(Please visit the site to view this file)(Please visit the site to view this file)Hi Kevin, Please see below for answers to the questions you need to know more about. 1. Could you compare the actual ADC results (ADCRESULTx register) in a good vs. bad case and provide them? This would be instead of looking at the current measurements recorded, which is calculated based on the read ADC values. The customer is working on finding a way to collect raw ADC through CCS, is this possible? If so, how? Also additional data was collected last night using scope to show measured ADC voltage reading @ ADC micro pin matching measured AC current, but reported reading by micro was incorrect (stopped updating). Phase C where is most susceptible is connected to ADCINA6 thought might indicate something to be last in the ADC queue. 2. Could you provide the ADC initialization code being used? Attached is source code for the ADC functions (including initialization). These were created by TI to their understanding and the customer did not modify them (they are from 2015), so not sure if they contain the errata fixes. 3. As mentioned, there are several ADC advisories in the device errata. Have these been understood and have the suggested workarounds been properly implemented? Currently, they are working through the ADC advisories, and have completed: Initial conversion, ADC result conversion when sampling ends on 14th cycle, ADC revision register limitation ADC can become non-responsive when ADCNONOVERLAP or RESET is called They are still working on the others, and would be happy to review them with us to make sure we implemented them correctly. 4. Could you provide more details on when this issue seems to happen? Is there something in the application that is systematically triggering the ADC results to be bad? Does it appear to happen after a certain length of operation, i.e. what is meant by eventually? Any details you can provide around these occurrences would be helpful in diagnosing the issue. The issue seems to happen after a low power sleep mode is induced and then the controller is woken back up. It is very repeatable on the motor dyne, issue happens after 30 sleep cycles. On the actual project it seems more random. Thanks, Barend
↧