• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Subscribers 46 subscribers
  • Views 1302 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

ADS1258EVM-PDK mechanical (PCB) and electrical question, ADCPro

Pio Baettig
Prodigy 50 points
Other Parts Discussed in Thread: ADS1258EVM-PDK, ADCPRO, OPA365, ADS1258

Hi,

I am looking into acquiring an ADC board for data acquisition of 6 bridge (pressure) sensors with low output levels. After whittling down the options, I think the ADS1258EVM-PDK would fit the bill.

I plan on layouting a little board to plug in on top of the EVM to which the input connectors will be routed. For this I am looking for some mechanical data. I found the design data here e2e.ti.com/.../2680.ads1158-ads1258 but as I don't own protel, I have no way of extracting them. (I'd also be interested of the screw positions of the PDK motherboard, so I can design the whole thing into an enclosure (with fitting connectors to the outside world, differential amp and VRef if needed). (I am using eagle)

As the signals are small, my plan is to also design 2xOPA365 on this board to loop in between the mux out and the adc in. How stable is the included DAC as a voltage reference, or would it be better to use an additional voltage reference (LM4140ACM-1.0 1.024 V) in order to not have to add too high an amplification factor?
And lastly, is there a possibility to timestamp the acquisition values (individual lines) in the ADCPro software, so I can correlate these measurements with measurements from other sensors? (In the integrated sine generator demo mode, only the values were written to the log file, so I don't know how the program behaves with the real hardware.

I have not totally understood yet whether to use bipolar(supply) differential(signal input) or single supply differential input would be better. As the signal is differential anyways, I would not lose the sign bit anyways even if just driving with a single 5 V supply?


Thank you very much for your help.

Pio

  • 0
    TI__Mastermind 44690 points

    Hi Pio,

    Welcome to the TI E2E Forums!

    You're welcome to use the ADS1258EVM as you like for evaluation purposes and I can provide you with the EVM mechanical information, but please note that assembling the EVM into a finished product would violate TI's standard terms and conditions for evaluations modules. If the assembly is part of your evaluation, than likely this is not an issue.


    The EVM headers have the following dimensions:

    (Units are in inches. Note that the ADS1258EVM may not be exactly 3.2 x 1.7 inches in dimension; however, the relative header positions will be the same as indicated.)

    Note the latest MMB0 (Rev D.) does not have any screw-holes:

    (It does have four adhesive bumpers on the back side, for which only the through hole components would limit the placement of these. The MMB0 is about 4.75 x 3.25 inches.)

    Regarding your other questions:

    Pio Baettig said:
    As the signals are small, my plan is to also design 2xOPA365 on this board to loop in between the mux out and the adc in. How stable is the included DAC as a voltage reference, or would it be better to use an additional voltage reference (LM4140ACM-1.0 1.024 V) in order to not have to add too high an amplification factor?

    The included DAC uses an on-board REF1004-2.5 as its voltage reference. This reference is a bit older and I know that newer references (such as the REF50xx series) will certainly be more stable over temperature.

    The reference voltage value is not too important with delta-sigma ADC's. Typically, the LSB size on a 24-bit delta size ADC (i.e. the quantization noise) is much smaller than the thermal noise (on lower resolution SAR ADCs this is not always the case). Therefore, using a smaller reference voltage (to make the LSB size smaller) will not have much effect because the thermal noise will still limit the resolution. In general, I'd recommend using a larger reference voltage so that you have a larger full-scale input range (FSR).

     

    Pio Baettig said:
    And lastly, is there a possibility to timestamp the acquisition values (individual lines) in the ADCPro software, so I can correlate these measurements with measurements from other sensors? (In the integrated sine generator demo mode, only the values were written to the log file, so I don't know how the program behaves with the real hardware.

     

    You would have to make significant modifications to allow for this. Note that the data is collected in blocks, buffered, and then transmitted to the PC (data is not streaming). It probably wouldn't be enough to only modify ADCPro; I think this would require changes to the MMB0's embedded code. These code modifications would not be something for which we could provide support. For this reason, I might suggest beginning with a new PCB design, using an MCU you are already familiar with programming.

     

    Pio Baettig said:
    I have not totally understood yet whether to use bipolar(supply) differential(signal input) or single supply differential input would be better. As the signal is differential anyways, I would not lose the sign bit anyways even if just driving with a single 5 V supply?

    Correct, your input signal will be differential either way and the data will always have a sign bit...the only difference is the signal's common mode voltage. If you want the input common-mode voltage to be near 0V, you would need to use a bipolar supply. A bipolar supply requires that you generate a negative supply voltage for the ADC, while a single-supply configuration may require you to level-shift the input common-mode voltage (often times the voltage reference can help provide the level-shift voltage).

     

    I hope that helps!

    Best Regards,
    Chris

  • 0 in reply to Christopher Hall
    Prodigy 50 points

    Christopher Hall said:

    Hi Pio,

    Welcome to the TI E2E Forums!

    You're welcome to use the ADS1258EVM as you like for evaluation purposes and I can provide you with the EVM mechanical information, but please note that assembling the EVM into a finished product would violate TI's standard terms and conditions for evaluations modules. If the assembly is part of your evaluation, than likely this is not an issue.

    Hi Christopher,


    thank you very much. Yes, the assembly is part of the evaluation, it will go into an aluminium box to reduce interferences and there will be a pcb on top to mount the connectors neatly and put the amplifier (OPA365 from ADS1258 data sheet) in between MUX out and ADC in.

    Christopher Hall said:


    The EVM headers have the following dimensions:

    (Units are in inches. Note that the ADS1258EVM may not be exactly 3.2 x 1.7 inches in dimension; however, the relative header positions will be the same as indicated.)

    The second analog port seems centered to the left one in y direction and offset to the right by about 500 mils?

    Christopher Hall said:

    The included DAC uses an on-board REF1004-2.5 as its voltage reference. This reference is a bit older and I know that newer references (such as the REF50xx series) will certainly be more stable over temperature.

    The reference voltage value is not too important with delta-sigma ADC's. Typically, the LSB size on a 24-bit delta size ADC (i.e. the quantization noise) is much smaller than the thermal noise (on lower resolution SAR ADCs this is not always the case). Therefore, using a smaller reference voltage (to make the LSB size smaller) will not have much effect because the thermal noise will still limit the resolution. In general, I'd recommend using a larger reference voltage so that you have a larger full-scale input range (FSR).

     

    Thank you very much, so amplification is the way to go.

    Christopher Hall said:

    You would have to make significant modifications to allow for this. Note that the data is collected in blocks, buffered, and then transmitted to the PC (data is not streaming). It probably wouldn't be enough to only modify ADCPro; I think this would require changes to the MMB0's embedded code. These code modifications would not be something for which we could provide support. For this reason, I might suggest beginning with a new PCB design, using an MCU you are already familiar with programming.

     

    So there is no timing information provided (or numbering of the lines which can be traced back to time?) ADCPro seems to be able to do FFT as I saw on some youtube videos, so is the timing just used/transmitted internally and not written to a file?

    Christopher Hall said:

    Correct, your input signal will be differential either way and the data will always have a sign bit...the only difference is the signal's common mode voltage. If you want the input common-mode voltage to be near 0V, you would need to use a bipolar supply. A bipolar supply requires that you generate a negative supply voltage for the ADC, while a single-supply configuration may require you to level-shift the input common-mode voltage (often times the voltage reference can help provide the level-shift voltage).

    Thank you very much for the explanations and for all your help. So using a wheatstone bridge running of a 5 V supply my common mode voltage will be 2.5 V and I can just use the unipolar supply variant with the 5V VCC as reference (which has the advantage that my measurement will be ratiometric.

    Thank you very much for your help.

    Best regards,

    Pio Baettig

  • 0 in reply to Pio Baettig
    TI__Mastermind 44690 points
    Hi Pio,

    I should have also mentioned that the 20-pin connector on the top of the EVM is a Samtec TSM-110-01-T-DV. Its mating pair is the Samtec SSW-110-F-D-VS-K. So the pins have 100 mil pitch and the diagram above indicates the center of the connector. Therefore, the "Serial" connector need to be 2,700 mils to the right of the "Analog" connector (when looking from the top), for example.

    (I believe the power connector is a Samtec TSM-105-01-L-DV-P, which mates to Samtec SSW-105-22-F-D-VS-K).

    Regarding timing information in ADCPro...
    ADCPro only knows the approximate data rate of the ADC and it plots the FFT based on this information. However, it does not have any time stamping information to tell when the sample was taken, not even with respect to the previous sample.

    A wheatstone bridge should work well with a single supply, and yes, measuring it ratiometrically is the way to go!

    Let me know if you have any additional questions!

    Best Regards,
    Chris