TMS570LS0714: ADC self-test issue

Snaku Lee1

Prodigy 140 points

Part Number: TMS570LS0714
Other Parts Discussed in Thread: HALCOGEN

Hi,

I try to use the function 'SL_SelfTest_ADC()' of the SafeTI Diagnostic Library to test a ADC input pin,

but sometimes I get the result 'ADC_PIN_UNDETERMINED' because the ADC conversion value 'Vu' it lower than 'Vn' value,

which mean the ADC sample the input channel with SELF_TEST=1 and HILO=1, the result is lower than SELF_TEST=0.

Here is the ADC value when it perform the self testing.

I find there is a similar question in the forum, but no answer about it.

Regards,

Snaku

over 5 years ago

0 Sunil Oak over 5 years ago

TI__Mastermind 49120 points

Snaku,

I see that the HALCoGen project used to create the demo app uses the default sample time (2 ADCLK cycles, or 200ns) setting for the ADC conversions. This is not enough, depending on how much is the input impedance on the input channel being tested.

Please increase the sample time to be close to 1us and check the results.

Regards, Sunil

0 Snaku Lee1 over 5 years ago in reply to Sunil Oak

Prodigy 140 points

Yes, I notice the sample time is too short, I already change the sample time to 2us when I did the ADC self-test.

Regards,

Snaku

0 Snaku Lee1 over 5 years ago in reply to Sunil Oak

Prodigy 140 points

Yes, I notice the sample time is too short, I already change the sample time to 2us when I did the ADC self-test.

Regards,

Snaku

0 Sunil Oak over 5 years ago in reply to Snaku Lee1

TI__Mastermind 49120 points

I ran the ADC pin self-test on a setup here and see the results as expected. Please let me know what is connected to the ADC input channel 0 in your case. Even the Vd result in your case is too high. Are you running your test on a TI development kit?

0 Snaku Lee1 over 5 years ago in reply to Sunil Oak

Prodigy 140 points

Hi Sunil,

It is the circuit on the ADC channel 22 pin.

Regards,

Snaku

0 Sunil Oak over 5 years ago in reply to Snaku Lee1

TI__Mastermind 49120 points

Hi Snaku,

I cannot duplicate your observations on my setup. The self-test for the ADC input channels is fairly straightforward and is described in the reference manual.

Vd is the conversion result with the input connected along with a "pull down" to ADREFLO, so it is expected that Vd will be significantly lower than Vn (normal input). In your case, Vd is only slightly lower than Vn.

Similarly Vu is the conversion result with the input connected along with a "pull up" to ADREFHI, so it is expected that Vu is larger than Vn.

Can you extend the sample time to the max value configurable and see if the results change?

Regards, Sunil

0 Snaku Lee1 over 5 years ago in reply to Sunil Oak

Prodigy 140 points

Hi Sunil,

I change the ADC sample time to 20us, and the Vd value is better than 2us, but not close to 0.

Regards,

Snaku

0 Sunil Oak over 5 years ago in reply to Snaku Lee1

TI__Mastermind 49120 points

Hi Snaku,

With the above results you should get ADC_PIN_GOOD, right? Isn't that what you expected?

Regards, Sunil

0 Snaku Lee1 over 5 years ago in reply to Sunil Oak

Prodigy 140 points

Hi Sunil,

Yes, the result is ADC_PIN_GOOD. Thanks for your help.

Regards,

Snaku

Arm-based microcontrollers

Arm-based microcontrollers forum

TMS570LS0714: ADC self-test issue