ADC Clarification: TRIGSEL and ADCINTSOCSELn

James McLaughlin

Other Parts Discussed in Thread: TMS320F28069

Hello,

We use the TMS320F28069 and are having some strange behavior with the ADC after we started using the AdcOffsetSelfCal() function (as recommended by the data sheet and TRM).

Basically when we reconfigure back to our normal ADC configuration (after running AdcOffsetSelfCal()), it doesn't seem like it should run at all (we weren't setting ADCINTSOCSELn back to 0). This is based on the register description for ADCINTSOCSELn, which states that if it is set to anything other than 0, TRIGSEL is ignored (page 506-607 in the TRM). However if TRIGSEL is ignored, our ADC should not be sampling at all. Additionally, Figure 8-2 on page 486 shows that the TRIGEL input and the ADCINTSOCSELn input are OR'd together, which conflicts with the statements made on page 506, but would make sense why our ADC still runs.

I found this post that discusses it, but never got confirmation:
http://e2e.ti.com/support/microcontrollers/c2000/f/171/p/43844/170847.aspx#170847

So when TRIGSEL is configured, and ADCINTSOCSELn is configured for an ADCINTn response...are both inputs used, or is only the ADCINTn trigger?

Thanks.

- James

over 12 years ago

0 Devin Cottier over 12 years ago

TI__Guru 60865 points

Hi James,

Yes, it looks like there is a discrepancy in the documentation and it looks like this was never resolved. I am not sure which is correct.

Are you using software triggering or something else (e.g. ePWM triggering)? If you are using ePWM, most likely the diagram is correct. The only other way I could see samples still occurring is if the ping-pong sampling from the self-cal routine never stopped.

Note that you don't have to use the provided self-calibration routine to do self-calibration. You can use ePWM triggered (or other trigger sources) to collect the data to adjust the offset trim register if it is more convenient for your application.

0 James McLaughlin over 12 years ago in reply to Devin Cottier

Expert 1090 points

Thank you for the reply, Devin.

We normally use ePWM triggering, so the diagram being correct seems like the only possibility, but I would just like to confirm this. The ping-pong sampling definitely stops, since ADCINT2 is disabled by the provided calibration routine, and we don't enable it during our reconfiguration. We verified it is disabled during normal run-time.

Is there a way we can get a final word that the statements on page 508-609 are incorrect regarding the TRIGSEL?

- James

0 Devin Cottier over 12 years ago in reply to James McLaughlin

TI__Guru 60865 points

James,

Yeah, give me a few days to confirm this on the bench. I will also begin the process of preparing the document changes.

0 Devin Cottier over 11 years ago in reply to Devin Cottier

TI__Guru 60865 points

James,

No updates yet, but I haven't forgotten about this.

0 Devin Cottier over 11 years ago in reply to Devin Cottier

TI__Guru 60865 points

James,

I have confirmed that the diagram is correct, the bit description is incorrect. I will submit for the documentation to be updated.

0 James McLaughlin over 11 years ago in reply to Devin Cottier

Expert 1090 points

Thank you very much, Devin. I appreciate you looking into that for me.

- James

C2000™︎ microcontrollers

C2000 microcontrollers forum

ADC Clarification: TRIGSEL and ADCINTSOCSELn