• State Resolved
  • Locked Locked
  • Replies 6 replies
  • Subscribers 82 subscribers
  • Views 885 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.

F28069 Controlstick ADC Calibration

Andrew Green1
Intellectual 335 points
Other Parts Discussed in Thread: CONTROLSUITE, TMDS28069USB

Hi

I have a 28069 controlstick and when I run the ControlSuite example TriggeredADC I get A/D conversion values with errors (as some other posts have alluded to). I am not using the PWM DAC as input source as there is too much ripple on the signal, instead I am feeding in a dc voltage direct to the ADC. According to my Fluke 187 multimeter and the debugger I have the following measured input voltage and ADResult  (theoretically 0 to 4095)

11.7mV = 0

0.4995V = 587 to 589

0.9940V = 1193 to 1196

2.0095V =   2412 to 2414

2.6540V  = 3203 to 3206

These values do not correspond accurately enough to a 3.3V internal reference.

The TI example code supplied includes a call to the calibration procedure for the ADC, and I note during running the debugger shows:

ADCREFTRIM = 0x15AA

ADCOFFTRIM = 0x0000

I have seen exactly these same calibration values also quoted in a similar forum posting, so I am questioning if these are the factory calibration values or something else.

Given this is supplied TI code, what could be the reason?  Calibration not implemented while debugging? Calibration data false or corrupted? Defective chip? Conversion time not long enough?

I have no reason to believe my DMM is inaccurate. When using the external reference  the values are converted more accurately.

I would be grateful of some assistance. 

Andrew

  • 0
    TI__Guru 60865 points

    Andrew,

    Is your F28069 device symbolized as "TMS" or "TMX"?  If TMX, you may have statically trimmed devices (essentially a best guess trim global to all devices, not trim tailored to your specific device).  

     

  • 0 in reply to Devin Cottier
    TI__Guru 60865 points

    Searching around the forum a bit I am pretty sure you have TMX (experimental) silicon on your control card.  When did you buy this control card?

  • 0 in reply to Devin Cottier
    Intellectual 335 points

    Hi Devin

    Thanks for the quick reply. It does say TMX. We have a few lying around now, some bought last year and some bought this year, so I cannot be sure. I will see if we have one that says TMS.

    Thanks again.

    Andrew

  • 0 in reply to Andrew Green1
    TI__Guru 60865 points

    Andrew,

    Ok.  If you don't have a TMS kit, you have a few options:

    • Develop with the TMX kits as-is.  
      • The gain/offset error will exceed datasheet limits.
    • Swap out the parts for TMS parts.  
      •  This may be somewhat difficult depending on what kind of soldering setup you have.  I think you will want a heat gun to lift the part, which will be tricky, then hand-solder the new part down, which should be ok if you have a good lab setup.
      • There is a risk that you will damage the kit
      • You may be able to get free TMS parts by requesting samples from the TI website.
    • Try to trim the device yourself
      • If you look around the forums and ControlSUITE code, you should find the procedure for offset self-trim.  This does not require any external instruments.  This is something we recommend you do periodically in the end application in any case.
      • Gain trim is trickier.  You will need a multi-meter, current meter, and precision DC voltage source.  I can walk you through this if desired.  Alternatively you can just measure the gain error (which you have essentially done already) and then use the CPU to scale your conversions to cancel the error for development purposes. 
    • If you complain enough, you can probably get TI to send you a new kit
  • 0 in reply to Devin Cottier
    Intellectual 335 points

    Hi Devin

    The distributor we have been using sells only the TMDX28069USB and not the TMDS28069USB. I was unaware that there were actually two different types. 

    I have adjusted the offset and gain as suggested using:

    EALLOW;

    AdcRegs.ADCOFFTRIM.bit.OFFTRIM = 0x10;

    AdcRegs.ADCREFTRIM.bit.BG_FINE_TRIM = 0x02;

    EDIS;

    Now I get noise averaged values of

    0V = 0

    0.5005V = 622 to 623

    0.9992V = 1238 to  1239

    1.4997V = 1859

    1.9995V = 2479

    2.4997  = 3097 to 3098

    3.0003V  = 3718 to 3720

    Numbers look much better with respect to a 3.3V ref.

    Thanks

    Andrew

  • 0 in reply to Andrew Green1
    TI__Guru 60865 points

    Andrew,

    Glad you got things working.  It looks like TI is behind on getting out TMS kits, even though the particular part has been TMS for awhile.  This has been noted as an issue; sorry for the hassle.  Good luck developing!