View MAX1460_4825574.PDF datasheet online --- IC-ON-LINE

Datasheet File OCR Text:

general description the MAX1460 implements a revolutionary concept in signal conditioning, where the output of its 16-bit ana- log-to-digital converter (adc) is digitally corrected over the specified temperature range. this feature can be readily exploited by automotive, industrial, and medical market segments, in applications such as sensors and smart batteries. digital correction is provided by an internal digital signal processor (dsp) and on-chip 128- bit eeprom containing user-programmed calibration coefficients. the conditioned output is available as a 12-bit digital word and as a ratiometric (proportional to the supply voltage) analog voltage using an on-board 12-bit digital-to-analog converter (dac). the uncommit- ted op amp can be used to filter the analog output, or implement a 2-wire, 4?0ma transmitter. the analog front end includes a 2-bit programmable- gain amplifier (pga) and a 3-bit coarse-offset (co) dac, which condition the sensor? output. this coarsely corrected signal is digitized by a 16-bit adc. the dsp uses the digitized sensor signal, the temperature sen- sor, and correction coefficients stored in the internal eeprom to produce the conditioned output. multiple or batch manufacturing of sensors is support- ed with a completely digital test interface. built-in testa- bility features on the MAX1460 result in the integration of three traditional sensor-manufacturing operations into one automated process: pretest: data acquisition of sensor performance under the control of a host test computer. calibration and compensation: computation and storage of calibration and compensation coefficients determined from transducer pretest data. final test operation: verification of transducer cali- bration and compensation, without removal from the pretest socket. the MAX1460 evaluation kit (ev kit) allows fast evalua- tion and prototyping, using a piezoresistive transducer (prt) and a windows -based pc. the user-friendly ev kit simplifies small-volume prototyping; it is not necessary to fully understand the test-system interface, the calibra- tion algorithm, or many other details to evaluate the MAX1460 with a particular sensor. simply plug the prt into the ev kit, plug the ev kit into a pc parallel port, con- nect the sensor to an excitation source (such as a pres- sure controller), and run the MAX1460 ev kit software. an oven is required for thermal compensation. features low-noise, 400a single-chip sensor signal conditioning high-precision front end resolves less than 1v of differential input signal on-chip dsp and eeprom provide digital correction of sensor errors 16-bit signal path compensates sensor offset and sensitivity and associated temperature coefficients 12-bit parallel digital output analog output compensates a wide range of sensor sensitivity and offset single-shot automated compensation algorithm?o iteration required built-in temperature sensor three-state, 5-wire serial interface supports high-volume manufacturing ________________________applications hand-held instruments piezoresistive pressure and acceleration transducers and transmitters industrial pressure sensors and 4?0ma transmitters smart battery charge systems weigh scales and strain-gauge measurement flow meters dive computers and liquid-level sensing hydraulic systems automotive systems customization maxim can customize the MAX1460 for unique require- ments. with a dedicated cell library of more than 90 sen- sor-specific functional blocks, maxim can quickly provide customized MAX1460 solutions, including customized microcode for unusual sensor characteristics. contact maxim for further information. MAX1460 low-power, 16-bit smart adc ________________________________________________________________ maxim integrated products 1 19-4784; rev 0; 10/99 part MAX1460ccm 0? to +70? temp. range pin-package 48 tqfp ordering information functional diagram appears at end of data sheet. pin configuration appears at end of data sheet. windows is a registered trademark of microsoft corp. evaluation kit available for free samples & the latest literature: http://www.maxim-ic.com, or phone 1-800-998-8800. for small orders, phone 1-800-835-8769.
pga gain MAX1460 low-power, 16-bit smart adc 2 _______________________________________________________________________________________ absolute maximum ratings electrical characteristics (v dd = +5v, v ss = 0, f xin = 2mhz, t a = t min to t max , unless otherwise noted.) stresses beyond those listed under ?bsolute maximum ratings?may cause permanent damage to the device. these are stress rating s only, and functional operation of the device at these or any other conditions beyond those indicated in the operational sections of the specificatio ns is not implied. exposure to absolute maximum rating conditions for extended periods may affect device reliability. supply voltage, v dd to v ss ......................................-0.3v to +6v all other pins ...................................(v ss - 0.3v) to (v dd + 0.3v) short-circuit duration, all outputs .............................continuous continuous power dissipation (t a = +70?) 48-pin tqfp (derate 12.5mw/? above +70? ).....1000mw operating temperature range...............................0? to +70? storage temperature range .............................-65? to +160? lead temperature (soldering, 10sec) .............................+300? co-dac code = 011 co-dac code = 010 continuous conversion co-dac code = 001 co-dac code = 000 during operation co-dac code = 100 co-dac code = 101 co-dac code = 110 co-dac code = 111 pga gain code = 11 pga gain code = 10 from v ss to v dd pga gain code = 01 pga gain code = 00 conditions 134 149 164 81 96 111 32 47 62 -20 -5 10 -10 5 20 -62 -47 -32 -111 -96 -81 % v dd -164 -149 -134 coarse offset 90 93 96 74 77 80 59 61 64 v/v 43 46 49 ? 400 700 i dd supply current (note 2) v 4.75 5.0 5.25 v dd supply voltage (note 1) db 90 cmrr common-mode rejection ratio nv/? ?200 input-referred offset tc hz 15 throughput rate m ? 1.0 r in input impedance ppm/? ?0 gain temperature coefficient (tc) units min typ max symbol parameter t a = 0? to +70? 5k ? input impedance pga gain code = 00, co-dac code = 000 ? 1.3 linearity lsb/? 260 resolution lsb rms 2 output-referred noise nv rms 1700 input-referred noise % 0.006 inl integral nonlinearity (note 5) bits 16 resolution adc (notes 3, 4) temperature sensor (note 6) general characteristics analog input pga and coarse-offset dac (notes 3, 4) pga gain
MAX1460 low-power, 16-bit smart adc _______________________________________________________________________________________ 3 conditions units min typ max symbol parameter dac resolution 12 bits integral nonlinearity inl 1 lsb differential nonlinearity dnl 0.5 lsb op amp supply current 100 ? input common-mode range cmr v ss + 1.3 v dd - 1.0 v open-loop gain a v 60 db offset voltage (as unity-gain follower) v os v in = 2.5v (no load) -30 +30 mv output voltage swing no load v ss + 0.05 v dd - 0.05 v output current range v out = (v ss + 0.2v) to (v dd - 0.2v) ?00 ? note 1: eeprom programming requires a minimum v dd = 4.75v. i dd may exceed its limits during this time. note 2: this value does not include the sensor or load current. this value does include the uncommitted op amp current. note that the MAX1460 will convert continuously if repeat mode is set in the eeprom. note 3: see the analog front-end, including pga, coarse offset dac, adc, and temperature sensor sections. note 4: the signal input to the adc is the output of the pga plus the output of the co-dac. the reference to the adc is v dd . the plus full-scale input to the adc is +v dd and the minus full-scale input to the adc is -v dd . this specification shows the con- tribution of the co-dac to the adc input. note 5: see figure 2 for adc outputs between +0.8500 to -0.8500. note 6: the sensor and the MAX1460 must always be at the same temperature during calibration and use. note 7: the output dac is specified using the external lowpass filter (figure 8). note 8: sdio is an input/output digital pin. it is only enabled as a digital output pin when the MAX1460 receives from the test sys- tem the commands 8 hex or a hex (table 4). note 9: xin is a digital input pin only when the test pin is high. note 10: guaranteed by design. not subject to production testing. electrical characteristics (continued) (v dd = +5v, v ss = 0, f xin = 2mhz, t a = t min to t max , unless otherwise noted.) input high voltage v ih 4.0 v input low voltage v il 1.0 v input hysteresis v hyst 1.0 v input leakage i in v in = 0 or v dd ?0 ? input capacitance c in (note 10) 50.0 pf output voltage low v ol i sink = 500? 0.5 v output voltage high v oh i source = 500? 4.5 v three-state leakage current i l cs = 0 ?0 ? three-state output capacitance c out cs = 0 (note 10) 50.0 pf output voltage low v ol i sink = 500? 0.3 v output voltage high v oh i source = 500? 4.7 v three-state leakage current i l cs = 0 ?0 ? three-state output capacitance c out cs = 0 (note 10) 50.0 pf output dac (note 7) uncommitted op amp digital inputs: start, cs1, cs2, sdio (note 8) , reset, xin (note 9) , test digital outputs: d[11...0] digital outputs: sdio (note 8) , sdo, eoc, out
MAX1460 low-power, 16-bit smart adc 4 _______________________________________________________________________________________ name function 1, 2, 12, 13, 18, 19, 31, 32, 36, 41?5 n.c. no connection. not internally connected. 3 agnd analog ground. connect to v dd and v ss using 10k ? resistors (see functional diagram ). pin 4 start optional conversion start input signal, used for extending sensor warm-up time. internally pulled to v dd with a 1m ? (typical) resistor. 5 i.c. internally connected. leave unconnected. 9 d9 parallel digital output - bit 9 8 d8 parallel digital output - bit 8 7 d7 parallel digital output - bit 7 6 d6 parallel digital output - bit 6 16, 17 cs1, cs2 chip-select input. the MAX1460 is selected when cs1 and cs2 are both high. when either cs1 or cs2 is low, all digital outputs are high impedance and all digital inputs are ignored. cs1 and cs2 are internally pulled high to v dd with a 1m ? (typical) resistor. 15 v ss negative supply input 14, 37, 38 v dd positive supply voltage input. connect a 0.1? bypass capacitor from v dd to v ss . pins 14, 37, and 38 must all be connected to the positive power supply on the pcb. 11 d11 parallel digital output - bit 11 (msb) 10 d10 parallel digital output - bit 10 20 sdio serial data input/output. used only during programming/testing, when the test pin is high. the test system sends commands to the MAX1460 through sdio. the MAX1460 returns the current instruction rom address and data being executed by the dsp to the test system. sdio is internally pulled to v ss with a 1m ? (typical) resistor. sdio goes high impedance when either cs1 or cs2 is low and remains in this state until the test system initiates conversion. 21 sdo serial data output. used only during programming/testing. sdo allows the test system to monitor the dsp registers. the MAX1460 returns to the test system results of the dsp current instruction. sdo is high impedance when test is low. 22 reset reset input. when test is high, a low-to-high transition on reset enables the MAX1460 to accept commands from the test system. this input is ignored when test is low. internally pulled high to v dd with a 1m ? (typical) resistor. 26 d2 parallel digital output - bit 2 25 d1 parallel digital output - bit 1 24 d0 parallel digital output - bit 0 (lsb) 23 eoc end of conversion output. a high-to-low transition of the eoc pulse can be used to latch the parallel digital output (pins d[11...0]). pin description
MAX1460 low-power, 16-bit smart adc _______________________________________________________________________________________ 5 rail-to-rail is a registered trademark of nippon motorola, ltd. pin description (continued) name function pin 33 ampout general-purpose operational amplifier output 30 out output dac. the bitstream on out, when externally filtered, creates a ratiometric analog output volt- age. out is proportional to the 12-bit parallel digital output. 29 d5 parallel digital output - bit 5 28 d4 parallel digital output - bit 4 27 d3 parallel digital output - bit 3 34 amp+ noninverting input of general-purpose operational amplifier 35 amp- inverting input of general-purpose operational amplifier 39 xout internal oscillator output. connect a 2mhz ceramic resonator (murata cst200) or crystal from xout to xin. 40 xin internal oscillator input. when test is high, this pin must be driven by the test system with a 2mhz, 50% duty cycle clock signal. the resonator does not need to be disconnected in test mode. 46 inp positive sensor input. input impedance is typically > 1m ? . rail-to-rail input range. 47 test test/program mode enable input. when high, enables the MAX1460 programming/testing operations. internally pulled to v ss with a 1m ? (typical) resistor. 48 inm negative sensor input. input impedance is typically > 1m ? . rail-to-rail input range.
MAX1460 low-power, 16-bit smart adc 6 _______________________________________________________________________________________ detailed description the main functions of the MAX1460 include: analog front end: includes pga, coarse-offset dac, adc, and temperature sensor test system interface: writes calibration coefficients to the dsp registers and eeprom test system interface: observes the dsp operation. the sensor signal enters the MAX1460 and is adjusted for coarse gain and offset by the analog front end. five bits in the configuration register set the coarse-offset dac and the coarse gain of the pga (tables 1 and 2). these bits must be properly configured for the optimum dynamic range of the adc. the digitized sensor signal is stored in a read-only dsp register. the on-chip temperature sensor also has a 3-bit coarse-offset dac that places the temperature signal in the adc operating range. digitized temperature is also stored in a read-only dsp register. the dsp uses the digitized sensor, the temperature signals, and the cor- rection coefficients to calculate the compensated and corrected output. the MAX1460 supports an automated production envi- ronment, where a test system communicates with a batch of MAX1460s and controls temperature and sen- sor excitation. the three-state digital outputs on the MAX1460 allow parallel connection of transducers, so that all five serial interface lines (xin, test, reset , sdio, and sdo) can be shared. the test system selects an individual transducer using cs1 and cs2. the test system must vary the sensor? input and tem- perature, calculate the correction coefficients for each unit, load the coefficients into the MAX1460 nonvolatile eeprom, and test the resulting compensation. the MAX1460 dsp implements the following character- istic equation: where gain corrects the sensor? sensitivity, g 1 and g 2 correct for gain-tc, t and signal are the digitized out- puts of the analog front end, of 0 corrects the sensor? offset, of 1 and of 2 correct the offset-tc, and d off is the output offset pedestal. the test system can write the calibration coefficients into the MAX1460 eeprom or write to the dsp regis- ters directly. the MAX1460 can begin a conversion using either the eeprom contents or the register con- tents. when the test system issues commands, the MAX1460 is a serially controlled slave device. the test system observes the MAX1460 dsp operation in order to acquire the temperature and signal adc results, to verify the calibration coefficients, and to get the output d. the MAX1460 places the contents of sev- eral important dsp registers on the serial interface after the tester issues a start conversion command. after calibration, compensation, and final test, the MAX1460 is adapted to its sensor and the pair can be removed from the test system. use the resulting trans- d gain 1 g t g t signal of of t of t d 2 2 01 2 2 off =++ () ++ + ? ? ? ? + 1 93 1 1 3 77 0 1 2 61 1 0 1 46 0 0 0 nominal gain (v/v) pga-0 pga-1 pga setting 1 (mv rti) (v dd = 5v) co-0 co-1 co-s pga setting 3 (mv rti) (v dd = 5v) pga setting 2 (mv rti) (v dd = 5v) pga setting 0 (mv rti) (v dd = 5v) % v dd (at adc input) co setting 122 79 39 80 97 162 1 1 0 +3 149 0 1 0 +2 52 1 0 0 +1 62 104 96 25 31 51 47 -4 4 -39 -79 -122 -3 -3 -5 -5 3 3 5 5 -25 -31 -47 -52 -62 -104 -96 -80 -97 -162 -149 0 0 0 +0 0 0 1 -0 1 0 1 -1 0 1 1 -2 1 1 1 -3 table 2. typical coarse offset dac settings table 1. nominal pga gain settings pga setting -51
MAX1460 low-power, 16-bit smart adc _______________________________________________________________________________________ 7 ducer by applying power and the start signal. latch the 12-bit parallel digital output using the eoc pulse. the maximum conversion rate of the MAX1460 is 15hz, using a 2mhz resonator. if an analog output is desired, build a simple lowpass filter using the out pin, the uncommitted op amp, and a few discrete components (figure 8). analog front end, including pga, coarse offset dac, adc, and temperature sensor before the sensor signal is digitized, it must be gained and coarse-offset corrected to maximize the adc dynamic range. there are 2 bits (four possible settings) in the configuration register for the pga gain, and 3 bits (eight possible settings) for the co dac. the flowchart (figure 1) shows a procedure for finding the optimum ?ake a test system variable called ?omoregain. ?et the temperature to where the sensor? sensitivity is highest. this is normally cold for silicon prts. ?et the pga gain settings to minimum. ?lear the variable ?omoregain. ?pply maximum sensor excitation. ?est for clipping (digitized signal > 0.85). record the pga and coarse offset settings. caution: clipping is still possible for large sensor? offset tc and large temperature ranges. if necessary, guardband against clipping by reducing the ?.85 clipping constants above. ?pply midscale excitation to the sensor. ?ind the coarse offset dac setting where the digitized signal register is closest to zero (midscale). ?educe the pga gain one step. ?et the variable ?omoregain. increase the pga gain one step. did adc clip? is the pga at minimum gain? the sensor sensitivity is too large. add a resistor between the top of the bridge and v dd , then start over. yes yes series resistor sensor no yes yes no no no ?pply minimum sensor excitation. ?est for clipping (digitized signal < -0.85). v dd is the pga at maximum gain? is ?omoregain?set? figure 1. flowchart for determining pga and co settings
MAX1460 low-power, 16-bit smart adc 8 _______________________________________________________________________________________ analog front-end settings when the sensor? character- istics are unknown. use the tabulated values (tables 1 and 2) if the peak sensor excursions are known. see the test system interface section for details on writing these analog front-end bits. the pga gain and the co are very stable, but are not accurate. manufacturing variances on the gain and off- set of the MAX1460 analog front-end superposition the residual sensor errors, and are later removed during final calibration. for example, suppose the sensor? sensitivity is +10mv/v with an offset of -12mv/v. let the supply volt- age be +5v. the full scale (-fs) output of the sensor is then +5v(-12mv/v) = -60mv; +fs is then +5v (-12mv/v + 10mv/v) = -10mv. following through the flowchart, the pga gain setting is +3 (gain = 93v/v) and the co correction setting is +1 (+25mv rti) - (referred-to input). the coarsely corrected -fs input to the adc is (-60mv + 25mv)93 = -3.255v. the +fs input to the adc is (-10mv + 25mv)93 = +1.395v. the input range of the adc is ? dd . thus the maximum and minimum digitized sensor signals become -3.255 / 5 = -0.651 and +1.395 / 5 = +0.279. notice that the bridge multiplies the signal by v dd and the adc divides the signal by v dd . thus, the system is ratiometric and not dependent on the dc value of v dd . the adc output clips to ?.0 when input values exceed ? dd . the best signal-to-noise ratio (snr) is achieved when the adc input is within ?5% of v dd (figure 2). the MAX1460 includes an internal temperature-sensing bridge allowing the MAX1460 temperature to be used as a proxy for the sensor temperature. for this reason, the MAX1460 must be mounted in thermal proximity to the sensor. the output of the temperature-sensing bridge is also corrected by a 3-bit coarse-offset dac and processed by the adc. the selection of the temperature sensor offset (tso) bits in the configura- tion register should be made so that the digitized tem- perature signal is as close to 0.0 as possible at midscale temperature. this is done to maximize the dynamic range of the thermal-calibration coefficients (table 3). -0.010 -0.006 -0.008 -0.002 0.004 0.002 0 0.004 0.008 0.006 0.010 -100 -60 -40 -20 -80 0 20 40 80 60 100 sensor signal input or adc input/output range (%) nonlinearity error (%fs) figure 2a. analog front-end inl (typical) -4 -3 0 -1 -2 1 2 3 4 -100 -20 -40 -80 -60 0 20 40 60 80 100 sensor signal input or adc input range (%) error (16-bit lsbs) figure 2b. analog front-end differential nonlinearity (dnl) (typical) 0 1.0 0.5 1.5 3.0 3.5 2.5 2.0 4.0 -100 -60 -40 -20 0 -80 20 40 60 80 100 sensor signal input or adc input/output range (%) noise standard deviation (16-bit lsbs) figure 2c. analog front-end noise standard deviation of the samples (typical)
MAX1460 low-power, 16-bit smart adc _______________________________________________________________________________________ 9 test-system interface: writing calibration coefficients to the dsp registers and eeprom to make the MAX1460 respond to commands from the test system, raise the test pin and drive xin with a 2mhz clock signal. it is not necessary to remove the resonator. reset must be low for at least 16 clock cycles to initialize the MAX1460. then, a rising transi- tion on reset begins a 32-bit serial transfer of the test- system command word through sdio. the test system transitions sdio on falling edges of the xin clock; the MAX1460 latches data is on the rising edge (figure 3). the 32-bit command word generated by the test-sys- tem is divided into four fields (figure 3). the 4-bit com- mand field is interpreted in table 4. the other fields are usually ignored, except that command 1 hex uses the two register fields, and command 2 hex requires an eeprom address. the command word fields are: register data field: holds the calibration coeffi- cients to be written into the MAX1460 16-bit registers eeprom address field: holds the hexadecimal address of the eeprom bit to be set (from 00 hex to 7f hex) register address field: contains the address of the register (0 to 7) where the calibration coefficient is to be written command field: instructs the MAX1460 to take a particular action (table 4) min 16 clk cycles command 1 register data field eeprom address field reg. add command 2 command 3 command n command 00 01 02 03 29 30 31 00 01 02 03 29 30 31 00 01 02 03 29 30 31 00 01 02 03 29 30 31 d0 d1 d2 d3 c3 nu nu d0 d1 d2 d3 c3 nu nu d0 d1 d2 d3 c3 nu nu d0 d1 d2 d3 c3 nu nu d0 d1 d2 d3 d4 d5 d6 d7 d8 d9 d10 d11 d12 d1 3 d14 d15 e0 e1 e2 e3 e4 e5 e6 r0 r1 r2 c0 c1 c2 c3 nu nu msb lsb lsb msb lsb msb lsb msb note: all transitions must occur within 100ns of the xin clock edge. xin test reset sdio figure 3. test-system command timing diagram tso-0 tso-1 tso-2 temperature bridge offset tso setting 1 1 1 7 0 1 1 6 1 0 1 5 minimum maximum 0 0 1 4 1 1 0 3 0 1 0 2 1 0 0 1 0 0 0 0 table 3. temperature sensor offset (tso) settings
MAX1460 low-power, 16-bit smart adc 10 ______________________________________________________________________________________ writing to the dsp registers command 1 hex writes calibration coefficients from the test system directly into the dsp registers. tester com- mands 8 hex and c hex cause the MAX1460 to start a conversion using the calibration coefficients in the reg- isters. this direct use of the registers speeds calibra- tion and compensation because it does not require eeprom write-access time. bringing reset low clears the dsp registers, so the test system should always write to the registers and start a conversion in a single command timing sequence. as shown in table 5, seven registers hold the calibra- tion coefficients of the characteristic equation [d out = gain (1+g 1 t + g 2 t 2 ) (signal + of 0 + of 1 t + of 2 t 2 ) + d off ] implemented by the MAX1460 dsp. all of the registers are 16-bit, two? complement coding format. when a register is interpreted as an integer, the deci- mal range is from -32768 (8000 hex) to +32767 (7fff hex). fractional coefficient values range from -1.0 (8000 hex) to +0.99997 (7fff hex). the register at address 0 is called the configuration register. it holds the coarse offset, pga gain, op amp power-down , temperature-sensor offset, repeat mode, and reserved bits, as shown in table 6. the functionali- ty of the coarse offset, pga gain, and temperature-sen- sor bits are described in the analog front end section. the op amp power-down bit enables the uncommitted op amp when set. the repeat-mode bit is tested by the last instruction of the dsp microcode, and, if set, imme- diately initiates another conversion cycle. the maxim reserved bits should not be altered. write a calibration coefficient into a dsp register. 1 0 1 hex 4 hex 0 0 0 0 0 2 hex 0 hex 0 0 8 hex block-erase the entire eeprom (writes ??to all 128 bits). 0 1 0 a hex write ??to a single eeprom bit. 0 0 1 0 noop (no-operation) 0 0 0 0 start conversion command. the registers are not updated with eeprom values. sdio and sdo are enabled as dsp outputs. 1 0 0 c hex e hex c0 hex code reserved 3, 5, 6, 7, 9, b, d, f hex - - - - start conversion command. the registers are updated with eeprom values. sdio and sdo are enabled as dsp outputs. 1 0 1 start conversion command. the registers are not updated with eeprom values. sdio and sdo are disabled. 1 1 0 start conversion command. the registers are updated with eeprom values. sdio and sdo are disabled. 1 1 1 command c3 c2 c1 table 4. test system commands gain 6 7 1 2 of 2 gain correction -32768 to +32767 integer quadratic tc offset 3 4 -1.0 to +0.99997 fraction d off g 1 linear tc gain -1.0 to +0.99997 fraction output midscale pedestal g 2 quadratic tc gain -1.0 to +0.99997 fraction -32768 to +32767 of 0 offset correction -1.0 to +0.99997 fraction integer register address of 1 5 linear tc offset -1.0 to +0.99997 fraction coefficient function range format table 5. dsp calibration coefficient registers
MAX1460 low-power, 16-bit smart adc ______________________________________________________________________________________ 11 writing to the internal eeprom the test system writes to the eeprom with commands 4 hex (block-erase the entire eeprom), 2 hex (write ??to a single eeprom bit) and 0 hex (noop). during normal operation (when the test pin is low) or when the test system issues instructions a hex or e hex (start conversion from eeprom values), the dsp reads the calibration coefficients from the eeprom. in the normal production flow, determine the calibration coefficients using direct register access. then load the calibration coefficients into the eeprom with tester instruction 2 hex. instruction 4 hex block-erases the eeprom and is necessary only for a rework or reclaim operation. for each part, the maxim reserved bits in the configuration register should be read before instruc- tion 4 hex is issued, and restored afterwards. the MAX1460 is shipped with its internal eeprom uninitial- ized, except for the reserved bits. the internal 128-bit eeprom is arranged as eight 16- bit words. these eight words are the configuration register and the seven calibration-coefficient values (table 7). the MAX1460 eeprom is bit addressable. the final cal- ibration coefficients must be mapped into the eeprom locations that are to be set. there is no bit- clear instruction. any eeprom write operation is neces- sarily long because the internal charge pump must cre- ate and maintain voltages above 20v long enough to cause a reliably permanent change in the memory. writing an eeprom bit requires 6ms, so writing the eeprom typically requires less than 400ms. do not decrease the eeprom write times. to write an eeprom bit, the test system must be compli- ant with the command timing diagram shown in figure 3, performing the following operations: 1) issue command 0 hex, including the eeprom address field of the bit to be written. 2) issue command 2 hex, with the address field used in step 1. continuously repeat this command 375 times (6ms). 3) issue command 0 hex, including the eeprom address field used in steps 1 and 2. the procedure for using command 4 hex (block-erase the eeprom) is similar. record the maxim reserved bits in the configuration register prior to using this com- mand, and restore them afterwards. the number of block-erase operations should not exceed 100. 1) issue command 0 hex. 2) issue command 4 hex. continuously repeat this command 375 times (6ms). 3) issue command 0 hex. test system interface: observing the dsp operation test system commands 8 hex and a hex initiate a con- version while allowing the test system to observe the operation of the dsp. to calibrate a unit, the test sys- tem must know the digitized temperature and sensor signals, stored in dsp registers 8 and 9, and the cali- brated and compensated output stored in dsp register 10. the test system should also verify the eeprom con- tents, registers 0?. all these signals pass through dsp register s during the execution of the instruction rom microcode. the sdo pin outputs the s register values, and the sdio pin tells the tester which signal is currently on s. 01 co-0 (lsb) 09 maxim reserved tso-0 (lsb) co-1 (msb) 0 (lsb) 8 tso-1 0a 9 0b co-s (sign) 10 tso-2 (msb) 02 1 0c 11 0d 03 2 12 maxim reserved pga-1 (msb) 04 3 maxim reserved maxim reserved 0e 05 4 13 0f 14 repeat mode pga-0 (lsb) 10 15 (msb) maxim reserved maxim reserved 06 5 07 6 op amp power-down 08 7 eeprom address (hex) description bit position table 6. configuration register bitmap
MAX1460 low-power, 16-bit smart adc 12 ______________________________________________________________________________________ there are three internal dsp registers that are directly observable on the sdio and sdo pins: s: 16-bit dsp scratch or accumulator register, con- taining the result of the execution of the current microcode instruction. p: 8-bit dsp program pointer register, which holds the address of the instruction rom microcode. ps: 8-bit dsp program store register. ps is the instruction that the dsp is currently executing. ps is the instruction rom data at address p. the dsp instructions relevant to the test system are list- ed in table 8. after the test system sends the start conversion com- mands 8 hex or a hex, sdio and sdo are both enabled as MAX1460 serial outputs. the test system should disable (high impedance) its sdio driver to avoid a bus conflict at this time so that the MAX1460 can drive the pin. after the dsp executes each one of the microcode instructions, the contents of the registers s, p, and ps are output in a serial format (figure 4). a new dsp instruction and a new state of the s, p, and ps registers are delivered every 16n + 9 clock cycles, where n = 0, 1, 2... after the start conversion command completes. the tester should latch the sdio and sdo 0f msb configuration lsb 0d 0e 1f msb 0b gain lsb ee address (hex) 10 1d 09 contents 1e 0a 1b ee address (hex) 20 19 contents 1a 0c 1c 18 16 08 17 14 12 06 13 15 11 07 2f msb g 1 04 lsb 2d 02 2e 03 3f msb 05 2b g 2 lsb 01 ee address (hex) 30 3d 29 contents 3e 2a 3b ee address (hex) 40 39 contents 3a 2c 3c 38 36 28 37 34 32 26 33 35 31 27 24 22 23 25 21 4f msb of 0 lsb 4d 4e 5f msb 4b of 1 lsb ee address (hex) 50 5d 49 contents 5e 4a 5b ee address (hex) 60 59 contents 5a 4c 5c 58 56 48 57 54 52 46 53 55 51 47 6f msb of 2 44 lsb 6d 42 6e 43 7f msb 45 6b d off lsb 41 ee address (hex) 70 7d 69 contents 7e 6a 7b ee address (hex) 00 79 contents 7a 6c 7c 78 76 68 77 74 72 66 73 75 71 67 64 62 63 65 61 table 7. eeprom memory map d7 register 7? off register 8?emperature signal register 0?onfiguration 56 register 9?ensor signal d8 01 d9 register 1?ain 3b register 10compensated output d d0 66 or 6c ea 65 or 6b d1 47 register 2? 1 d2 11 d3 2e register 3? 2 register 4?f 0 register 5?f 1 d4 38 d5 03 register 6?f 2 d6 22 instruction code (ps) (hex) s register value program counter (p) (hex) table 8. subset of dsp instruction
MAX1460 low-power, 16-bit smart adc ______________________________________________________________________________________ 13 s0 s1 s2 s3 s4 s5 s6 s7 s8 s9 s10 s11 s12 s13 s14 s15 s0 s1 s2 s3 s4 s5 s6 s7 s8 s9 s10 s11 ps4 ps5 ps6 ps7 p0 p1 p2 p3 p4 p5 p6 p7 ps0 ps 1 ps2 ps3 ps4 ps5 ps6 ps7 p0 p1 p2 p3 p4 p5 p6 p7 ps0 ps1 ps2 ps3 note: all transitions must occur within 100ns of the xin clock edge. lsb lsb msb lsb msb msb (16 n + 9) th clock cycle (16 (n + 1) + 9) th clock cycle dsp cycle n-1 dsp cycle n dsp cycle n+1 xin sdo sdio s12 s13 s14 s15 figure 4. dsp serial output timing diagram bits on the falling edge of the xin clock signal. when the p and ps registers in table 8 appear on sdio, the tester should save the corresponding sdo data. the conversion timing of the MAX1460 is shown in figure 5 and table 9. in the figure, the conversion is initiated by a rising transition on the start pin. equivalently, conversion can be initiated in test mode after completion of tester commands 8 hex or a hex, or reinitiated by the state of the repeat mode bit in the configuration register. after a conversion is initiated, the 16-bit adc digitizes the temperature and sensor signals during t adc . then, the dsp executes the instruction rom microcode during t dsp . in test mode, and during t dsp , sdio and sdo outputs carry useful information. at 130,586 clock cycles after the start conversion command is received, the lsb of the s and p dsp registers is available on sdo and sdio. the last dsp instruction is d0 hex. the tester can now start a new communication sequence by lowering the reset pin for at least 16 clock cycles, and then resume dri- ving sdio. sdio becomes high impedance when reset is low. v dd start (optional) sdio & sdo (test mode) d [11...0] t warm t eoc t conv t adc t dsp eoc figure 5. MAX1460 conversion timing
MAX1460 low-power, 16-bit smart adc 14 ______________________________________________________________________________________ applications information calibration and compensation procedure perform fine calibration by characterizing the sensor/ MAX1460 pair using the test system and then finding the calibration coefficients gain, g 1 , g 2 , of 0 , of 1 , and of 2 using the equations below. this simple fine-calibra- tion procedure requires three temperatures, denoted a, b, and c, and two sensor excitations, named s and l for small and large. thus, there are six data points (as, al, bs, bl, cs, and cl); six unknown calibration coeffi- cients; and six versions of the characteristic equation, in the form: equation (1) where d l , d s , and d off are determined by the end product specification. d l is the desired MAX1460 out- put corresponding to the l sensor excitation; d s is the desired MAX1460 output corresponding to the s sen- sor excitation; d off is the desired midscale output; signal cl is the digitized sensor reading at temperature c with the l sensor excitation applied; and t c is the digitized temperature reading at temperature c. unstable digitized temperature readings indicate that thermal equilibrium has not been achieved, necessitat- ing increased soak times or a better thermal control. averaging many readings from the MAX1460 will help filter out ac variations in the sensor excitation and oven temperature. begin calibration by soaking the sensor and the MAX1460 pair at the first temperature, a, and apply the l excitation to the sensor. start a conversion and record the digitized temperature t a and the digitized signal signal al . apply the s sensor excitation, and record the digitized signal signal as . repeat this proce- dure for temperatures b and c, recording t b , signal bl , signal bs , t c , signal cl , and signal cs . the a l and a s versions of equation 1 may be ratioed to obtain: equation (2a) similarly, equation (2b) equation (2c) where equation (3) equations 2a, 2b, and 2c form a system of three linear equations, with three unknowns, of 0 , of 1 , and of 2 . solve for of 0 , of 1 , and of 2 . the small sensor excitation versions of equation 1 can be ratioed to obtain: equation (4a) equation (4b) y y g ty ty g ty ty 0 cs bs 1 b cs c bs 2b 2 cs c 2 bs ? () +? () + ? ? ? ? ? = yy g tyty g t y t y 0 cs as 1 a cs c as 2a 2 cs c 2 as ? () +? () + ? ? ? ? ? = x dd dd l off s off = ? ? signal x signal 1x of of t of t 0 cl cs 01c2c 2 ? ? ++ + = ? signal x signal 1x of of t of t 0 bl bs 01b2b 2 ? ? ++ + = ? signal x signal 1x of of t of t 0 al as 01a2a 2 ? ? ++ + = ? d d gain 1 g t g t signal of of t of t l off 1 c 2 c 2 cl 0 1 c 2 c 2 ?= ++ ? ? ? ? ++ + ? ? ? ? table 9. MAX1460 conversion timing parameter symbol min max units sensor warm-up time t warm 35 ms adc time t adc 130,585 130,585 xin clk cycles dsp time t dsp 3,220 3,364 xin clk cycles eoc pulse width t eoc 8 8 xin clk cycles conversion time t conv 133,805 133,949 xin clk cycles
MAX1460 low-power, 16-bit smart adc ______________________________________________________________________________________ 15 where: equation (5a) equation (5b) equation (5c) equations 4a and 4b form a system of two linear equa- tions and two unknowns, g 1 and g 2 . solve for g 1 and g 2 . equation 1 can now be readily solved for the last unknown, gain. arithmetic manipulation can magnify measurement errors and noise. quantization of the calibration coeffi- cients is another reason to consider adjusting the gain and d off coefficients. to do this, load the MAX1460 registers with the calculated coefficients gain, g 1 , g 2 , of 0 , of 1 , of 2 , and d off . assuming the oven is still at temperature c and the s sensor excitation is still applied, measure the output d cs . change to the l sen- sor excitation, and measure d cl . compute the new gain coefficient using equation 6. remeasure d cl , and compute the new d off coefficient, given by equation 7. equation (6) equation (7) the final calibration coefficients may now be written into the MAX1460 eeprom. the unit is now ready for final test. dddd off new off l cl =+? gain gain dd dd new ls cl cs = ? ? y dd signal of of t of t cs s off cs 0 1 c 2 c 2 = ? ++ + y dd signal of of t of t bs s off bs 0 1 b 2 b 2 = ? ++ + y dd signal of of t of t as s off as 0 1 a 2 a 2 = ? ++ + -10 -4 -6 -8 0 -2 8 6 4 2 fso 10 0 10203040506070 uncompensated sensor error temperature ( c) error (%fso) offset figure 6. sensor characteristics before compensation -0.20 -0.10 -0.15 0 -0.05 0.15 0.10 0.05 0.20 020 10 30 40 50 60 70 compensated transducer error temperature ( c) error (% span, 4000 codes) fso offset figure 7. compensated sensor/MAX1460 pair agnd amp- amp+ out unfiltered bitstream filtered analog output r d1 10k r d2 10k ampout c f 1 f r f 500k r1 500k MAX1460 op amp v dd figure 8. filtering the output dac
MAX1460 low-power, 16-bit smart adc 16 ______________________________________________________________________________________ this algorithm minimizes the error directly at the six test conditions, as, al, bs, bl, cs, and cl. space the temperatures a, b, and c widely to minimize the signal- to-noise ratio of the measurement. if there is a large error remaining in the finished product, move the cali- bration temperatures closer to the peak error tempera- tures. similarly, full-scale sensor excitation may not be the best calibration condition if the sensor has non- linearities. move s and l away from full scale. figure 6 shows the characteristics of an individual lucas-novasensor model nph8-100-eh, 0 to 15psig, silicon pressure sensor. figure 7 shows the result of the compensated sensor/MAX1460 pair. using the compensated sensor/MAX1460 pair after calibration and removal from the test system, the MAX1460 and the sensor form a mated pair. the start pin can be connected to v dd or left unconnected if the sensor does not require a significant warm-up time. now operation is simple: just apply power and latch the paral- lel output d when eoc falls. temperature is digitized dur- ing the first half of t adc , so the MAX1460 provides a minimum sensor warm-up time of 35ms. using a 2mhz resonator, the conversion time t conv is nominally 67ms. if the repeat mode bit is set, conversions repeat at a rate of 15hz. if the sensor requires more than 35ms of warm-up time, the start pin may be used to initiate conversion (figure 5). if the repeat mode bit is set, start should remain high. if the repeat mode bit is reset, start may be used to externally control the conversion rate of the MAX1460. after the 12-bit parallel output d is latched, end the con- version by taking start low for at least one clock cycle. the output dac converts the parallel digital output into a serial bitstream on out. a simple external lowpass filter, using the MAX1460 op amp, converts the out bitstream into a ratiometric analog voltage (figure 8). the filter shown is an inverting configuration, but the gain and d off coefficients of the characteristic equa- tion can be adjusted to obtain either polarity. if the op amp is not used, it can be powered down using the op amp power-down bit in the configuration register. the MAX1460 requires a minimum of external compo- nents: one power-supply bypass capacitor (c1) from v dd to v ss . one 2mhz ceramic resonator (x 1 ). two 10k ? resistors for the agnd pin. if an analog output is desired, two 500k ? resistors and a 1? capacitor are needed for filtering. MAX1460 evaluation/ development kit the MAX1460 evaluation kit (ev kit) speeds the devel- opment of MAX1460-based transducer prototypes and test systems. first-time users of the MAX1460 are strongly encouraged to use this kit, which includes: 1) evaluation board, with a MAX1460 sample and a sili- con pressure sensor, ready for customer evaluation. 2) interface board that must be connected to a pc par- allel port. 3) MAX1460 communication/compensation software (windows compatible), which enables programming of the MAX1460 one module at a time. 4) detailed design/applications manual, developed for sensor-test engineers. the evaluation kit order number is MAX1460evkit. pin configuration n.c. amp- amp+ ampout n.c. n.c. out d5 d4 d3 d2 d1 n.c. n.c. agnd start i.c. d6 d7 d8 d9 d10 d11 n.c. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 48 47 46 45 44 43 42 41 40 39 38 37 36 35 34 33 32 31 30 29 28 27 26 25 MAX1460 n.c. v dd v ss cs1 cs2 n.c. n.c. sdio sdo reset eoc d0 inm test inp n.c. n.c. n.c. n.c. n.c. xin xout v dd v dd
MAX1460 low-power, 16-bit smart adc ______________________________________________________________________________________ 17 functional diagram 16-bit interface to all signals MAX1460 op amp instruction rom control logic eeprom configuration register correction coefficients registers temperature & sensor signal registers pga & coarse offset correction 12-bit digital output ref = v dd ref = v dd oscillator 16-bit digital signal processor (dsp) dac mux 16-bit adc temperature sensor cs1 cs2 start test reset sdio sdo eoc amp- amp+ ampout out d [11...0] v ss +5v +5v 10k 10k xin xout v dd 2mhz resonator x 1 c1 0.1 f sensor inp inm agnd chip information transistor count: 59,855 substrate connected to v ss
MAX1460 low-power, 16-bit smart adc 18 ______________________________________________________________________________________ notes
MAX1460 low-power, 16-bit smart adc ______________________________________________________________________________________ 19 notes
MAX1460 low-power, 16-bit smart adc maxim cannot assume responsibility for use of any circuitry other than circuitry entirely embodied in a maxim product. no circu it patent licenses are implied. maxim reserves the right to change the circuitry and specifications without notice at any time. 20 ____________________maxim integrated products, 120 san gabriel drive, sunnyvale, ca 94086 408-737-7600 1999 maxim integrated products printed usa is a registered trademark of maxim integrated products. notes

▲Up To Search▲

Price & Availability of MAX1460

	To Download MAX1460 Datasheet File
If you can't view the Datasheet, Please click here to try to view without PDF Reader .