Strange problem with analog output on PCI 6251 and BNC 2110

[johnsold]

Google "isolation amplifier".  Wikipedia has an explanation. TI, Analog Devices, and others make devices.  Allied, Digi-Key, Mouser, and other distributors sell them.  If you want an instrument rather than a component, look at the NI SCXI-1125.  All of these are found on the first page of the search results.

Lynn

[hmalissa · Accepted solution]

Thank you very much.

I solved the problem using a simple op amp based subtractor.

[mayankb]

Hi,

I am experiencing a similar issue with Analog Outputs on my NI-PCIe 6361 board (with BNC 2110). I am working on an AFM device and I am trying to do system identification using the NI board. I am able to receive signals properly. However, there is a problem with Analog Output (both A0 and A1).

I used oscilloscope to monitor the AO0/AO1 signal (with AFM unit not in-loop) and it works fine. However, the moment I connect the signal to the AFM unit, the signal becomes zero (as seen on oscilloscope) and the drive current is constant 21m. However, when I use an external function generator to generate the same sine wave with AFM unit in loop, the drive current varies sinusoidally with 6m pk-pk current as expected and I was able to actuate the device properly. Just wondering if you could advise me on that.

I have also tried to flip the FS/GS switches on the BNC 2110, but that doesn't make any difference. The AFM control signal is a single-ended input in this case.

[johnsold]

It is probably better to start a new thread and link to this one to indicate that you saw this discussion. The reason is that many people pay less attention to threads marked as Solved.

It certainly sounds like a similar situation. Try measuring the voltage between the grounds on the BNC 2110 and the AFM (while both are powered on but not connected). Use a multimeter or a battery powered oscilloscope so the measuring instrumetn does not introduce additional ground connections. I expect that you will find enough difference to cause the problem.

Lynn

[mayankb]

Hi Lynn,

Thanks. Actually the oscilloscope reads the voltages right. All the devices - the PCIe card, oscilloscope and AFM unit have the same grounds. Let me explain you the scenario more precisely.

I have an AFM unit and the input to AFM unit is a voltage signal from -10V to 10V. I think the input of the AFM unit is internally pulled low (as it reads -10V with the multimeter when turned on). Before feeding in the voltage to AFM from the PCIe board, I first scoped the signal on an oscilloscope and it read the correct signal. However, the moment I feed the signal to the AFM unit, the signal drops down to zero and the AFM piezos do not get actuated.

Then, I used a function generator (again sharing the same ground) with 50ohms output resistance and generated the same sinusoidal signal and the AFM unit works fine.

I have tried using a voltage follower for the input signal to the AFM unit in case the AFM unit was trying to draw an excess current from the PCIe card but that didn't help either.

[dsFlux]

Hi mayankb,

What is the input impedance of the AFM unit? The output impedance of the PCIe-6361 is 0.2 ohms. Has this setup worked correctly in the past?

[mayankb]

Hi David,

The setup has worked in the past. Actually, today only I was able to make the device work.

As Lynn had suggested that "overloading" might be an issue and a normal op-amp wouldn't help. So, I used a dsp board with higher output current rating and simply implemented a voltage follower using that and now the AFM setup works fine.

Mayank