ADC strange behaviour
2595 Views - Created 05/02/2017
05/02/2017
Posted by:
MhicAoidh
OK, I'm using six ADC channels on an ExpanderPi to measure four voltage levels via resistor dividers and two currents using ACS712 sensors.Each channel works individually, both with a known source voltage and the real input. I've checked all 8 channels on the ADC and all works fine.However. For some reason, they don't work together in certain combinations that use adjacent pins. For example, inputs connected to:pins 1, 3, 5, 7 or 2, 4, 6, 8 workspins 1 & 2 doesn't work.pins 2 & 3 works.pins 2, 3 & 4 doesn't work.When it fails, the symptoms are consistent. For example, with an actual voltage of about 2.4V measured at the pin, pin X reports 0V, pinX+1 reports ~0.6V, and pinX+2 reports the correct value. Disconnect X+2, or move it to X+3, and everything reads correctly. What on earth is going on? Could this be an impedance issue?
05/02/2017
Posted by:
andrew
Location:
United Kingdom
The problem could be that the ADC is set in the wrong mode. The ADC used on the expander pi can be used as 8 single-ended inputs or differential pairs with pins 1 and 2, 3 and 4, 5 and 6 and 7 and 8 making up the four differential inputs. If the ADC has been set to be in differential mode then it would be trying to measure the difference between the input pins instead of measuring relative to ground which would explain the problem you are having.
Are you using our python library for reading from the ADC? If so then you should be able to set the inputs to single-ended mode by using read_adc_voltage(channel, 0). The second parameter on the read_adc_voltage function sets the mode for the inputs with 0 being single-ended and 1 being differential. Setting the parameter to 0 should hopefully solve the problem.
Are you using our python library for reading from the ADC? If so then you should be able to set the inputs to single-ended mode by using read_adc_voltage(channel, 0). The second parameter on the read_adc_voltage function sets the mode for the inputs with 0 being single-ended and 1 being differential. Setting the parameter to 0 should hopefully solve the problem.
06/02/2017
Posted by:
MhicAoidh
Yes that would certainly explain it. The version of the python(3) libraries I was using doesn't seem to support the set single/double-ended mode setting, so I downloaded the new version and now it bombs out when it tries to setup the GPIO ports. I get the following runtime error in file ABE_ExpanderPi.py, line 259, in set_pin_direction:AttributeError: 'IO' object has no attribute 'address'??
06/02/2017
Posted by:
andrew
Location:
United Kingdom
Sorry about the bugs. I have uploaded a new version of the library to GitHub which should be working. I tested the ADC inputs using the demo-adcread.py script and it appears to be reading the inputs correctly in single-ended mode.
07/02/2017
Posted by:
MhicAoidh
No, sorry, same error but in line 265 this time. Am I doing something wrong?
07/02/2017
Posted by:
andrew
Location:
United Kingdom
Sorry, there was still a bug with the IO functions. Can you try downloading the updated version
07/02/2017
Posted by:
MhicAoidh
Yes! It works, thank you. That's got to be the fastest tech support in history.
Note: documents in Portable Document Format (PDF) require Adobe Acrobat Reader 5.0 or higher to view.
Download Adobe Acrobat Reader or other PDF reading software for your computer or mobile device.