• State Resolved
  • Locked Locked
  • Replies 5 replies
  • Subscribers 46 subscribers
  • Views 454 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

AMC7812B: CH0 (pin33) signal inluence others channels

Nicola Martinelli
Prodigy 165 points
Part Number: AMC7812B

Hi all.
I'm using a AMC7812B to read 8 channels (CH0 to CH7).
The eight channels are configured in the same way: I must read a DC voltage (max is 24V) and I'm using a "voltage divider" -> R1=40.2k, R2=10k. Between the voltage divider and the ADC channel there is a 10nF capacitor.
I'm using the "direct mode" to read the channels.

I see a very strange problem: the signals in the ch0 influence the measure in the others channels and the measure in the ch0 is always accurate.
Besides, greater is the difference between the ch0-chx signals, worst is the measure in the others chx channels.
For example:
ch0-ch7=20.00V -> all measures are very accurate because ch0=19.97V, ch1-ch7=19.99V
ch0=0V ch1-ch7=20V ->  ch0=0V, ch1-ch7=20.11 to 20.12V
ch0=4V ch1-ch7=20V ->  ch0=3.997V, ch1-ch7=20.08 to 20.1V
ch0=12V ch1-ch7=20V ->  ch0=11.98V, ch1-ch7=20.04 to 20.05V

Any suggestions?
Thank you.
Nicola

  • 0
    TI__Guru 53773 points
    Hi Nicola,

    Can you share a schematic? What do you have on the ADC-REF-IN/CMP pin? In addition, if you disable channel 0 (ADC Channel Register 0), does the output of the other channels still get affected?

    Thanks,
    Paul
  • 0 in reply to Paul_Frost
    Prodigy 165 points

    Hi Paul

    Here attached you can see the schematic.

    AD gain is 2xVREF=5V

    I read one channel at a time and so when I read CHx the others are disabled.

    Let me know if you need other info.

    Thank you, Nicola 

  • 0 in reply to Nicola Martinelli
    TI__Guru 53773 points
    Hi Nicola,

    It may be interesting to look at the input pin with an oscilloscope during the sample event. The AMC has a SAR converter in it, so when it samples it draws a slight current from the input pin. Having the 10nF cap should help with this, but it may be interesting to see if the high impedance of the resistor divider is causing the voltage to drop during the conversion.

    To clarify, you are writing to ADC Channel register 1 and 2 (0x50 and 0x51) between each read to disable the input channels?

    Thanks,
    Paul
  • 0 in reply to Paul_Frost
    TI__Guru 53773 points
    Any updates?
  • 0 in reply to Paul_Frost
    Prodigy 165 points

    Hi Paul
    When I read a channel I disabled all the others channels.

    With the oscilloscope I saw the signals on the 10nF cap. The signals are different when I check ch0 and the others channels.
    It would seem that when I read ch1-ch7, is also connected ch0.

    Anyway, I changed the reading procedure and I solved the problem. In the old procedure, the reading was made with the average of 8 samples and to read the samples I switched 8 times.
    In the new procedure, I switch only one time and I bring 9 samples, I discard the first (because it is inaccurate) and the reading is the average of 8 samples. In this way all the measures are very accurate.
    Thank you.
    Nicola