = serial output mode is enabled. the dout parallel data bus is replaced with 4 serial output busses -- one per channel. see 0 *06h - 0*08h and table 30. 30 reserved. set to 0. 29 sclk reset. 1 = serial clock divider is reset by syncinx if the reset serial output bit is set in iwa = *000h, bits 13 or 29 of any of the 4 channels. 28 sclk polarity. 1 = low to high transitions at the center of the data bit. 0 = high to low transitions at the center of the data bit. 27:25 sclk rate. 000 = sclk disabled. 001 = input clock rate. 010 = input clock rate / 2. 011 = input clock rate / 4. 100 = input clock rate / 8. 101 = input clock rate /16. 24:21 reserved. set to 0. 20 route channel 2 to channel 3. 1 = route the output of fir2 of channel 2 to the i nput of the cic to fir1 gain block in channel 3. 19 route channel 1 to channel 2. 1 = route the output of fir2 of channel 1 to the i nput of the cic to fir1 gain block in channel 2. 18 route channel 0 to channel 1. 1 = route the output of fir2 of channel 0 to the i nput of the cic to fir1 gain block in channel 1. 17 channel 0 external agc source select. 0 = when bit 14 is set, channel 1 controls channel 0 gain. 1 = when bit 14 is set, channel 3 controls channel 0 gain. 16 channel 2 external/internal gain control. 1 = channel 2 gain is controlled by channel 3. 15 channel 1 external/internal gain control. 1 = channel 1 gain is controlled by channel 3. 14 channel 0 external/internal gain control. 1 = channel 0 gain is controlled by channel 3 or channel 1 depending on the state of bit 17. 13 clko2 or intrpt . 1 = clko2/intrpt is intrpt (2 clock period wide pulse). 0 = clko2/intrpt is clko2. note: for intrpt iwa = 0*0ah, bit 31 must be set for the channel that is the interrupt source.. 12 fsyncx polarity. 0 = active high 1 = active low. isl5416
33 this register (iwa = 0001h) ro utes the bits from the four range select blocks to the four nibbles of the eout bus. each nibble of the eout bus can be controlled by any of the range selectors. each range select block contains a mapping table with three input bits and eight output bits. the three input bits represent 0 to 42 db of attenuation in 6 db steps. the programming of the input to output bit map depends on the type of attenuator or vga used. typically four bits would be sufficient for an attenuator while 8 bits might be needed for a d/a controlled vga. an example programming for 4 bits per attenuator control is: 0x00001357h range controls a, b, c, and d control eout bits 15:12, 11:8, 7:4, 3:0, respectively. for 8 bits per attenuator control: 0x00000145 range controls a and c control eout bits 15:8 and 7:0, respectively. 11 clko2 polarity. 1 = low to high transition in the middle of data period. 0 = high to low transition in the middle of data period. if bit 13 is set: 1 = active high interrupt 0 = active low interrupt. 10 clko1 polarity. 1 = low to high transition in the middle of data period. 0 = high to low transition in the middle of data period. 9:8 hold clko after reset. 11 = hold clko after hard reset until syncin2. 10 = hold clko after hard reset until syncin1. (to re-sync the cloc k after it has been started, disable it by setting these bit s to 00 and then set the bits back to 1x and apply syncinx). 01 = enable clko after programming this regist er (this applies only when the clko rate is not equal to clkin rate; there is no need to align the start if clko = clk, and if that is the case, it starts imm ediately after programming). 00 = disable clko. 7:0 clko1 rate. clkin / n for n = 1 to 16. 1 = 0x80h or 0x00h 2 = 0x81h 3 = 0x82h 4 = 0x93h 5 = 0x94h 6 = 0xa5h 7 = 0xa6h 8 = 0xb7h 9 = 0xb8h 10 = 0xc9h 11 = 0xcah 12 = 0xdbh 13 = 0xdch 14 = 0xedh 15 = 0xeeh 16 = 0xffh table 15. common output control functions (gwa = 0000h) reset state = 0x00000001h p(31:0) function isl5416
34 a test register is provided for the processor to input data to the part. the register can be se lected as the input for any or all of the channels. the test input is 16 bits of data plus 3 bits of exponent. the input enable is also controllable by the up. the choices are always active, always inactive, and one clock wide input enable each time the up writes to iwa = 0003h. this last choice is for inputting test data via the up interface. the input enable choice is done in the channel at iwa = *001h, bits 19:17. table 16. eout (range control) source select (gwa = 0001h) reset state = 0x00000000h p(31:0) function 31:16 unused. 15:12 eout(15:12). 1xxx = set to 0000. 0000 = ah (a range select mapped bits 7:4). 0001 = al (a range select mapped bits 3:0). 0010 = bh (b range select mapped bits 7:4). 0011 = bl (b range select mapped bits 3:0). 0100 = ch (c range select mapped bits 7:4). 0101 = cl (c range select mapped bits 3:0). 0110 = dh (d range select mapped bits 7:4). 0111 = dl (d range select mapped bits 3:0). see iwa = 0*17h and 0*18h. 11:8 eout(11:8). see bits 15:12. 7:4 eout(7:4). see bits 15:12. 3:0 eout(3:0). see bits 15:12. table 17. test input data register (gwa = 0002h) reset state = 0x00000000h p(31:0) function 31:26 reserved. set to 0. 25 reserved. set to 1. 24 reserved. set to 0. 23 enable testout d(15:0) to dout(15:0). set to 0 for normal operation. see table 18 below. 22 enable testout c(15:0) to cout(15:0). set to 0 for normal operation. see table 18 below. 21 enable testout b(15:0) to bout(15:0). set to 0 for normal operation. see table 18 below. 20 enable testout a(15:0) to aout(15:0). set to 0 for normal operation. see table 18 below. 19 reserved. set to 0. 18:16 exponent. 000 = add 0 db of gain. 111 = add 42 db of gain. 15:0 data. twos complement. table 18. testout x(15:0) bus signals (can be or?d with no rmal output on xout by setting bits 23:20, gwa = 0002h) 15 range control end of integration period signal. 14 range control end of delay signal 13 range control counter load sign al (sync and slot counter generated) 12 range control magnitude greater than immediate threshold 11 range control average magnitude greater than threshold 2 10 range control average magnitude smaller than threshold 3 9 resampler nco2 2x carry out isl5416
35 for synchronization of the isl5416 channels to system timing and/or to the processing of other isl5416 channels, one synchronization output (synco) and two synchronization inputs (syncin1 and syncin2) are provided. the synco of one isl5416 might be connected to the syncin1 of all the isl5416s to allow the up to synchronously start or update parameters in all of the isl5416s. a write to iwa = 0005h also internally routes the synco to the syncin1 input with the same delay as connecting the synco pin to the syncin1 pin. for alignment to system timing, the syncinx pins can be connected to any one-clock-wide signal synchronous to clkc. a second synchroniza tion input, syncin2, is provided to synchronize different channels to different event or to allow the processor to control certain events and the system timing to control others. there are three resets to the isl5416 -- the reset pin, the chip hard reset (iwa = 0006h), and the soft channel reset (iwa = *018h). the pin reset and chip hard reset have the same effect. the soft channel reset only affects the selected channel and does not reset the control registers, only the slave (active) registers of master/slave pairs. 8 resampler nco1 2x carry out 7 agc counter load signal (sync and slot counter generated) 6 agc end of delay counter 5 agc loop gain select signal 4 agc update forward gain signal 3 carrier nco msb (l.o. test point) 2:0 fifo read address (2:0); (fifo depth, empty = 0) table 18. testout x(15:0) bus signals (can be or?d with no rmal output on xout by setting bits 23:20, gwa = 0002h) table 19. test input strobe (gwa = 0003h) reset state = inactive n/a function n/a strobe. a write to this location generates a one-clock-wi de test input enable (for use with the test input register). table 20. synco (gwa = 0004h) reset state = inactive n/a function n/a strobe. a write to this location generat es a one-clock-wide pulse on the synco pin. table 21. synco with internal feedback (gwa = 0005h) reset state = inactive n/a function n/a strobe. a write to this location generates a one-clock-wide pulse on the synco pin that is also internally fed back to the syncin1 pin. table 22. chip hard reset via software (gwa = 0006h) reset state = inactive n/a function n/a reset strobe. writing to this location generates a hardware re set, resetting all control registers. identical function to th e reset pin. isl5416
36 tables of i/o registers table 23. channel input format (iwa = 0*00h) reset state = 0x00000000h p(31:0) function 31:16 unused. 15 fixed gain mode in vga. 14 reserved. set to 1. 13 invert input clock. 1 = high -> low edge of the input clock samples input data. 0 = low -> high edge of the input clock samples input data. 12 input clock source. 1 = clkc 0 = clka for ain (address 0000 xxx1 0000 0000) clkb for bin (address 0000 xx1x 0000 0000) clkd for din (address 0000 1xxx 0000 0000) clkc is always used for cin the rising edge of clkc is always used fo r the channel processing and output timing. 11 input format. 1 = offset binary. 0 = 2?s complement 10 gated & interpolated. 1 = input interpolated. 0 = input gated. 9:7 demux delay. 000 = no delay, take sample aligned with the enable. 111 = take 7th sample after the enable. 6:5 reserved. set to 0. 4 lsb to msb swap on data. 1 = input bus reversed msb for lsb -- xin(0) = msb. 0 = input bus normal xin(16) = msb provided to simplify circuit board routing. 3:2 fixed/floating point mode. 00 = 16-bit fixed point (bits 16:1, 0 unused). 01 = 14/3 floating point (bits 16:3 mantissa, 2:0 exponent). 10 = 15/2 floating point (bits 16:2 mantissa, 1:0 exponent). 11 = 16/1 floating point (bits 16:1 mantissa, 0 exponent). 1:0 reserved. set to 0. isl5416
37 table 24. xout data versus time slot rout ing, time slots 0, 1 (iwa = 0*01h) reset state = 0x00000000h x = a, b, c, d (descriptions below are for the channels 0/1 multiplexer iwa = 0101h (mux 01ac). see figure 10 for corresponding signals and buses for the other multiplexers) p(31:0) function 31:30 reserved. set to 0. 29:28 xout timing source. the output slot counter is started by a new sample from 00 = channel 0 01 = channel 1 10 = channel 2 11 = channel 3 this allows two output busses to be used by one channel if the other channel is disabled. 27 invert (ones complement) xout bus. 1 = ones complement xout(15:0). by setting this bit for bout or dout, the aout and bout and/or cout and dout can provide complementary output signals. 26 assert fsyncx during slot 1. 1 = assert fsyncx for this time slot. 0 = no fsyncx. 25 enable to cout during slot 1. (oe_01c) 1 = route the data selected in 23:16 and sync enabled in bit 26 to the cout and syncc pins. provided for multiplexing a and b w ith c and d channels. 24 enable to aout during slot 1. (oe_01a) 1 = route the data selected in 23:16 and sync enabled in bit 26 to the aout and synca pins. provided for multiplexing a and b w ith c and d channels. 23:20 slot 1, high byte. data output from the a high byte multiplexer for slot 1. see table 29 for coding. 19:16 slot 1, low byte. data output from the a low byte multiplexer for slot 1. see table 29 for coding. 15:11 reserved. set to 0. 10 assert fsyncx during slot 0. 1 = assert fsyncx for this time slot. 0 = no fsyncx. 9 enable to cout during slot 0. 1 = route the data selected in 7:0 and sync enabled in bit 10 to t he cout and syncc pins. provided for multiplexing a and b wit h c and d channels. 8 enable to aout during slot 0. 1 = route the data selected in 7:0 and sync enabled in bit 10 to the aout and synca pins. provided for multiplexing a and b wit h c and d channels. 7:4 slot 0, high byte. data output from the a high byte multiplexer for slot 0. see table 29 for coding. 3:0 slot 0, low byte. data output from the a low by te multiplexer for slot 0. see table 29 for coding. table 25. xout data versus time slot routing, time slots 2, 3 (iwa = 0*02h) reset state = 0x00000000h p(31:0) function 31:27 reserved. set to 0. 26:16 slot 3 control (see table 24) 15:11 reserved. set to 0. 10:0 slot 2 control (see table 24) table 26. xout data versus time slot routing, time slots 4, 5 (iwa = 0*03h) reset state = 0x00000000h p(31:0) function 31:27 reserved. set to 0. 26:16 slot 5 control (see table 24) 15:11 reserved. set to 0. 10:0 slot 4 control (see table 24) isl5416
38 the isl5416 can be programmed to respond to the synchronization inputs, syncin1 and syncin2, in a number of ways. this register controls the range control response to the sync inputs. the response of the channels to the sync inputs is controlled by iwa = *000h. clko start up from syncin1 or syncin2 is controlled in iwa = 0000h. table 27. xout data versus time slot routing, time slots 6, 7 (iwa = 0*04h) reset state = 0x00000000h p(31:0) function 31:27 reserved. 26:16 slot 7 control (see table 24) 15:11 reserved. 10:0 slot 6 control (see table 24) table 28. input syncin1/syncin2 function select (iwa = 0*05h) reset state = 0x00000000h p(31:0) function 31 enable syncin2 to range control. 1 = range control block responds to synci n2 according to bits 19:16 settings. 0 = range control block does not respond to syncin2. 30:20 reserved. set to 0. 19 read update on syncin2. 1 = attenuator control accumulator contents are transferred to the read hol ding register on syncin2. 18 load accumulator on syncin2. 1 = attenuator control accumulator is loaded fr om the up load regist er (0*19) on a syncin2. 17 range control startup on syncin2. 1 = enable range control on syncin2 if 0*10, bit 1 is set. range cont rol is disabled on reset or by writing a zero to iwa = 0*1 0h, bit 0. if this bit is set, bit 0 of 0*10 is set on a syncin2. 16 timing reset on syncin2. 1 = reset range control timing on syncin2 if 0*10, bit 1 is set. on ce started, the timers will r un until disabled (bit 1 of 0*1 0h or hw reset). 15 enable syncin1 to range control. 1 = range control block responds to syn cin1 according to bits 3:0 settings. 0 = range control block does not respond to syncin1. 14:4 reserved. set to 0. 3 read update on syncin1. 1 = attenuator control accumulator contents are transferred to the read hol ding register on syncin1. 2 load accumulator on syncin1. 1 = attenuator control accumulator is loaded fr om the up load regist er (0*19) on a syncin1. 1 range control startup on syncin1. 1 = enable vga startup control on syncin2 if 0*10, bit 1 is set. ra nge control is disabled on reset or by writing a zero to 0*1 0, bit 0. if this bit is set, bit 0 of 0*10 is set on a syncin1. disable on reset. 0 timing reset on syncin1. 1 = reset range control timing on syncin1 if 0*10, bit 1 is set. on ce started, the timers will r un until disabled (bit 1 of 0*1 0h or hw reset). isl5416
39 serial output: when bit 31 of gwa = 0000h is set, the dout bus is used for serial outputs. four bits are allocated to each channel as follows: a common serial clock generator is used for all four outputs, so the four sclks are synchronous. four separate outputs are provided to simplify pwb routing. each sclk output can be separately enabled, so that unused clock outputs can be turned off. serial outputs are always msb first. addresses 0106h to 0108h control the serial output from channel 0. addresses 0206h to 0208h control the serial output from channel 1. addresses 0406h to 0408h control the serial output from channel 2. addresses 0806h to 0808h control the serial output from channel 3. table 29. high, low byte data type codes (after rounding in the channel) code channel 0, channel 1 muxes channel 2, channel 3 muxes 0000 ch 0 i(23:16) ch 2 i(23:16) 0001 ch 0 i(15:8) ch 2 i(15:8) 0010 ch 0 i(7:0) ch 2 i(7:0) 0011 ch 0 q(23:16) ch 2 q(23:16) 0100 ch 0 q(15:8) ch 2 q(15:8) 0101 ch 0 q(7:0) ch 2 q(7:0) 0110 ch 0 agc(15:8) ch 2 agc(15:8) 0111 ch 0 agc(7:0) ch 2 agc(7:0) 1000 ch 1 i(23:16) ch 3 i(23:16) 1001 ch 1 i(15:8) ch 3 i(15:8) 1010 ch 1 i(7:0) ch 3 i(7:0) 1011 ch 1 q(23:16) ch 3 q(23:16) 1100 ch 1 q(15:8) ch 3 q(15:8) 1101 ch 1 q(7:0) ch 3 q(7:0) 1110 ch 1 agc(15:8) ch 3 agc(15:8) 1111 ch 1 agc(7:0) ch 3 agc(7:0) table 30. serial output bits allocation channel 0 channel 1 channel 2 channel 3 sclkx dout0 dout4 dout8 dout12 ssyncx dout1 dout5 dout9 dout13 sd1x dout2 dout6 dout10 dout14 sd2x dout3 dout7 dout11 dout15 isl5416
40 table 31. serial output control (iwa = 0*06h) reset state = 0x00000000h p(31:0) function 31:30 reserved. set to 0. 29 sclk enable. 1 = enable the serial clock generator is shar ed by the four serial outputs. a serial clock pin is provided for each output. each pin can b e enabled or disabled independent of the other channels. 28 ssyncx polarity. 1 = active low. 0 = active high. the ssyncx signal is asserted for one serial clock period for each time slot where ssync is enabled. 27:26 reserved. set to 0. 25:24 ssyncx position. 00 = early ssync. ssyncx is asserted during the serial clock period prior to the firs t data bit for each slot where ssync is en abled. 01 = late ssync. ssyncx is asserted during the serial clock period following the last data bit for each slot where ssync is ena bled. 1x = coincident ssync. ssyncx is asserted during the serial cl ock period of first da ta bit for each slot where ssync is enabled . 23:20 reserved. set to 0. 19:16 sd1x routing mask. these bits gate the serial output of each ch annel to any or all 4 serial output pins . the gated serial outputs from all of the channels are ored together. this allows channels to be multiplexed together on a single serial output by offsetting the serial data stre ams from each other using the hold off delay below or by using empty time slots. note that the serial data from each channel is zeroed a fter all of the slots have been output, so it will not interfere with a delayed channel. the multiplexing of the ssyncx signals matches the data multiplexing. 19 - enable the serial output to the sd1d pin. 18 - enable the serial output to the sd1c pin. 17 - enable the serial output to the sd1b pin. 16 - enable the serial output to the sd1a pin. 15:12 sd2x routing mask. 15 - enable the serial output to the sd2d pin. 14 - enable the serial output to the sd2c pin. 13 - enable the serial output to the sd2b pin. 12 - enable the serial output to the sd2a pin. 11:0 output hold off delay. these bits control a programmable hold off del ay from the time a set of data samples is provided to the serial output section t o the time that the serial output begins. the delay is programmed in serial clocks. program with the desired number of serial clocks: 0 = no delay table 32. serial output sd1x slot control (iwa = 0*07h) reset state = 0x00000000h p(31:0) function 31 slot 4 sync enable. 1 = ssyncx active for this time slot 0 = no ssyncx for this time slot 30:27 slot 4 word width. 0000 = 0-bit* 0001 = 4-bit 0010 = 6-bit 0011 = 8-bit 0100 = 10-bit 0101 = 12-bit 0110 = 16-bit 0111 = 20-bit 1000 = 24-bit 1001 = 32-bit (8 lsbs zeroed) all other codes are invalid. note that if the channel output is rounded to fewer than 24 bits and fewer than 24 bits is selecte d for the slot width, the output will be doubly rounded. * if 0-bit is selected for slot 1, 2, or 3, one sclk period will actually be used, though no data will be output. isl5416
41 26:24 slot 4 data type. 000 = zeros. 001 = i 010 = q 011 = agc (real time). 23:16 slot 3. see bits 31:24. 15:8 slot 2. see bits 31:24. 7:0 slot 1. see bits 31:24. table 32. serial output sd1x slot control (iwa = 0*07h) reset state = 0x00000000h p(31:0) function table 33. serial output sd2x slot control (iwa = 0*08h) reset state = 0x00000000h p(31:0) function 31 reserved, set to 0. 30:27 slot 4 word width. 0000 = 0-bit* 0001 = 4-bit 0010 = 6-bit 0011 = 8-bit 0100 = 10-bit 0101 = 12-bit 0110 = 16-bit 0111 = 20-bit 1000 = 24-bit 1001 = 32-bit (8 lsbs zeroed) all other codes are invalid. note that if the channel output is rounded to fewer than 24 bits and fewer than 24 bits is selecte d for the slot width, the output will be doubly rounded. * if 0-bit is selected for slot 1, 2, or 3, one sclk period will actually be used, though no data will be output. 26:24 slot 4 data type. 000 = zeros. 001 = i 010 = q 011 = agc (sampled real time, not sampled by p write or real time updated every agc input). 23:16 slot 3. see bits 31:24. 15:8 slot 2. see bits 31:24. 7:0 slot 1. see bits 31:24. table 34. range control dc blocking filter control (iwa = 0*09h) reset state = 0x00000000h p(31:0) function 31:19 reserved. set to 0. 18 dc filter disable. 1 = clear filter based dc offset, shut off filter 17:16 dc filter gain. 00 = widest 01 = medium 10 = narrowest (roughly 120 hz hpf at 61.44 msps) 11 = use p programmed dc offset value 15:0 p loaded dc offset. twos complement. these bits are subtract ed from the 16 mantissa bits. bit 15 has the same weighting as xin16 (msb). isl5416
42 a/d range control registers the range control section monito rs the output of the a/d and adjusts the rf/if gain to ma intain a desired a/d output range. the gain adjustments are in 6 db steps. the levels, adjustment rates, and gain to bit mapping are programmable. a code is programmed for output on the eout bus for each of the eight states of the three msbs of the attenuator control register. these codes can be up to 8 bits, but if four gain control sections are used, only four bits are available for each gain control section. the ro uting of the gain control bits to eout bits is done in iwa = 0001h. the range control registers will be explained later in: iwa = 0*09h - 0*16h and iwa = 0*19h - 0*1dh direct addresses 4-7 have a sequenced read mode for quickly reading i/q output data, real time agc gain, sampled agc gain, or range control data. the read order is programmed in this register. the arrival of a new channel output sample resets the read pointer to the first data type. the rising edge of the rd signal (or dstrb with rd/wr high in up mode 1) will advance the pointer for that channel to the next data to get it set up at the interface for fast access. the clko2 signal can be programmed as an interrupt signal for this mode (see gwa = 0000h, bit 13) to tell the processor when there is a new data. alternatively, the frame strobe signals from the channels can be used. table 35. p sequenced read mode, read order select (iwa = 0*0ah) reset state = 0x00000000h p(31:0) function 31 1 = generate an interrupt with eac h new data output for this channel. 30:28 read eight. see bits 3:0. mask off the msb of the code. 27:24 read seven. see bits 3:0. 23:20 read six. see bits 3:0. 19:16 read five. see bits 3:0 15:12 read four. see bits 3:0 11:8 read three. see bits 3:0 7:4 read two. see bits 3:0 3:0 read one. (first read). 0000 = agc gain -- sampled by a write to iwa = *010h 0001 = agc gain -- real ti me (updated with every i/q) 0010 = q(7:0), 8 zeros 0011 = q(24:8) 0100 = i(7:0), 8 zeros 0101= i(24:8) 0110 = range control i&d (see iwa = 0*1ch) 0111 = range accumulator (see iwa = 0*1bh) note that the i/q data will be rounded to the number of bits pr ogrammed in iwa = *001h bits 5:3, and the real time agc data wil l have the format selected in iwa = *001h, bits 2:0. table 36. adc range control -- main (iwa = 0*10h) reset state = 0x00000000h p(31:0) function 31:24 immediate hold off. delay in clock cycles after the immediate threshold adjustm ent before another adjustment is allowed (00000000 = 1 clock delay). load with the desired value minus one. 23:16 update exponent delay. delay in clock cycles from a change in t he vga value before it is added to the input exponent value. this should be set to equa l the rf/adc plus the isl5416 pipel ine delays (6 clocks). 15 disable accumulator updates. 1 = disable range control accumulator updates. up can still load. isl5416
43 14 time using clocks/samples. count intervals and delays usin g clocks or input enables. 1 = clocks. 0 = input enables. 13:11 upper limit. upper attenuator control limit, 0 to 42 db in 6 db steps. 000 = 0 db. 111 = 42 db. 10:8 lower limit. lower attenuator control limit, 0 to 42 db in 6 db steps. 000 = 0 db. 111 = 42 db. 7:4 normalization shift/divide select. normalizing shifter control. divi des the integrated magnitude by 2 n prior to threshold comparison: accumulator input bits: 2 0 . . . 2 -15 accumulator output bits: 2 16 . . . 2 -15 0000 = select 2 1 to 2 -14 1111 = select 2 16 . . .to 2 1 3 enable input exponent bits. enable the exponent bits from the input to be added to t he attenuation control bits and routed to the channel(s). 2 enable range control exponent bits. enable the attenuator control register (accumulator msbs) bits to be added to the exponent bits from the input and routed to th e channel(s). 1 enable range control. 1 = enable the range control. 0 = disable the range cont rol (including timers). 0 enable range control bits enable changes in the attenuator control register (if this bit is cleared, the timers still run but changes to the register are inhibited. note: the range control can be enabled by writing to bit 0 or bit 0 can be set by syncinx to start updates. see iwa = 0*05h. tim ing reset by syncinx is enabled by bits 0 and 16 of iwa = 0*05h. table 36. adc range control -- main (iwa = 0*10h) reset state = 0x00000000h (continued) p(31:0) function table 37. time slot period, delay from syncinx to start of integration (iwa = 0*11h) reset state = 0x00000000h this control is shared for ain/bin and for cin/din p(31:0) function 31:16 sync delay. delay from sync (external or counter generated) to the start of integration. range of delay is 1 to 65536, load with the desire d value minus 1. delay in input samples or clocks as selected by bit 14, iwa = 0*10h. 15:0 slot period. time interval between counter generated syncs in samples or cloc ks as selected by bit 14, iwa = 0*10h. range for period is 2 to 65536, load with period minus 1. table 38. number of integrations per slot, integration time (iwa = 0*12h) reset state = 0x00000000h this control is shared for ain/bin and for cin/din p(31:0) function 31:16 number of integrations per slot. number of input magnitude integration per iods per slot. range is 1 to 32768, lo ad with number of integrations minus 1. 15:0 integration time. number of input samples to average before making an upper or lowe r threshold decision in samples or clocks as selected by bit 1 4, iwa = 0*10h. range is 2 to 65536, load with samples minus 1. isl5416
44 notes on thresholds: the input range is -32768 to 32767 (0x8000 to 0x7fff) the magnitude range is 0 to 32768 ( 0x0000 to 0x8000) the immediate threshold is 0 to 65536 (0x0000 to 0xffff) the magnitude must be greater than the immediate threshold to trigger. the integration period is 2 to 2 16 . the integrated magnitude range is 0x00000000 to 0x80000000 the normalization divider range is 2 1 to 2 16 , so with the maximum dividers: the averaged output is 0x0000 to 0x8000 the upper and lower thresholds are 0 to 65536 (0x0000 to 0xffff) to trigger the upper and lower comparisons: the magnitude must be greater than the upper threshold. the magnitude must be less than lower threshold. table 39. lower threshold, attenuator change (iwa = 0*13h) reset state = 0x00000000h this control is shared for ain/bin and for cin/din p(31:0) function 31:30 reserved. set to 0. 29:16 attenuation step 3. amount to decrease the attenuation control register if the aver age input magnitude is below the lower threshold. 14-bit value l oads the lsbs of a 17-bit accumulator. the 3 msbs of the 17-bit lo ad are hard-wired to ?111? , so the load value should be masked fr om a negative number. the msb of the 14-bit value is weighted at 6 db so the value at all zeros would be ~12 db (decrease in attenuation). 15:0 threshold 3. lower threshold value. if the accumulated and normalized input magni tude is below this value, the attenuation control accumulat or is adjusted by the amount programmed in bits 29:16. table 40. upper threshold, attenuator change (iwa = 0*14h) reset state = 0x00000000h this control is shared for ain/bin and for cin/din p(31:0) function 31:30 reserved. set to 0. 29:16 attenuation step 2. amount to increase the attenuation control register if the av erage input magnitude is above the upper threshold. 14-bit value l oads the lsbs of a 17-bit accumulator. the 3 msbs of the 17-bit load are hard-wired to ?000? . the ms b of the 14-bit value is weight ed at 6 db so the value at all ones would be ~12 db (increase in attenuation). 15:0 threshold 2. upper threshold value. if the accumulated and normalized input m agnitude is above this value, the attenuation control accumulat or is adjusted by the amount programmed in bits 29:16. table 41. immediate threshold, attenuator gain (iwa = 0*15h) reset state = 0x00000000h this control is shared for ain/bin and for cin/din p(31:0) function 31:30 reserved. set to 0. 29:16 attenuation step 1. amount to increase the attenuation control register if the samp le-by-sample input magnitude is above the immediate threshold. 1 4- bit value loads the lsbs of a 17-bit accumulator. the 3 msbs of the 17-bit load are hard-wired to ?000? . the msb of the 14-bit value is weighted at 6 db so the value at al l ones would be ~12 db (increase in attenuat ion). this change happens immediately. subseq uent changes due to crossing this threshol d are held off by the delay in location iwa = 0*10, bits 31:24. 15:0 threshold 1. immediate threshold value. if the input magnitude is above this va lue, the attenuation control accumulator is adjusted by the a mount programmed in bits 29:16. isl5416
45 notes on loading and reading the range control accumulator: master bus -> holding register -> accumulator -> holding register the accumulator is 17 bits. the lower 16 bits are loaded from th e micro-processor interface master register into a holding regi ster. the msb of the input holding register is always zero. the accumu lator is loaded from the holding register. when the accumulator is read, the most significant 16 accumulator bits are returned (the lsb of the accumulator is not read). range control bit weights are listed in table 85 in the back of the data sheet. table 42. accumulator 16:0 xxxxxxxxxxxxxxxxx master 15:0 xxxxxxxxxxxxxxxx iwa = 0*19h z, 15:0 0xxxxxxxxxxxxxxxx iwa = 0*1bh 16:1 xxxxxxxxxxxxxxxx table 43. leak factor (bias) (iwa = 0*16h) reset state = 0x00000000h this control is shared for ain/bin and for cin/din p(31:0) function 31:16 unused. 15:0 leak factor. this signed value is added to the attenuator control register if the average input m agnitude is between the upper and lower thr esholds when updates are enabled (at the end of integration) s [ s -1 . . . -15 ] into the 17-bit accumulator. msb of 16 programmable bits (s) has a weight of -24 db. table 44. gain mapping look-up table register 1 (iwa = 0*17h) reset state = 0x00000000h p(31:0) function 31:24 attenuation code 3 (18 db code). 8 bit output code when th e attenuation register msbs = 011b. see output e routing control for bit to pin mapping. 23:16 attenuation code 2 (12 db code). 8 bit output code when th e attenuation register msbs = 010b. see output e routing control for bit to pin mapping. 15:8 attenuation code 1 (6 db code). 8 bit output code when the attenuation register msbs = 001b. see output e routing control f or bit to pin mapping. 7:0 attenuation code 0 (0 db code). 8 bit output code when the attenuation register msbs = 000b. see output e routing control fo r bit to pin mapping. table 45. gain mapping look-up table register 2 (iwa = 0*18h) reset state = 0x00000000h p(31:0) function 31:24 attenuation code 7 (42 db code). 8 bit output code when th e attenuation register msbs = 111b. see output e routing control for bit to pin mapping. 23:16 attenuation code 6 (36 db code). 8 bit output code when th e attenuation register msbs = 110b. see output e routing control for bit to pin mapping. 15:8 attenuation code 5 (30 db c ode). 8 bit output code when the attenuation r egister msbs = 101b. see output e routing control for bit to pin mapping. 7:0 attenuation code 4 (24 db c ode). 8 bit output code when the attenuation r egister msbs = 100b. see output e routing control for bit to pin mapping. table 46. p attenuator control accumulator load (iwa = 0*19h) reset state = 0x00000000h p(31:0) function 31:16 reserved. set to 0. isl5416
46 15:0 attenuator load value. up loading of attenuator control. msb = 24 db, next-msb = 12 db, et c. only top three bits are us ed to address the look-up table . table 46. p attenuator control accumulator load (iwa = 0*19h) reset state = 0x00000000h p(31:0) function table 47. p attenuator control accumulator load strobe (iwa = 0*1ah) reset state = inactive p(31:0) function n/a a write to this location generates a one-clock-wide strobe that transfers the p attenuator control regist er preload from iwa = 0*19h to the accumulator. the transfer can also be caused by a syncinx signal (see iwa = 0*05h) table 48. p attenuator control accumulator read strobe (iwa = 0*1bh) reset state = inactive p(31:0) function n/a a write to this location generates a one- clock-wide strobe that transf ers the contents of the attenuator control accumulator to a holding register to stabilize it for up reading. the transfer can also be caused by a synci nx signal (see iwa = 0*05h). msb = 48 db (al ways zero). note: 0x8000h is written to iw a = 0*19h is read 0x4000 here. table 49. p range control status read strobe (iwa = 0*1ch) reset state = inactive p(31:0) function n/a a write to this location generates a one- clock-wide strobe that transfers the content s of the range control (averaging) accu mulator after normalization to a holding register to stabilize it for up reading. bit 16 - 1 = immediate thres hold crossed since last read. 15:0 - integrated magnitude after normalization (c omparison value for aver age magnitude thresholds). table 50. p range control dc offset read strobe (iwa = 0*1dh) reset state = inactive p(31:0) function n/a a write to this location generates a one- clock-wide strobe that transfers the contents of the range control dc offset filter accumulator to a holding register to stabilize it for up reading. 21:0 - signed dc offset value +/- 0.25 * full scale (input) range. isl5416
47 tables of channel indirect write address registers the response of the channels to the syncin1 and syncin2 inputs is controlled by iwa *000h. bits 31:16 control the response to syncin2 and bits 15:0 control the response to syncin1. most processing blocks can be individually reset by a sync. some control registers are double buffered with the up loading a master regist er and the data transfer to a slave/active register occurring on either a write to a strobe iwa location or on a syncinx signal. updating with a syncinx signal allows the f unctions in more than one isl5416 to be updated simultaneously. set bits to 1 to enable. . table 51. channel reset/syncin1, syncin2 control (iwa = *000h) reset state = 0x00000000h p(31:0) function 31 channel syncin2 enable. enables c hannel overall response to syncin2. 30 reserved. set to 0. 29 serial output. reset the serial output section on syncin2. this will reset the serial clock divider if gwa = 0000h, bit 29 is set. 28 fifo. reset the fifo at the ihbf/resampler input on syncin2. 27 leap counter. reset the ihbf/resampler leap counter on syncin2. 26 resampler ncos. reset back end - resampler ncos on syncin2. 25 fifo/interpolation half band filter/resampler. reset back en d (interpolation half band filter, resampler, decimation counter, and fifo) on syncin2. 24 agc reset. resets agc processing on syncin2 and sets the agc gain to 0 db (or to the lower limit if greater than 0 db). 23 agc timing. resets agc timing on syncin2. 22 fir2 reset. resets fir2 on syncin2. resets the datapath registers and decimation counter. 21 fir1 reset. resets fir1 on syncin2. resets the datapath regist ers and decimation counter. 20 cic reset. resets the cic on syncin2. 19 resampler update. update (start) the resampler ncos frequency from the master/holding register to the slave/active register on syncin2. 18 agc gain load. update/load agc gain from the master/ holding register to the slav e/active register on syncin2. 17 carrier center frequency update. updates carrier center fr equency from the master/holding r egister to the slave/active register on syncin2. 16 data path update. update channel pr ocessing control register (*001) from the master/holding regi ster to the slave/active register on syncin2. 15 channel syncin1 enable. enables c hannel overall response to syncin1. 14 reserved. set to 0. 13 serial output. reset the serial output section on syncin1. this will reset the serial clock divider if gwa = 0000h, bit 29 is set. 12 fifo. reset the fifo at the ihbf/resampler input on syncin1. 11 leap counter. reset the ihbf/resampler leap counter on syncin1. 10 resampler ncos. reset back end - resampler ncos on syncin1. 9 fifo/interpolation half band filter/resampler. reset back en d (interpolation half band filter, resampler, decimation counter, and fifo) on syncin1. 8 agc reset. resets agc processing on syncin1 and sets the agc gain to 0 db (or to the lower limit if greater team 0 db). 7 agc timing. resets agc timing on syncin1. 6 fir2 reset. resets fir2 on syncin1. resets the datapath registers and decimation counter. 5 fir1 reset. resets fir1 on syncin1. resets the datapath regist ers and decimation counter. 4 cic reset. resets the cic on syncinq1. 3 resampler update. update (start) resampler ncos frequency from the master/holding register to the slave/active register on syncin1. isl5416
48 the channel processing control register enables and disables the major processing blocks in the channel. this register is double buffered. on re set, the slave/active register is cleared, disabling the processing in the channel. the processing is enabled by updat ing the slave register with either a write to location iwa *019h or by a syncin signal, if enabled (see iwa *000h). 2 agc gain load. update/load agc gain from the master/holding register to the slave/active register on syncin1. 1 carrier center frequency update. updates carrier cent er frequency from the master/holding register to the slave/active register on syncin1. 0 data path update. update c hannel processing control regi ster (*001) on syncin1. table 51. channel reset/syncin1, syncin2 control (iwa = *000h) reset state = 0x00000000h (continued) p(31:0) function table 52. channel processing control (iwa = *001h) reset state = 0x00000000h p(31:0) function 31:29 cof enable. this is a serial carrie r offser frequency control loaded through the din bus at the clock rate. the serial bit s are shown in table 53. msb justified with unused lsbs zeroed. cof is sh ifted in one bit per clock, msb first. cofsync is one-clock-wide, active high pulse, asserted during the cloc k period before the first bit (msb). the o ffset frequency is added to the center fre quency loaded by the p. the offset word is a 32-bit twos complement value. 0xx = disabled. 100 = 8 bits offset. 101 = 16 bits offset. 110 = 24 bits offset. 111 = 32 bits offset. 28:23 reserved. set to 0. 22 hoif interpolate by 1. 1 = set this bit to enable hoif in the ihbf/resampler block. this is provided to use the hoif as a phase (time) shift with no r ate change. 21 agc bypass. 1 = bypass agc; 0 = agc enabled. 20 agc timer enable. 1 = timing counters in the agc are enabled. 0 = timing counters in the agc are disabled. note: counters must be enabled if one of the time d modes is selected in register iwa = *008h. 19:17 channel input select. 000 = disabled. 001 = up test register (iwa = 00 02h) used as an input, always enabled. 010 = up test register used as an input, enabled by a strobe at iwa = 0003h. 011 = reserved. 100 = ain. 101 = bin. 110 = cin. 111 = din. table 53. cof and cofsync to din bit mapping channel cof cofsync 0 din16 din15 1 din12 din11 2din8din7 3din4din3 isl5416
49 16 pn noise enable. pn noise is added to the output of the mixer at the level selected in location iwa = *006h. this bit enables the pn generator to the channel. when 0, iwa = *006h will be a dc offset. 1 = pn added to the mixer output. 0 = pn disabled. 15 cic filter enable. 1 = cic filter enabled (minimum decimation is 2). 0 = cic filter disabled (the cic shifter is used for floating point to fi xed point conversion). 14:13 cic to fir1 gain adjust. the 24-bit output from the nco/mi xer/cic block to fir1 is scaled and rounded to 20 bits. saturati on detection is included. this gain stage is always present in the data path. 00 = x1. 01 = x2. 10 = x4. 11 = x8. 12 fir1 enable. 1 = fir1 enabled. 0 = fir1 bypassed. 11:10 fir1 to fir2 gain adjust. the 24-bit output from the fi r1 block to fir2 is scaled and rounded to 20 bits. saturation detec tion is included. this gain stage is always present in the data path. 00 = x1. 01 = x2. 10 = x4. 11 = x8. 9 fir2 enable. 1 = fir2 enabled. 0 = fir2 bypassed. 8 interpolation half band filter enable. 1 = ihbf enabled. 0 = ihbf bypassed. 7 resampler enable. 1 = resampler enabled. 0 = resampler bypassed. 6 leap counter enable. 1 = resampler leap counter enabled. 0 = resampler leap counter disabled. 5:3 i/q output rounding. i and q data busses to the output sect ion are rounded to the selected number of msbs. bits below the programmed number of bits are zeroed. 000 = 24 bits, data on bits 23:0. 001 = 20 bits, data on bits 23:4, bits 3:0 zeroed. 010 = 16 bits, data on bits 23:8, bits 7:0 zeroed. 011 = 12 bits, data on bits 23:12, bits 11:0 zeroed. 100 = 8 bits, data on bits 23:16, bits 15:0 zeroed. 101 = 6 bits, data on bits 23:18, bits 17:0 zeroed. 110 = 4 bits, data on bits 23:20, bits 19:0 zeroed. 111 = reserved. table 52. channel processing control (iwa = *001h) reset state = 0x00000000h (continued) p(31:0) function isl5416
50 2:0 agc gain output rounding. the agc gain is rounded to the selected number of bits. the agc gain word into the rounder is: e e e e m m m m m m m m m m m m m m. the agc gain is 2^e * (1.0 + m / 16384) eeee = 0 - 15, mmm...m = 0 - 16383 0 1. m m m m m m m m m m m m m m * 2^(e e e e) z = zero all other codes are reserved. table 52. channel processing control (iwa = *001h) reset state = 0x00000000h (continued) p(31:0) function table 54. fedcba98 76543 210 resolution 000= e e e emmmmmmmmmmmm ~ 96/ 65536db 001= e e e emmmmmmmm z z z z ~ 96/ 4096db 010=eeeemmmmzzzzzzzz~ 96/ 256db table 55. nco center frequency (iwa = *002h) reset state = 0x00000000h p(31:0) function 31:0 carrier center frequency (ccf). the range is -fs/2 to +fs/2, where fs is the input sample rate to the part. the carrier frequency is: fc = fs * ccf / (2^32). the ccf control is double buffered. the active carrie r center frequency register is read at this address. table 56. nco phase offset (iwa = *003h) reset state = 0x00000000h p(31:0) function 31:16 unused. 15:0 phase offset. this is the value of the phase offset added to the carrier nco phase. this can be interpreted as either an un signed number with a range from 0 to 2 ? ( / 32768) in 65536 steps, or as a signed number with a range of - to ? ( / 32768) in 65536 steps. table 57. nco center frequency update (iwa = *004h) reset state = inactive n/a function n/a strobe. a write to this location generat es a one-clock-wide strobe that transfers the carrier center frequency written to th e holding register (iwa = *002h) to the active r egister. the transfer can also be caused by a synci signal (see iwa = *000h). the content s of the active register are read at location iwa = *002h. table 58. nco/mixer/cic control (iwa = *005h) reset state = 0x00000000h p(31:0) function 31:29 maximum exponent shift saturation level. the maximum allo wed shift factor from the input a/d range control block and the exponent from the input bus. if the sum exc eeds this value, it will be saturated at this value. this is forced to 111 if th e up test register is selected as the input source. 28 force zero feedback. set to 1 to force nco phase accumulator feedback to 0. (provided for test) 27 clear feedback on update. 1 = clear nco phase accumulator on frequency load/update (f or synchronizing the nco phas e of multiple channels). 0 = phase continuous frequency updates. 26 reserved. set to 0. 25:20 cic base shift factor. the shift factor (exponents) from the input and the range control block are added to this value. 19 reserved. set to 0. isl5416
51 18:16 cic stages. number of cic integrat or/comb pairs (1-5). stages are enabled starting with the largest accumulator. 000 = 1 stage; 001 = 2 stages; 010 = 3 stages; 011 = 4 stages; 100 = 5 stages; load with the number of stages minus 1. 15:0 cic decimation factor. load with (decimation - 1). table 58. nco/mixer/cic control (iwa = *005h) reset state = 0x00000000h p(31:0) function table 59. pn gain (iwa = *006h) reset state = 0x00000000h p(31:0) function 31:16 unused. 15:0 pn level. level of the pn generator data that is added to the mixer output when the pn generator is enabled in iwa = *001h, bit 16. table 60. fir1, fir2, post-resampler control (iwa = *007h) reset state = 0x00000000h p(31:0) function 31:19 unused. 18:16 decimation of the resampler decimator block. load with the desired decimation factor minus one. 000 = decimation of 1 111 = decimation of 8 15 reserved. set to 0. 14:12 fir2 required clocks. this value should be set to: ceil (#taps/8) - 1. 11 reserved. set to 0. 10:8 fir2 decimation. available decimation through fir2 bl ock is 1-8. set the value to: decimation factor - 1. 7 fir1 enable hbf mode. 1 = half band filter mode. 0 = normal mode. 6:4 fir1 required clocks. set to: ceil (#taps/4) - 1. 3 reserved. set to 0. 2:0 fir1 decimation. available decimation through fir1 bl ock is 1-8. set the value to: decimation factor - 1. table 61. agc control, mean/median, delay, mode (iwa = *008h) reset state = 0x00000000h p(31:0) function 31:16 unused. 15 reserved. set to 0. 14 real time agc gain output source. 0 = the continuously updated loop filter output is provi ded to the output section as the ?real time? agc gain. 1 = the sampled loop filter output (sampled when the interval c ount is 0) is provided to the output section as the ?real time? agc gain. 13:8 data delay. this value is the number of samples that data is delayed to allow for agc settling. 0x00 = no delay. 0x3f = 1 sample . . . 0x01 = 63 samples. 7 loop gain 1 mean/median. 1 = agc settles to mean of the signal if lg1 is selected. 0 = agc settles to median of the signal. isl5416
52 d 6 loop gain 2 mean/median. 1 = agc settles to mean of the signal if lg2 is selected. 0 = agc settles to median of the signal. 5 up loop gain select. set to: 1 = loop gain 2. 0 = loop gain 1. 4 up loop gain control. set to: 1 = up loop gain select. 0 = counter controlled loop gain select. 3 disable loop filter update. set to: 1 = disable loop filter (holds the last value, the up can load). 0 = normal agc operation. 2 sampled mode. 1 = continuous forward gain updates. 0 = the forward gain held during the interval count (when loop gain 1 is used). the forward gain is continuously updated after the interval count (when loop gain 2 is used). 1 enable delay. set to: 1 = enable delay 0 = no delay note: delay must be enabled for sampled mode. 0 reserved. set to 0. table 63. threshold (set point) (iwa = *009h) reset state = 0x00000000h p(31:0) function 31:16 unused. 15:0 set point. bit weights: 2^ 2 1 0. 1 2 3 4 5 6 7 8 9 0 1 2 3, 2 0 is the full scale vector magnitude on one axis before the magnitude computation gain is applied. note: there is a gain in the magnitude computation prior to t he threshold compare. the gain versus number of clocks is: clocks gain 16 1.64676 8 1.647 4 1.642 3 1.630 2 1.581 the set point should be programmed to t he desired level times the magnitude gain. table 61. agc control, mean/median, delay, mode (iwa = *008h) reset state = 0x00000000h (continued) p(31:0) function table 62. control bit settings for agc modes agc mode iwa=*001h iwa = *008h bit 20 bit 4 bit 2 bit 1 continuous update p controls loop gain 0 110 contiuous with delay 0 1 1 1 timed loop gain selection 1 0 1 0 timed with delay 1 011 sampled gain updates (always with delay) 1 001 isl5416
53 table 64. upper/lower gain limit (iwa = *00ah) reset state = 0x00000000h p(31:0) function 31:16 upper limit. e e e e m m m m m m m m m m m m 01.mmmmmmmmmmmm * 2^eeee example: a gain of 48 (33.6 db) would be: 1.5 * 2 5 or 0101100000000000 15:0 lower limit. same weighting as upper limit. table 65. loop gains (1 & 2) (iwa = *00bh) reset state = 0x00000000h p(31:0) function 31:24 attack 1. loop gain value used for increasing signal levels (decreasing gain). bits 31:28 are the exponent, bits 27:24 are the mantissa. loop gain is 0.mmmm * 2 eeee . a loop gain of 00000000 weights the error by 2 -19 to the input of the accumulator. 23:16 decay 1. loop gain value used for decr easing signal levels (increasing gain). 15:8 attack 2. loop gain value used for increasing signal levels ( decreasing gain). bits 15:12 ar e the exponent, bits 11:8 are t he mantissa. loop gain is 0.mmmm * 2 eeee . a loop gain of 00000000 weights the error by 2 -19 to the input of the accumulator. 7:0 decay 2. loop gain value used for decr easing signal levels (increasing gain). table 66. agc counter preloads 1 (iwa = *00ch) reset state = 0x00000000h p(31:0) function 31:16 slot period (in clocks). sets the timing interval (slot peri od) counter period. the interval counter can be reset by synci nx (see iwa = *000h) to align it to the system timing. load with number of clocks minus 1. 15:0 delay (in clocks). sets the delay from the start of the timing interval (bits (31: 16 above) to the start of the interval se t in iwa = *00dh. load with number of clocks minus 1. table 67. agc counter preloads 2 (iwa = *00dh) reset state = 0x00000000h p(31:0) function 31:16 unused. 15:0 adjustment interval (in clocks). sets the length of the adjustment period for timed agc changes. (when loop gain 2 is used in timed and sampled modes). load with number of clocks minus 1. table 68. agc up gain load value (iwa = *00eh) reset state = 0x00000000h p(31:0) function 31:16 unused. 15:0 gain. this location allows the up to set the agc gain dire ctly. if the gain loaded by the up is outside the limits set in i wa = *00ah, it will be set to the limit value. to set a fixed gain, set the loop gains to zero and the limits to full scale (or set both li mits to the desired gain). bit weightings are the same as for the limits: eeee mmmm mmmm mmmm. table 69. agc up gain load strobe (iwa = *00fh) reset state = inactive n/a function n/a strobe. writing to this location generates a strobe, synchronized to the clock, that updates the agc with the gain value in register iwa = *00eh. the transfer can also be c aused by a syncinx signal (see iwa = *000h). table 70. readback sample agc gain and magnitude (iwa = *010h) reset state = inactive n/a function n/a strobe. writing to this location copies the agc loop filter accumulator contents to a holding register to stabilize it for r eading by the up at direct addresses 4 - 7. (the magnitude is read at this location after the strobe. magnitude bit weightings are: 2 2 . . . 2 -13 ) isl5416
54 table 71. nco 1 output rate -- top 32 bits (iwa = *011h) reset state = 0x00000000h p(31:0) function 31:0 rsout(48:16). this nco sets the re-sampl er output sample rate when the ihbf/resam pler block is enabled. if the hoif is bypa ssed, this is 2x the agc (and fir2) output rate. note that the re-sam pler only interpolates. the re-s ampler output can be decimated u sing a counter. the decimation is progra mmed in iwa = *007h, bits 18:16. fout = fs * rsout / (2^48), where rsout = 0 to 2^48 -1, fout = 0 to ~fs. bits 48:16 of the active register are read back at this location. if hoif is enabled, then f out must be f clk /2 if ihbf is enabled, hoif disabled, then f out must be f clk table 72. nco 1 output rate -- lower 16 bits (iwa = *012h) reset state = 0x00000000h p(31:0) function 31:16 rsout(15:0). see iwa = *011h. bits 15:0 of the ac tive register are read back at this location. 15:0 unused. table 73. nco 2 input rate -- top 32 bits (iwa = *013h) reset state = 0x00000000h p(31:0) function 31:0 rsin(48:16). this nco sets the output sample rate for the fi fo (input sample rate to the ihbf/resampler block) when the ihbf/resampler block is enabled. if the ho if is bypassed, this has no effect. fin = fout * rsin / (2^48), where rsin = 0 to 2^48 -1, fin = 0 to ~fout. fin should be set equal to the output sample rate of fir2 (fs / (cicdeci * fir1deci * fir2deci)). bits 48:16 of the active register are read back at this location. table 74. nco 2 input rate -- lower 16 bits (iwa = *014h) reset state = 0x00000000h p(31:0) function 31:16 rsin(15:0). see iwa = *013. bits 15:0 of the active register are read back at this location. 15:0 unused. register iwa = *015h is used to add delay to a channel. the output timing is unaffected (except at startup from reset). but the group delay of the channel is adjusted. if the ihbf and hoif are both enabled, the total delay range is 0 to 1023/256 fir2 outp ut sample periods. note: the hoif can be enabled with no rate c hange by setting the nco2 register (iwa = *013h and *014h) to zero and setting the bit 22 of iwa = *001h. table 75. phase offset (iwa = *015h) reset state = 0x00000000h p(31:0) function 31:11 reserved. set to 0. 10:9 fifo depth. after reset, the fifo waits until it reaches a depth of 2 + this value (total of 2, 3, 4, or 5) before allowing fifo reads. these bits can be used to add delay in in crements of the fir2 output sample rate. hoif or ihbf must be enabled. 8 hoif input delay. 1 = one hoif input sample delay is added. hoif must be enabled. 7:0 resampler phase offset. allows the output sample timing to be shifted in increments of 1/256 the hoif input period. 00 = least delay. ff = most delay. hoif must be enabled. isl5416
55 locations iwa = *01a and *01b delete or repeat a sample at the input to fir1 to shift the system timing. these locations should not be used if the ihbf or re-sampler is enabled. table 76. leap counter (iwa = *016h) reset state = 0x00000000h p(31:0) function 31:0 leap counter preload. when the leap counter reaches zero, it rese ts the phase accumulators in nco1 and nco2 to zero (if enabled). because some frequencies cannot be represented exactly by t he nco control word, this function is prov ided to zero out any error that builds up due to the limited nco resolution. load with th e desired number of clocks between resets mi nus 1. note that the ceiling function sh ould be used when computing the nco frequency control words so that the ncos build up a positive phase error and phase accumulator carries are not suppressed. table 77. resampler frequency/phase update strobe (iwa = *017h) reset state = inactive n/a function n/a strobe. a write to this location transfers the contents of iwa = *011/*012h and iwa = *013/*014h holding registers to the ac tive resampler nco frequency registers. the transfer can al so be caused by a syncinx signal (see iwa = *000h). table 78. channel soft reset (iwa = *018h) reset state = inactive n/a function n/a soft reset. a write to this location generates a soft reset of the channel. this clears the slave registers of master/slave pairs and stops processing in the channel. table 79. update data path control register (iwa = *019h) reset state = inactive n/a function n/a strobe. a write to this location updates the active data path control register from its holdi ng register. the transfer can a lso be caused by a syncinx signal (see iwa = *000h) . the active register is cleared on reset, disabling channel proc essing, so this up date is used to synchronously star t the channel processing. table 80. advance (iwa = *01ah) reset state = inactive n/a function n/a strobe. a write to this location repeats one sample at the out put of the cic. should only be used if there is sufficient ove rhead for processing an extra sample. table 81. retard (iwa = *01bh) reset state = inactive n/a function n/a strobe. a write to this location deletes one sample at the output of the cic. table 82. fir 1 coefficients (32 x 20) (iwa = *100h thru *11fh) reset state = inactive p(31:0) function 31:20 unused. 19:0 coefficients. twos complement -524287 to + 524287 (-524288 not allowed). +524288 would be unity gain. table 83. fir 2 coefficients (64 x 20) (iwa = *200h thru *23fh) reset state = undefined p(31:0) function 31:20 unused. 19:0 coefficients. twos complement -524287 to + 524287 (-524288 not allowed). +524288 would be unity gain. isl5416
56 table 83. fir 2 coefficients (64 x 20) (iwa = *200h thru *23fh) reset state = undefined p(31:0) function each multiplier has up to 8 coefficients. if the six bit coefficient addresses are labeled m2 m1 m0 c2 c1 c0, where the m-bit s select the multiplier (000 is the first mult iplier encountered by the data samples), and the c-bits select the coefficient (000 is th e first coefficient encounter ed by the data samples). m2 is not used for fir1. for example, if an eight tap filter is used for fir1, the coefficients would be loaded: (earliest) c0 --- 000000 (*100h) c1 --- 000001 c2 --- 001000 c3 --- 001001 c4 --- 010000 c5 --- 010001 c6 --- 011000 c7 --- 011010 (*11ah) when the half band filter mode is enabled for fir1, the zero valued zero crossing coefficients are not loaded, but the coefficients must be centered ar ound the hard wired center tap. this means that zeros valued coefficients must be added if the number of coefficients per multiplier is not uniform. fo r example, with an 11-tap half band filter with coefficients: c0, 0, c1, 0, c2, 0.5, c3, 0, c4, 0, c5, the coefficient loading would be: 0 ----- 000000 (*100h) c0 --- 000001 c1 --- 001000 c2 --- 001001 c3 --- 010000 c4 --- 010001 c5 --- 011000 0 ----- 011001 (*119h) the filter computation requires two clock cycl es for the computation. du e to the added zero -valued coefficient s, the group dela y has increased from 6 to 8 sample periods. isl5416
57 absolute maximum ratings thermal information supply voltage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.5 (core) 4.6 (i/o?s) input, output or i/o voltage . . . . . . . . . . . . . . . . . gnd -0.5v to 5.5v esd classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . class i operating conditions voltage range i/o, v ccio . . . . . . . . . . . . . . . . . +3.135v to +3.465v voltage range core, v ccc . . . . . . . . . . . . . . . . . . +1.71v to +1.89v temperature range industrial. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . -40 o c to 85 o c input low voltage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0v to +0.8v input high voltage . . . . . . . . . . . . . . . . . . . . . . . . . . 2v to i/o v ccio thermal resistance (typical) ja ( o c/w) 256 lead bga package. . . . . . . . . . . . . . . . . . . . . . 31 w/200 lfm air flow . . . . . . . . . . . . . . . . . . . . . . . . . 28 w/400 lfm air flow . . . . . . . . . . . . . . . . . . . . . . . . . 26 maximum junction temperature . . . . . . . . . . . . . . . . . . . . . . .150 o c maximum storage temperature range . . . . . . . . . . -65 o c to 150 o c for recommended soldering conditions see tech brief 334. caution: stresses above those listed in ?a bsolute maximum ratings? may cause permanent damage to the device. this is a stress o nly rating and operation of the device at these or any other conditions above those indicated in the operational sections of this specification is not implied. note: note : ja is measured with the component mounted on a high effective therma l conductivity test board with ?direct attach? features in fr ee air or with the airflow. see tech brief tb379 for details electrical specifications v ccc = core supply: 1.8v 0.09v, t a = -40 o c to 85 o c, industrial parameter symbol test conditions min typical max units logical one input voltage v ih v ccio = 3.45v 2.0 v logical zero input voltage v il v ccio = 3.15v 0.8 v output high voltage v oh i oh = -2ma, v ccc = 3.15v 2.6 3 v output low voltage v ol i ol = 2ma, v ccc = 3.15v 0.1 0.4 v input leakage current i i v in = v ccio or gnd, v ccio = 3.465v -10 < 1 (note 3) 10 a output leakage current i o v in = v ccio or gnd, v ccio = 3.465v -10 10 a standby power supply current (core) i ccsb-cr v ccc = 1.89v, see ac test load circuit, no clk 12 (note 6) ma standby power supply current (io?s) i ccsb-io v ccc = 1.89v, see ac test load circuit, no clk 0.75 ma operating power supply current (core) i ccop-cr f = 80mhz, v in = v ccc or gnd, v ccc = 1.89v, see ac test load circuit, 660 ma operating power supply current (io?s) i ccop-io f = 80mhz, v in = v ccc or gnd, v ccc = 1.89v, see ac test load circuit, 150 ma operating power supply core current (typical) i ccop-typ f = 80mhz, v in = v ccc or gnd, v ccc = 1.89v, see ac test load circuit, 500 (note 1, 4) ma input capacitance c in freq = 1mhz, v ccc open, all measurements are referenced to device ground 5pf (note 2) output capacitance c out 5pf (note 2) notes: 1. power supply current is proportional to frequency of operation and programmed configurat ion of the part. typical rating for i ccop is 6.5 ma/mhz. 2. capacitance: t a = 25 o c, controlled via design or pr ocess parameters and not directly tested. c haracterized upon initial design and at major process or design changes. 3. typical pull-up/down resistance values are 60 k ? . 4. typical measured power on the evaluation board: . umts: clk = 76.8 mhz, total power = 1.1 w (core current = 500 ma; i/o current = 60 ma) . cdma2000: clk = 80 mhz, total power = 750 mw (core current = 360 ma; i/o current = 30 ma). isl5416
58 electrical specifications v ccc = core supply: 1.8v 0.09v, v ccio = io?s supply: 3.3v 0.165v, t a = -40 o c to 85 o c industrial parameter symbol min max units input and control timing, clock skew (figure 11 and 12) clkc frequency f clk -95mhz clkc high t ch 3-ns clkc low t cl 4-ns setup time - data inputs, input enables to clkx high t ds 5.5 - ns hold time - data inputs, input enables to clkx high t dh 0-ns setup time - syncinx to clkc high t syncs 3 hold time - syncinx to clkc high t synch 0 clkc to output valid - synco t pdc -5ns reset pulse width low t rw 5-ns reset setup time to clkc high (note 5) t rs 5-ns clkx to clkc skew 1 (figure 12) t cks1 2.5 clkx to clkc skew 2 (figure 12) t cks2 2.5 microprocessor write timing ( p mode = 0, figure 13) p(15:0) setup time to rising edge of wr t psw 7.5 - ns p(15:0) hold time from rising edge of wr t phw -1 - ns a(2:0) setup time to rising edge of wr t asw 8.5 - ns a(2:0) hold time from rising edge of wr t ahw -2 - ns ce setup time to rising edge of wr t csw 7-ns ce hold time from rising edge of wr t chw -2 - ns wr low time t wpwl 5-ns microprocessor read timing ( p mode = 0, figure 14) a(2:0) hold time from rising edge of rd (note 7) t ahr -4 - ns a(2:0) to p(15:0) data valid time t dv -19ns rd enable time t re -8.5ns rd disable time (note 6) t rd -6ns ce setup time to falling edge of rd t csf 2.5 - ns ce hold time from rising edge of rd (note 6) t chr 0-ns read cycle time (note 7) t rcy 25 - ns microprocessor write timing ( p mode = 1, figure 15) p(15:0) setup time to rising edge of dstrb t psr 8-ns p(15:0) hold time from rising edge of dstrb t phr -0.5 - ns a(2:0) setup time to rising edge of dstrb t asr 7.5 - ns a(2:0) hold time from rising edge of dstrb t ahr -1.5 - ns ce setup time to rising edge of dstrb t csr 8-ns ce hold time from rising edge of dstrb t chr -2 - ns r/w setup time to falling edge of dstrb t r/w sf 0-ns r/w hold time from rising edge of dstrb t r/w hr -1 - ns dstrb low time t dw 6-ns microprocessor read timing ( p mode = 1, figure 16) a(2:0) hold time from rising edge of dstrb (note 7) t ahr -1 - ns a(2:0) to p(15:0) data valid time t dv -16ns isl5416
59 ac test load circuit dstrb enable time t re -14ns dstrb disable time (note 6) t rd -6.5ns ce setup time to falling edge of dstrb t csf 7.5 - ns ce hold time from rising edge of dstrb (note 7) t chr -2 - ns dstrb low time t dw 5-ns read cycle time (note 8) t rcy 25 ns jtag timing (figure 17) tdi, tms set up ts tt 4.5 ns tdi, tms hold th tt 1.5 ns tclk to tdo valid tov tdo 7ns tclk to tdo disabled tod tdo 7ns tclk to tdo enabled toe tdo 7ns capture input setup time t istp 5ns capture input hold time t ihld 1.5 ns tclk to output valid (note 6) t dvld 26ns clock output timing and output enables (figure 18) clkc to parallel data, fsyncx and clko1 (divide-by 2 thru 16 modes) t pd 2 (note 6) 6.5 ns clkc low to clko1 low (divide-by 1 mode) t pdl 2 (note 6) 6.5 ns clkc high to clko1 high (divide-by 1 mode) t pdh 2 (note 6) 6.5 ns time skew between clko1 and parallel data or fsyncx (divide-by 2 thru 16 modes) t skew3 -1.5 1.5 ns time skew between clko1 and parallel data or fsyncx (divide-by 1 mode) t skew4 -2.5 1.5 ns output enable time t oen 6ns output disable time t odis 4ns notes: 5. the isl5416 goes into reset immediately on reset going low and comes out of reset on the 4th rising edge of clk after reset goes high. 6. controlled via design or process parameter s and not directly tested. characterized upon initial design and at major process o r design changes. 7. t ahr and t chr apply only to direct reads of addresses 4 - 7. 8. reading from direct addresses 4 - 7 (sequential read mode). electrical specifications v ccc = core supply: 1.8v 0.09v, v ccio = io?s supply: 3.3v 0.165v, t a = -40 o c to 85 o c industrial parameter symbol min max units equivalent circuit c l (note) i oh 1.5v i ol dut switch s1 open for i ccsb and i ccop s 1 note - test head capacitance, 50pf (typ) isl5416
60 waveforms figure 11. control timing figure 12. clock skews clkc xout, synco reset 1/f clkc t ch t cl t ds t dh t pdc t rs t rw x = a, b, c or d xin, enix low to high edge for active high clock high to low edge for active low clock t syncs t synch syncinx clkx clkc x = a, b, c or d t cks2 t cks1 isl5416
61 figure 13. microprocessor write timing ( p mode = 0) figure 14. microprocessor read timing ( p mode = 0) waveforms (continued) rd wr a<2:0> p<15:0> cs valid t csw t chw t asw t ahw valid t phw t psw t wpwl rd ce wr add(2:0) p(15:0) t re t dv t rd t ahr t csf t chr t rcy isl5416
62 figure 15. microprocessor write timing ( p mode = 1)) figure 16. microprocessor read timing ( p mode = 1)) waveforms (continued) ce rd/wr dstrb add(2:0) p(15:0) t psr t r/w sf t phr t dw t ahr t chr t csr t asr t r/w hr ce rd/wr dstrb add(2:0) p(15:0) t csf t ahr t dw t dv t re t rd t chr t rw sf t rcy isl5416
63 figure 17. jtag timing waveforms (continued) tclk tdi, tms tdo ts tt th tt tod tdo toe tdo tov tdo inputs (capture) outputs t istp t ihld t dvld isl5416
64 figure 18. output timing waveforms (continued) clkc clko1 (/2 thru /16) fsyncx aout, bout, cout, dout, clko1 (divide by 1) t pdh t skew4 t pd t pdl eout t skew3 oex t odis t oen xout isl5416
65 table 84. cic passband and alias levels frequency 5th order 4th order 3 rd order 2nd order 1st order f s / r passband alias passband alias passband alias passband alias passband alias 0 0<-2000<-2000<-2000<-2000<-200 0.01 -0.007 -199.564 -0.006 -159.651 -0.004 -119.738 -0.003 -79.825 -0.001 -39.913 0.02 -0.029 -169.041 -0.023 -135.233 -0.017 -101.425 -0.011 -67.617 -0.006 -33.808 0.03 -0.064 -151.023 -0.051 -120.818 -0.039 -90.614 -0.026 -60.409 -0.013 -30.205 0.04 -0.114 -138.129 -0.091 -110.503 -0.069 -82.877 -0.046 -55.252 -0.023 -27.626 0.05 -0.179 -128.048 -0.143 -102.438 -0.107 -76.829 -0.071 -51.219 -0.036 -25.610 0.06 -0.257 -119.749 -0.206 -95.799 -0.154 -71.849 -0.103 -47.900 -0.051 -23.950 0.07 -0.351 -112.683 -0.280 -90.146 -0.210 -67.610 -0.140 -45.073 -0.070 -22.537 0.08 -0.458 -106.522 -0.367 -85.218 -0.275 -63.913 -0.183 -42.609 -0.092 -21.304 0.09 -0.580 -101.054 -0.464 -80.843 -0.348 -60.633 -0.232 -40.422 -0.116 -20.211 0.10 -0.717 -96.135 -0.573 -76.908 -0.430 -57.681 -0.287 -38.454 -0.143 -19.227 0.11 -0.868 -91.662 -0.694 -73.330 -0.521 -54.997 -0.347 -36.665 -0.174 -18.332 0.12 -1.034 -87.558 -0.827 -70.047 -0.620 -52.535 -0.413 -35.023 -0.207 -17.512 0.13 -1.214 -83.766 -0.971 -67.013 -0.728 -50.260 -0.486 -33.507 -0.243 -16.753 0.14 -1.409 -80.241 -1.127 -64.193 -0.846 -48.145 -0.564 -32.096 -0.282 -16.048 0.15 -1.619 -76.947 -1.295 -61.558 -0.972 -46.168 -0.648 -30.779 -0.324 -15.389 0.16 -1.844 -73.855 -1.475 -59.084 -1.107 -44.313 -0.738 -29.542 -0.369 -14.771 0.17 -2.084 -70.943 -1.667 -56.754 -1.251 -42.566 -0.834 -28.377 -0.417 -14.189 0.18 -2.340 -68.189 -1.872 -54.551 -1.404 -40.913 -0.936 -27.276 -0.468 -13.638 0.19 -2.610 -65.579 -2.088 -52.463 -1.566 -39.347 -1.044 -26.231 -0.522 -13.116 0.20 -2.896 -63.098 -2.317 -50.478 -1.737 -37.859 -1.158 -25.239 -0.579 -12.620 0.21 -3.197 -60.734 -2.558 -48.587 -1.918 -36.440 -1.279 -24.294 -0.639 -12.147 0.22 -3.514 -58.477 -2.811 -46.782 -2.108 -35.086 -1.406 -23.391 -0.703 -11.695 0.23 -3.847 -56.319 -3.077 -45.055 -2.308 -33.792 -1.539 -22.528 -0.769 -11.264 0.24 -4.195 -54.252 -3.356 -43.402 -2.517 -32.551 -1.678 -21.701 -0.839 -10.850 0.25 -4.560 -52.269 -3.648 -41.815 -2.736 -31.361 -1.824 -20.907 -0.912 -10.454 0.26 -4.941 -50.363 -3.953 -40.291 -2.965 -30.218 -1.976 -20.145 -0.988 -10.073 0.27 -5.338 -48.531 -4.271 -38.825 -3.203 -29.119 -2.135 -19.412 -1.068 -9.706 0.28 -5.752 -46.767 -4.602 -37.413 -3.451 -28.060 -2.301 -18.707 -1.150 -9.353 0.29 -6.183 -45.066 -4.946 -36.053 -3.710 -27.040 -2.473 -18.026 -1.237 -9.013 0.30 -6.631 -43.426 -5.305 -34.740 -3.978 -26.055 -2.652 -17.370 -1.326 -8.685 0.31 -7.096 -41.842 -5.677 -33.473 -4.257 -25.105 -2.838 -16.737 -1.419 -8.368 0.32 -7.578 -40.311 -6.063 -32.249 -4.547 -24.187 -3.031 -16.125 -1.516 -8.062 0.33 -8.078 -38.832 -6.463 -31.066 -4.847 -23.299 -3.231 -15.533 -1.616 -7.766 0.34 -8.596 -37.401 -6.877 -29.921 -5.158 -22.440 -3.439 -14.960 -1.719 -7.480 0.35 -9.133 -36.015 -7.306 -28.812 -5.480 -21.609 -3.653 -14.406 -1.827 -7.203 0.36 -9.688 -34.674 -7.750 -27.739 -5.813 -20.804 -3.875 -13.869 -1.938 -6.935 0.37 -10.262 -33.374 -8.209 -26.699 -6.157 -20.024 -4.105 -13.349 -2.052 -6.675 0.38 -10.854 -32.114 -8.684 -25.691 -6.513 -19.268 -4.342 -12.845 -2.171 -6.423 0.39 -11.467 -30.892 -9.174 -24.713 -6.880 -18.535 -4.587 -12.357 -2.293 -6.178 0.40 -12.099 -29.707 -9.679 -23.766 -7.260 -17.824 -4.840 -11.883 -2.420 -5.941 0.41 -12.752 -28.557 -10.201 -22.846 -7.651 -17.134 -5.101 -11.423 -2.550 -5.711 0.42 -13.425 -27.442 -10.740 -21.953 -8.055 -16.465 -5.370 -10.977 -2.685 -5.488 0.43 -14.119 -26.359 -11.295 -21.087 -8.472 -15.815 -5.648 -10.544 -2.824 -5.272 0.44 -14.835 -25.308 -11.868 -20.246 -8.901 -15.185 -5.934 -10.123 -2.967 -5.062 0.45 -15.573 -24.287 -12.458 -19.430 -9.344 -14.572 -6.229 -9.715 -3.115 -4.857 0.46 -16.333 -23.296 -13.066 -18.637 -9.800 -13.978 -6.533 -9.318 -3.267 -4.659 0.47 -17.116 -22.334 -13.693 -17.867 -10.270 -13.400 -6.847 -8.933 -3.423 -4.467 0.48 -17.923 -21.399 -14.339 -17.119 -10.754 -12.840 -7.169 -8.560 -3.585 -4.280 0.49 -18.754 -20.492 -15.003 -16.393 -11.253 -12.295 -7.502 -8.197 -3.751 -4.098 0.50 -19.610 -19.610 -15.688 -15.688 -11.766 -11.766 -7.844 -7.844 -3.922 -3.922 isl5416
66 table 85. isl5416 range control bit weighting gfedcba9876543210.123456789abcdef input sxxxxxxxxxxxxxxx dc offset sxxxxxxxxxxxxxxx magnitude 0000000000000000 xxxxxxxxxxxxxxxx immediate threshold xxxxxxxxxxxxxxxx accumulator 0 x x x xxxxxxxxxxxxxxxxxxxxxxxxxxxx shift: 0000 xxxxxxxxxxxxxxxx 0001 xxxxxxxxxxxxxxxx 0010 x xxxxxxxxxxxxxxx 0011 x x xxxxxxxxxxxxxx 0100 x x x xxxxxxxxxxxxx 0101 xxxxxxxxxxxxxxxx 0110 xxxxxxxxxxxxxxxx 0111 x xxxxxxxxxxxxxxx 1000 x x xxxxxxxxxxxxxx 1001 x x x xxxxxxxxxxxxx 1010 xxxxxxxxxxxxxxxx 1011 xxxxxxxxxxxxxxxx 1100 x xxxxxxxxxxxxxxx 1101 x x xxxxxxxxxxxxxx 1110 x x x xxxxxxxxxxxxx 1111 0 x x x xxxxxxxxxxxx shifter output xxxxxxxxxxxxxxxx threshold 1, threshold 2 xxxxxxxxxxxxxxxx db value 24 12 6 gain accumulator 0xxxxxxxxxxxxxxxx up delta programmed xxxxxxxxxxxxxx (14 bits) max delta = 12 db delta3 (lower) 1 1 1xxxxxxxxxxxxxx (accumulator + delta 1) delta 2 (upper) 0 0 0 x xxxxxxxxxxxxx (accumulator + delta 2) delta 1 (immed.) 0 0 0xxxxxxxxxxxxxx (accumulator + delta 3) up load bits 0xxxxxxxxxxxxxxxx leak ssxxxxxxxxxxxxxxx (accumulator + leak) msb is sign extended limits 0xxx to mapping lut x x x table 86. agc bit weights agc bit weights f edcba9 8 7 6 5 4 3 2 1 0 . 12345678901234567890123456789 forward gain path inputs i/q s.12345678901234567890123 isl5416
67 barrel shifter shift eeee = 0 s.12345678901234567890123 eeee = 4 s.12345678901234567890123 gggg eeee = 8 s.12345678901234567890123 gggggggg eeee = 15 s.12345678901234567890123 ggggggggggggggg ------------saturate on overflow--------- -- bs out / mult in s.12345678901234567890123 mult 1.0 <= m<= 2.0 gv12345678901234 mult out s012345678901234567890123 saturate based on shift and mult out: feedback path; outputs i/q s.12345678901234567890123 mag in s.12345678901234567890123 magnitude (g = 1.647, max in = 1.414, max out = 2.33): g1012345678901234567890123 threshold 2101234567890123 mean error s101234567890123 median error ggg.vgggggggggggg v vv.vgggggggggggg lg mantissa g.1234 lg mant mult out s10123456789012345 lg exp shift = 0 sssssssssssssssss 1012345678901234 lg exp shift = 4 sssssssssssss 10123456789012345ggg lg exp shift = 8 sssssssss 10123456789012345g gggggg lg exp shift = 15 ss10123456789012345 gggggggggggggg lg accum 3 2 1 0. 12345678901234567890123456789 ll / ul ggg0123456789012345 output fwd gain e e e emmmmmmmmmmmmmm output up / parallel g e e e emmmmmmmmmmm bit weights (db); max gain = 96.3293 db 421 842631000000000000 s = sign .................. g = logic 0 1 00005731000000000 e=exponent 684210578942100000 m = mantissa 421005268473152100 v = logic 1 842631531105799574 table 86. agc bit weights (continued) agc bit weights f edcba9 8 7 6 5 4 3 2 1 0 . 12345678901234567890123456789 isl5416
68 0.0 -1.0 -2.0 -3.0 -4.0 -5.0 -6.0 0.0 0.1 0.2 0.3 0.4 0.5 f s /r db n = 5 n = 4 n = 2 n = 1 n = 3 figure 19. cic passband rolloff (n = # of stages, r = decimation factor, f s /r = 1 is cic output rate) 0 -20 -40 -60 -80 -100 -120 0.00 0.10 0.20 0.30 0.40 0.50 f s /r db n = 3 n = 5 n = 2 n = 1 n = 4 -140 figure 19a. cic first alias level (n = # of stages, r = decimation factor, f s /r = 1 is cic output rate) 0 -20 -40 -60 -80 -100 -120 0.0 0.5 1.0 1.5 2.0 2.5 f s /r db -140 3.0 figure 19b. 5th order (n = 5) cic response (r = decimation factor, f s /r = 1 is cic 0 1 2 3 4 5 6 x 10 4 -180 -160 -140 -120 -100 -8 0 -6 0 -4 0 -2 0 0 figure 20. a typical carrier nco spectrum plot (65536 point fft, blackman window, 800 hz bin nbw (29 db). peak bin is at -132 dbc, noise floor at -190 dbc/hz. isl5416
69 data sheet revisions: cdma2000-1xrtt: figure below shows the overall response using 5-stage cic filter, 32-tap first fir filter block an d 64-tap second fir filter block. table 87. revision history revision date description 6006.2 january 10, 2003 - trst pin description text added for circuit board schematics - added text in the vga/rf attenuator (range control) section, last paragraph - figure 3, ?nco, mixer and cic block di agram?, corrected fi r coefficients table - figure 4, ?fir1 and fir2 block diagrams?, updated - added power-up sequenci ng explanation on page 27 - table 24, iwa = 0*01h, updated register des cription (bit values remained the same) - table 51, iwa = *000h, updated register des cription (bit values remained the same) - table 52, iwa = *001h, bits 31:29, updated description (bit values remained the same) - table 52, iwa = *001h, bits 2:0, updated desc ription (bit values remained the same) - table 61, iwa = *008h, bit 0 changed to ?reserved, set to 0? - table 62, ?control bit settings for agc modes? added - table 83, added the fir coefficient loadi ng procedure at the bottom of the table - in electrical specifications table added the following values: t chr , t rcy , t dw - figure 12, ?clock skews?, updated - added package drawing note (last page) 1 2 3 4 5 6 7 8 9 10 x 10 6 -120 -100 -80 -60 -40 -20 figure 21. overall filter response of a single cdma2000 channel isl5416
70 umts / w-cdma: figure below shows the overall response using 5-stage cic filter, 32-tap first fir filter block and 64-tap second fir filter block. 0.5 1 1.5 2 2.5 3 x 10 7 -100 -80 -60 -40 -20 0 figure 22. overall filter response of a single umts channel isl5416
71 all intersil u.s. products are manufactured, asse mbled and tested utilizin g iso9000 quality systems. intersil corporation?s quality certifications c an be viewed at www.intersil.com/design/quality intersil products are sold by description only. intersil corporation reserves the right to make changes in circuit design, soft ware and/or specifications at any time without notice. accordingly, the reader is cautioned to verify that data sheets are current before placing orders. information furnishe d by intersil is believed to be accurate and reliable. however, no responsibility is assumed by intersil or its subsidiaries for its use; nor for any infringements of paten ts or other rights of third parties which may result from its use. no license is granted by implication or otherwise under any patent or patent rights of intersil or its subsidiari es. for information regarding intersil corporation and its products, see www.intersil.com isl5416 plastic ball grid array packages (bga) o top view d a1 bottom view p n l m j k g h f e 8 13 14 12 11 10 9 corner 765 3 42 c d a b 1 side view seating plane c a a1 a2 bbb c aaa a1 e a b e1 d1 b s all rows and columns s m a b c c 0.15 0.08 m 0.006 0.003 a1 a1 corner i.d. e a a c r t 15 16 corner corner i.d. v256.17x17 256 ball plastic ball grid array package symbol inches millimeters notes min max min max a - 0.059 - 1.50 - a1 0.012 0.016 0.31 0.41 - a2 0.037 0.044 0.95 1.13 - b 0.016 0.020 0.41 0.51 7 d/e 0.665 0.673 16.90 17.10 - d1/e1 0.587 0.595 14.90 15.10 - n 256 256 - e 0.039 bsc 1.0 bsc - md/me 16 x 16 16 x 16 - bbb 0.004 0.10 3 0.005 0.12 - rev. 2 01/03 notes: 1. controlling dimension: millimeter. converted inch dimensions are not necessarily exact. 2. dimensioning and tolerancing conform to amse y14.5m-1994. 3. ?md? and ?me? are the maximum ball matrix size for the ?d? and ?e? dimensions, respectively. 4. ?n? is the maximum number of bal ls for the specific array size. 5. primary datum c and seating plane are defined by the spherical crowns of the contact balls. 6. dimension ?a? includes standoff height ?a1?, package body thick- ness and lid or cap height ?a2?. 7. dimension ?b? is measured at the maximum ball diameter, paral- lel to the primary datum c. 8. pin ?a1? is marked on the top and bottom sides adjacent to a1. 9. ?s? is measured with respect to datum?s a and b and defines the position of the solder balls near est to package centerlines. when there is an even number of balls in the outer row the value is ?s? = e/2.

▲Up To Search▲

Price & Availability of ISL5416KIZ

	To Download ISL5416KIZ Datasheet File
If you can't view the Datasheet, Please click here to try to view without PDF Reader .