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.

TMS570LC4357: AD1[0] and AD7[7] consistently read lower values

Part Number: TMS570LC4357

Hello,

I've got a bunch of TMS570 evaluation boards and ADC1 pins 0 and 7 give consistently lower readings than the other pins.
For example I've connected all the pins to the same signal via breadboard bus and drive them via a trim pot. Then I just saturate the ADC's such that they all read 4095. But pins 0 and 1 show lower values:

one board:
AD1[0]: 395x
AD1[7]: 3960

another board:
AD1[0]: 398x-4000
AD1[7]: 4045

I can see these reduced values when I let all the inputs float (they usually just pick up 60Hz EMI).

One experiment I did was to ground pin 1. That didn't help pin 0 and made pin 2 start reading lower than expected values as well.

I'm just using ADC1 and triggering it in SW at 50Hz using adcStartConversion(). All the ADC timings are the default ones used by HCG.

Any help would be appreciated.

  • Ivan,

    Please increase the sample time for the conversions. This should help reduce the channel-to-channel impact of the conversions. See this application note to better understand the dependencies of external circuit impedance on sampling time:
    www.ti.com/.../spna118b.pdf

    Regards,
    Sunil
  • Thanks for the suggestion.

    I changed the sample time (and only the sample time, not the cycle time)

       ADC1>ADC1Group1>cycleTime 106.67

                      >SampleTime 1200

                      >tSample(ns): 1280.04

    from 300 to 1200 ns. This improved things a bit but AD1[0] and AD1[7] still stand out. See before/after data below. AD1[0] is blue and AD1[7] is purple.
    It occurs to me that signals 0 and 7 are at ends of an 8-bit "bus". Is there any significance to that?
    I'll also try to make this measurement with conditioned inputs (as opposed to launchpad+breadboard) and see if things improve and let you know.

  • Hello,

    I checked to ensure that there is nothing specific about these channel for them to convert lower than the other channels. Did you get a chance to "clean up" the inputs to these channels and perform the measurements again?

    Regards,
    Sunil
  • No, the "cleaned up" inputs are part of a circuit and from diverse sources so it's not easy to feed them all the same voltage.
    I'll close this as resolved since I don't know when I'll have the opportunity to verify this.