I am attempting to stream 32 differential channels using a T7 with a Mux80 multiplexer using Python-LJM on a laptop running ubuntu. For some reason, the readings that I'm getting don't seem to make any sense. If I put a one volt input into the first channel (Differential reading of AIN48/AIN56) I get a sensible answer, but if I put the same 1V signal into the other inputs the readings are incorrect.
As a sanity check, I wrote another script (TestLabjack.py) which takes a single sample of all of my channels using eReadNames and those values do agree with what I expect for all channels. I'm new to streaming, so perhaps I've made a mistake in initiating my stream? Please see the attached file StreamLabjack.py. I am essentially using the example script "stream_basic.py" except I modified the list of channels and set things up for differential input.
.png%3Fitok=PHGBxf6M "EEV Blog Reviews LabJack T7-Pro & Software Tools")
.jpg%3Fitok=B1YcuDFK)
.png%3Fitok=Ybt-OlgU "LabJack + JEDI One Can Create a Powerful Distributed DAQ Ecosystem")
First, try the latest beta firmware (v1.0208) and LJM driver version, and see if that helps. Use our Kipling application to update the firmware.
If that doesn't help, make sure your have secure connections to the Mux80 and T7. For a simple test, connect the negative channels to GND, and then connect the positive channels to GND or VS line so you can see them read 0 or 5 volts. In my simple test on positive channels 48 and 49, readings were 0 and 4.88 V as expected. My simple test was using your StreamLabJack.py code which looks fine codewise. Let us know your results and we can continue troubleshooting further.
I'm still having some issues. The lower value channels (AIN 48-AIN 55 and AIN 64-71) are giving reasonable values. The higher value channels (AIN80-AIN87 and AIN96-AIN103) sometimes work and other times they run away to +/- 10V. The odd thing is that they always work when I am not in stream mode. The output for TestLabjack.py is always correct. This makes me think that its not a connection issue. For my test I had Vs attached to the negative channel and GND attached to the positive channel. When things were working, the reading was -4.86 Volts.
One additional detail is that I have 700 Hz passive low pass anti-aliasing filters on the inputs (I drew a schematic for the first differential channel AIN48/AIN56 which is attached). I don't have ADC buffers after the filters. When I run either script I do see some residual voltage after unplugging the connections which goes away if I stick a 50-Ohm terminator on the BNC input to discharge the capacitor. Not sure why this would be causing issues with streaming, but I figured that I'd include this detail. I tried using the terminator when the stream values ran away to +/- 10 V thinking that perhaps the capacitor was getting charged for some reason. This didn't help. The readings remained railed.
A couple thoughts:
1. Too much source impedance could cause strange behavior like you are seeing, but looks like you have 475 ohms and anything <=1000 ohms requires no special thought to work to spec.
2. I see in your diagram that there is no ground connection. If your signal is otherwise referred to T7-GND somehow that is great, but if not your differential pair could be totally floating. Try adding a large resistor (perhaps 100k or even 1M) from the negative AIN to GND.
https://labjack.com/support/app-notes/differential-analog-inputs