TMS320F2800156-Q1: Differential sensor into single ended ADC inputs without differential/single ended amplifier circuit

Part Number: TMS320F2800156-Q1
Other Parts Discussed in Thread: C2000WARE

Tool/software:

I am trying to sample a differential input using the synchronous sampling method mentioned in the specification.  It seems to work but the values are way off.  I think the issue is the way the differential singles are terminated.  We have 2 options on our circuit.  One has parallel RC circuit from the positive input to the negative input, this is the one I have been testing with. The other as parallel TC circuit to ground for both the positive and negative inputs.  There is also a 49ohm resistor in series before the parallel RC circuits on both singles.  I just wanted to get some feedback on what you would recommend be used. 

  • Hi David,

    I am looking into this thread. I will reply back shortly.

    Thanks,

    Susmitha

  • Hi David,

    Are you running any example from C2000ware? Did you change ADC config to Differential input?

    Thanks,
    Susmitha

  • No, I am not using an example though I did reference them to confirm what I was doing.  I don't think there is a way to change the ADC's to differential on the 156 controllers since they only has single ended inputs.  I have 2 sensors with differential voltage outputs.  I have the positive input of one going to A0 and the negative input going to A1.  A0 is sampled on ADCA and the A1 is sampled by ADCC.  The other sensor has it positive input going to A2 and the negative input goes to A3/C5 and is sampled the same way as the other but at a different time.  I am using epwm1 to setup up the SOC triggers

    1. 50MHz clock
    2. 2500 base period
    3. SOCA is at .25 base period (for sensor 1)
    4. SOCB is at .75 base period (for sensor 2)
    5. Internal reference 2.5 volts
    6. ADC EOC interrupt is at end of conversion
    7.  I measure ~1Vdc at the pin A3/C5 and ADCC shows 0x31 (should be 0x746)
    8. I measure ~2Vdc at the A2 pin and ADCA shows 0x47 (should be 0xCCC)

    Neither are what they should be I am using an acqps of 31 but I have tried higher and lower values with no success.  I thought it might be the memory crosstalk so I added reads to A13 by both ADCA and ADCC.  I think that is the REFLO.  Any ideas would be appreciated.  Thanks

  • Hi David,

    I am reviewing this thread. I will get back to you shortly.

    Thanks,

    Susmitha

  • Hi David,

    Thats right. ADC channels in this device are single ended. Below is the image from TRM.

    Since you mentioned, you are using A0, A1 for first signal and A2 A3 for second signal. Please try A0 and A1 assigned to A module, A2 A3 assigned to C module...so you will be using C9 and C5...

    Can you please try above and let me know!

    Thanks,

    Susmitha

  • Thanks for the help.  This last change gave me a clue.  turns out I had missed setting up the AGPIO for A2 and A3.  Getting good readings now.

  • I do have a suggestion for you though.  On the pin diagrams could you put the default settings first?  For example for A2 on the pin diagram it says A2/C9,GPIO224 but the default setting is for GPIO224.  So if you made it so the pin diagram said GPIO224, A2/C9 then it would mean you would have to pick a new setting for it if you want analog.  This is done on other signals like A20/C20,GPIO12.  The default is analog so you know you have to use a new setting to use GPIO12.  Thanks for the help

  • Sure David. I will file Jira suggesting this when revising TRM.