"Saleae" clone logic analyzer and pulseview show inputs as HIGH when they aren't

[robduarte]

Hi,

I am trying to reverse engineer an RJ45 data cable on a laser cutter. I can see, with the logic analyzer and Pulseview, that there is really only one wire that does much. I could just stop there, since that wire tells me everything I need to continue with my project - it encodes the duty cycle of the laser tube.

The issue is that I can see, with a multimeter, that some of the other 7 wires are HIGH and some are LOW. But Pulseview shows them all as HIGH (except for the one with a changing signal). I tried swapping wires around and Pulseview always shows the data where it should, but the others remain HIGH on-screen (so it's not dead inputs on the LA). The voltages are essentially 3.3v and 0v so I wouldn't think it's a problem with the threshold. I have connected the ground pin from the LA to the circuit. 

What am I missing??

Thanks for any help!
Rob

[ledtester]

If the other wires aren't carrying any data it is possible they are floating in which case the logic analyzer could detect it as a high or low or even as an oscillating input.

Try attaching a 10K to 100K resistor from the wire to ground and see how the LA reads it. If the wire is floating then with the resistor it will consistently read it as low.

Btw, is ground one of the wires in the RJ45 cable? You mentioned that you see the issue with "7 other wires" but I would expect at most 6 wires in the RJ45 cable to be floating (8 minus the data and ground wires).

Finally, how is the computer/laptop connected to the LA powered? Are you using a laptop running on battery power or plugged into an AC adaptor?

[robduarte]

You are correct - one of the 8 wires is ground.

The computer is a desktop and it's plugged into the wall. This is a fancy facility with lots of outlets on independent circuits, so I guess it's possible that the outlet that the laser is plugged into is a different circuit than the computer. Do you think that matters? If there were a ground loop, would I be able to see the interesting signal with no troubles?

Re:floating that's an interesting thought.. I really think this company used an RJ45 connector/cable to connect the control board to the laser tube because it was convenient but only one wire (and ground) were actually needed. So maybe the other wires are floating because they are essentially ignoring them. I think it's also possible that I disconnected the laser tube's power at some point in my testing - maybe there's some active electronics in there that pull down the other pins?

Rob