LPC MicrocontrollersのトピックLPC54616J512BD208 ADC conversion reading drooping after 5-8 minutes

LPC54616J512BD208 ADC conversion reading drooping after 5-8 minutes

Heiser — Thu, 12 Aug 2021 08:07:24 GMT

Hello,

We are seeing an issue on our microcontroller where after around 5-8 minutes the ADC reading is dropping ~100 ADC bits on 3 ADC channels being converted, when the input being sampled on the channel's is not fluctuating.

We are using the conversion sequence A to get the ADC results from the 3 channels every 1ms. At the end of the sequence interrupt, the software trigger is being used to loop back through and get the conversions again.

The ADC reference voltage is tied to the 3.3V supply - this voltage is also not changing.

I have attached the ADC config routine and sequence handler plus the raw ADC readings captured over an hour.

Any help would be greatly appreciated.

Re: LPC54616J512BD208 ADC conversion reading drooping after 5-8 minutes

Alexis_A — Wed, 18 Aug 2021 02:52:08 GMT

Hello @Heiser,

Looking at the configurations this looks good but this behavior seems strange. I have some questions about this:

Could you let me know which SDK version are you using?

Is there any event that could trigger this behavior in your application? Some signal that could introduce noise to the ADC capture?

Is this present using any channel?

As reference, in the following document there's some recommendations to improve the ADC measurement.

Best Regards

Alexis Andalon

Re: LPC54616J512BD208 ADC conversion reading drooping after 5-8 minutes

Heiser — Wed, 18 Aug 2021 07:48:09 GMT

Hi @Alexis_A ,

I think we have found the issue - one of the other ADC input channels had a negative voltage that would slowly drift down from 0V to ~-0.7V. When the channel measured around -0.3V this would then affect the ADC readings. It was repeatable with an external power supply.

Best Regards,

Heiser

Re: LPC54616J512BD208 ADC conversion reading drooping after 5-8 minutes

Alexis_A — Wed, 18 Aug 2021 15:21:31 GMT

Hello @Heiser,

I'm glad that you could find the issue.

Only as a suggestion, to ensure the correct functionality of the MCU, the tolerance for negative voltage is limited to -0.5V.

Best Regards,

Alexis Andalon