DAQ card outputs wrong voltage / voltage clipped

[milifn]

Hi,

I am using both a PCI 6713 and PCI MIO 16E1 and have a very strange problem: When I try to output a voltage using LabView,I get instead of the voltage a clipped voltage with an unstable high phase. Basically everything is fine, as long as my voltage stays below about 0.3V. Everything is fine there, high level is stable, curve shape accurate. But as soon as it goes above 0.3 V, it is clipped and the voltage starts to drift. I attached a drawing of the problem to this post.

The whole system was running fine, but just a week ago the problem appeared. First it went away again, but now it reappeared. The first attempt to set a certain voltage might work, but already for the next pulse the problem kicks in. Now it won't even work correctly once. Restarting etc. didnt help.

The funny thing however is, the same PC has both devices listed above. I only and ever used the PCI6713. When the problem appeared, I tried to switch to the MIO 16E1 but it shows the same thing! Exactly the same problem there! I disconnected all leads to the cards and still it doesn't work. Anybody here with an advise?

Thanks!

Mike

[John_P1]

Hi Mike,

How are you measuring the output voltage? 

I'm taking a guess here, but it sounds like you might be using a 50 Ohm Scope to measure the signal.  The current drive per analog output line on both of the boards is only 5 mA.  5 mA across 50 Ohms would be 250 mV.

If this is the case, you should set your scope to high-impedance mode (if it supports that) so you don't draw too much current from the AO lines.

Best Regards,

[milifn · Accepted solution]

I was measuring the output voltage with a scope with 1 MOhm Input Impedance. But I found the problem by chance now! We got a whole batch of defective BNC cables! I already suspected the cable and tried various cables, pretty much the same result. Then somehow, as I wrote, the problem dissapeared once. And now I used a different type of cable to test the device and it worked! So I checked the other BNC cables, and some of them seem to have a shortage between Ground and Center, but in a way that it produces sort of a diode junction (probably due to an oxide layer somewhere). When I jiggled the cable suddenly I could measure a low impedance between ground and the center, but only for miliseconds, just like a bad contact.

Really weired! But thanks for the help!

[AnalogKid2DigitalMan]

Been there, done that. That's why I always purchase high quality coax cables and connectors from now on.

-AK2DM

[milifn]

Really hard to believe, the cables were new out of the box and certainly not the cheapest deal. Sad, I mean, why does one buy coax if not for high reliability?