MCC172 DC offset calibration

[keegan.brian]

 

In my project configuration I have 3 MCC172 boards on a Pi4 connected to two triaxial accelerometers. The triaxial  accelerometers I'm using are the 356A32 from PCB piezotronics which have a sensitivity of 100mV/g. I'm noticing that the DC offset on the 172 boards varies for board to board. In one of my stack ups that has three MCC172 boards the channels are all centered at zero. However, another stack up that has 3 different MCC172 boards the DC offset for the channels varies from -0.025g to 0.02g.

Is there any way to improve the DC offset at the output so that the channels are centered at 0g for the MCC172 boards?

Also is there anyway to show images besides a weblink?

[JRys]

IEPE Accelerometer voltage accuracy is especially great. You could be dealing with variations from one set to another. Swap the good ones with the bad and see if the problem follows the accelerometer. Otherwise, there is no DC adjustment other than doing it in software.

Best regards,

John

[keegan.brian]

Right now, I have 3 setups that have 3 172s each, so 9 in total. Using the same 2 accelerometers I cycled them through the 3 different stack up. I noticed that the DC offset would always change as the stack up changed. Then just to be thorough I cycled through with a different accelerometer pair through each of the 3 stack ups. When doing this the DC offset remained the same from HW stack up to stack up. Basically, the new accelerometer pair didn’t change the DC offset that I had measured before.

What about when the DC offset is out of spec of the datasheet value of 5.1mV? I had two measurements where the offset was measured at -6mV and -8.75mV as shown in the last image.
 

Edited February 14, 2023 by keegan.brian

[JRys]

Hello,

Could you retest but this time using a 2.7k or 3.3k ohm resistor instead of the accelerometer. The resistor simulates the bias voltage produced by a typical accelerometer. It should provide the best indication if the MCC 172 input is exceeding the 5.1mV offset spec with IEPE source current enabled. Use the continuous_scan example for the test and tell me the rms value after 30 seconds.

Best regards,

John

[keegan.brian]

Didn't have much luck with the resistor as the same problem was still showing up. Eventually I did figure out why I was getting the DC offset. We had built software on one raspberry pi that had 3 172 boards. We then created an image of that SD card and transferred that image the other three raspberry pi + 3x172s stack ups that we had. So the other three raspberry pi's had the 172 calibration from the first pi which was the source of the DC offset. Once we entered the sudo daqhats_read_eeproms on the other raspberry pi's the problem went away.