PXI 6273 unreliable analog output

[TimGo]

Hello,

I'm a new memeber to this forum, and relatively new to LabView. As I have stumbled my way through learning LabView over the last months, I have used this forum a lot. So thanks to everyone for that.

I am using a 6723 and 6259 (in a PXI 1033 chassis) to do a number of different things in a pretty elaborate test setup that we've built. One output on the 6723 is used to set the frequency on a variable frequency oscillator (Si504). This is a digital communication line, but I am using an analog output from the 6723 because the comm rate is relatively fast, and I don't think there is any way to do this with th DO lines. For background info, the communication is done by varying between two different periods between square wave transition edges (5.5µs represents 0, and 13.75µs represents 1). To generate the signal, I just send out an array which is built up of the values 0 and 3.3, and set the period to 2.75µs. The waveform doesn't look pretty, and it seems to be pushing the 6723 as fast as it can go, but it should work (see below). 

However, the problem is that sometime the PXI6723 doesn't output the waveform (to channel A10, in this case) correctly. It seems to get caught up in a mode where it outputs garbage instead of the signal. When this happens, you can see it "trying" to output the signal, but the voltage doesn't really raise above half a volt or so, and the signal is not recognizable. It seems to happen about 10-15 times in a row, before stopping and going back to the correct output.

To output to the 6723, I just use DAQ assistant setup for channel A10, with volage set to -10...10V. 

Any ideas why I might be having this frustrating intermittent problem?

[crossrulz]

My initial thought is the current draw needed to drive the oscillator.  The 6723 can source up to 5mA.  For a digital line, I would think that would be enough.  But 0.5V/5mA would come out to 100 Ohms.  If the oscillator has a variable impedance, make sure it is set as high as it will go.

It might also be worth looking for a Schmitt Triggered digital buffer to attempt to clean up your edges and possibly increase your drive capability.

[TimGo]

Thanks for the response. 

It's not hooked up to the oscillator yet (I will do that tomorrow), so the signal is only loaded down by the oscilliscope right now. I will probably put a buffer between just to make the signal cleaner as you suggest before we start testing.