topic DSC56F84452 ADC signals conflict using internal gains amplifiers x2 and x4. in Digital Signal Controllers

DSC56F84452 ADC signals conflict using internal gains amplifiers x2 and x4.

mirkoruggeri — Fri, 13 Nov 2015 14:12:58 GMT

Hello,

I am usigng three ADC inputs to read one external anlaog signal with three different gains (see the schematic attached).

When the input with the x2 gain goes in saturation, the read of an other input (used to read the Motor Temperature) is changed.

It seems that the signals go in conflict, because also the input signal with gain x2 goes down to a level lower than the saturation.

The signal in input to the microcontroller is stable and it is not affected from any saturation or alteration, such as the x1 and x4 inputs are not altered.

In order to better explain the problem I attach the Schematic and the Freemaster Recorder image of the signals.

Thank you in advance for your attention.

Best Regards

Ruggeri Mirko

R & D Engineer

ELSY s.r.l. Sistemi Elettronici

Re: DSC56F84452 ADC signals conflict using internal gains amplifiers x2 and x4.

xiangjun_rong — Mon, 16 Nov 2015 07:37:29 GMT

Hi, Mirko,

I think it is a known bug for the 12 bits ADC module. It says:

4578: Possible incorrect readings on second ADC conversion when sampling two ADC

channels and first channel input voltage < VREFL or > VREFH

Description: When sampling two ADC channels that are time-consecutively listed, if the first channel input

voltage is less than VREFL or greater than VREFH, incorrect readings on the second

conversion may result.

Impact: See Description.

Workaround: For the ADC pins used for the conversions, ensure that the voltages are between the values

VREFL + 0.1 V and VREFH - 0.1 V.

For your case, when you use x4 amplifier, the AMP output is saturated, which will take effect on the next sampling.

Xiangjun Rong

Re: DSC56F84452 ADC signals conflict using internal gains amplifiers x2 and x4.

mirkoruggeri — Tue, 17 Nov 2015 12:58:29 GMT

Hi Xiangjun Rong,

many thanks for your answer.

I solved the problem by changing the gain dinamically.

In first, I read the input with no gain and if the the reading value is lower than a threshold (calculated as the actual value amplified by the gain wanted), and afterwards I prepare the channel gain setted for the next sample.

In other words, I solved the problem by removing the amplification before that the output of the internal gain amplifiers reaches the saturation, and reintroducing them when the input signal go down within a secure tresholds.

Thank You for your help.

Best Regards

Ruggeri Mirko

R & D Engineer

ELSY s.r.l. Sistemi Elettronici