Hi all,
customer Diehl are asking a very specific question about the influence of the settings of ADCx_GC1 on the conversion result in case of an error.
This may need help from design to get an answer. Its a question of the customer's FMEA exercise.
Imagine the following setup:
ANA0 is configured to be amplified by 4 (ADCx_GC1_GAIN0 = 01)
ANA1 is configured to be amplified by 1 (ADCx_GC1_GAIN1 = 00)
The question: What happens to the conversion results in RSLT0 and RSLT1 if the analog input multiplexer of ANA is set to channel ANA0 due to an
internal short to 0 of the lowest bit of the multiplexer adressing signal? Meaning that channel ANA0 is selected due to the error although software initiated a conversion of channel ANA1.
Will the PGA amplify the input signal by 4 or not?
Thanks
Guenter
Solved! Go to Solution.
The signal scaling function, which incorporates the programmable gain feature, is independent of the signal select function. So, at the ADC level, the signal provided by the signal select function will be scaled in accordance with the signal scaling setting irrespective of if the signal was correctly selected or not.
Or, in orther words:
If the error causes ANA0 to be selected instead of ANA1, then the result would be ANA0 scaled by the gain control programmed for ANA1. This means that RSLT0 would be correct and RSLT1 would be ANA0 scaled by 1 (or RSLT0 divided by 4 in this case). There is only one PGA and it is after the analog muxing so it would be independent of the error in the muxing.
The signal scaling function, which incorporates the programmable gain feature, is independent of the signal select function. So, at the ADC level, the signal provided by the signal select function will be scaled in accordance with the signal scaling setting irrespective of if the signal was correctly selected or not.
Or, in orther words:
If the error causes ANA0 to be selected instead of ANA1, then the result would be ANA0 scaled by the gain control programmed for ANA1. This means that RSLT0 would be correct and RSLT1 would be ANA0 scaled by 1 (or RSLT0 divided by 4 in this case). There is only one PGA and it is after the analog muxing so it would be independent of the error in the muxing.
Design has been consulted. The answer will be posted soon.
Hi John,
have you got any feedback from design on that? Customer is pushing for an answer.
I posted the question on 11/5.
I reposted it today to engineering. I'll let you know any response to this hypothetical situation.