Noisy digital inputs only when laptop plugged in

[pallen]

I've had problems with ground loops and laptop power supplies in the past.  But this is the first time I've had it so bad that it renders digital inputs that are both tied to ground and software debounced, impossible to use. 

This is NOT NI hardware.  It's a Data Translation DT9816, a low-budget multifunction USB DAQ.   The laptop with the problem is a refurbished Dell, that has a very obvious non-Dell power supply that throws up the BIOS warning and scales back the processor.  (Danger Will Robinson!)

The problem with the digital inputs only shows up on the Dell when it's plugged in to AC power.  On every other machine, and with the Dell unplugged, the digital inputs do float.  But the ones I've tied to ground are rock solid and work fine. 

What surprises me most is that noise is so bad that even tied to ground with a half second debounce, I'm still getting false triggers.  I haven't scoped the power supply, but considering it actually sounds noisy, I suspect it looks pretty bad. 

I figure I'll try a powered USB hub.  If that doesn't work then I suspect the customer will have to replace the power supply on the laptop. 

I was surprised though.  I've used a lot of USB DAQ hardware, and this is the first time I've ever seen an issue like this.  The NI-USB 6002 I tested on this same laptop did not seem to display the same problems.  Although admittedly that was different software and the testing was brief.

I'm curious if any other developers have found something simlar in the past. 

[_Bryan]

I had the same problem years ago in our emissions lab, I am pretty sure it was a dell laptop with USB DAQ.  I have no idea on any part/model numbers.  Running on the battery everything was fine and having the AC plugged in I had a fair bit of noise.  I kind of remember getting a different power brick from the IT guys and everything went back to normal.

I also remember way back when I had to share my office/lab having some power issues with a desktop.  I was putting together something with temperature and voltage and everything was working fine.  Then all the signals started getting really noisy.  It came down to a guy plugging in his phone charger on the other side of the room.

[RTSLVU]

The noise is probably on the ground and it's making your ground reference float around.

That's why you still get false triggering on a grounded DIO pin.

[pallen · Accepted solution]

I can confirm that using an externally powered USB hub eliminated the noise problem in the original post. 

Interesting stuff.  I'd not seen this before.  Now I wish I still had the NI USB DAQ I had here last week so I could test that. 

Looks like I'll be using a powered hub on this project.  But I think I will still recommend replacing that power supply.  That much noise coming from the supply probably isn't good for anything plugged into it...or even sitting close to it.