scxi 1124, output voltages low on all channels

[NJank]

Have an SCXI-1000 chassis with a SCXI-1600 USB digitizer module and an SCXI-1124 Digital-to-Analog Converter. All 6 channels of the 1124 output incorrect voltages. in -10 to +10V mode, it puts out about -1.85V to 1.76V when sending 0 and 4095, respectively. In 0 to +1V mode, it puts out 0 and 0.188V. similar discrepancy for other modes. Again, it'son all 6 channels.

I am running Labview 7.1.1, NI-DAQmx 8.9, and have the Traditional NI-DAQ 7.4.4 as well.  My first thought is that somehow the internal calibration constants werecleared orsomething, so I put together a VI with the DAQmx External Calibration functions like what is shown in http://zone.ni.com/devzone/cda/tut/p/id/4845 under Analog Output.  However, if I put in the values I'm actually measuring I get an Error 200546 for the ones that don't match. Based on the 1124 user manual and the limits it states on calibration constant range, I assume that the constants it would need to correct for the values I'm getting would be too large (i.e., it's probably not a cal constant issue).  

I put the 1124 in another SCXI1100 chassis and got the same behavior, so it's not a chassis fuse (I also checked the ones behind the fans) or the 1600. I've been unable to read the current cal constants since I can't seem to set up teh 1124 for reading using the legacy drivers. not sure why. My guess is that I can't use the 1600 as ago-between since it's not listed as supported by Traditional NI-DAQ drivers, and DAQmx claims to still be unable to retrieve the constants, just to set new ones. 

I was tempted to believe this was a driver or software issue, but I'm starting to think otherwise. The chassis and board pass the self-test in MAX. Thoughts on howI can verify whether ornot the 1124 just went bad or if it's something user reparable? (some undocumented fuse...? )

[The_Red_Barron]

Howdy NJank!

When the chassis and board pass a self-test in MAX that is just verifying that the driver is able to communicate with the board.  Based on the tests you've described it sounds like the analog circuitry on the SCXI-1124 may be broken, though it's strange that it happened on all six channels.

What method are you using to output the voltage for testing (LabVIEW code, MAX test panel, etc...)?  Please use a MAX test panel if you are not already.

What equipment are you using to measure the output voltages and how are you connecting it?

Have you tried restoring the device to factory settings?

Regards,

[NJank]

Method: mainly the MAX test panel, although I've used a couple other VI's and gotten the same results.

Equipment: in voltage mode I've generally just been measuring the open circuit voltage across the Vout and gnd terminals using a handheld multimeter. multiple meters have indicated the same. I also put a series resistance across the terminals (~1kohm) and measured the same voltage behavior.

I'm pretty sure I tried the reset VI, but can't recall the specific result (i tried a number of things).  I think it being a Traditional NI-DAQ VI, I probably couldn't get it working. Haven't been able to get any of the legacy stuff to address the card, and I suspected it was because when trying to set up the chassis in MAX using the Traditional drivers, it won't let me add a 1600 as my slot 1 DAQ. Adding just the chassis and 1124 under Traditional in MAX, most of the legacy VI's give me some AO error indicating the device is not present, non-compatible, not an NI istrument, or not set up. (some catch-all statement like that).

[ManLongAtNI]

Hello NJank,

You are correct, you wll be unable to use the Traditional DAQ driver when using the 1600 as the controlling device. With that in mind, using the Traditional DAQ VIs with this device will not function properly. You will likely need to set new ones using the DAQmx driver.

[NJank]

thanks for confirming that. Thought for a while it was something I was doing wrong. As I mentioned above, I set up the DAQmx VI's to set new calibration constants, but I get an error 200546 when I actually try to commit the value. (no error if I pick cancel, or If I tell it the voltage I'm getting is the one I asked for. I.e., in 0 to 10V mode, I do get 0V correctly. So running the commit function for that point completes successfully. It's when I tell it I'm only getting 1.8V but should be getting 10V (i.e., for a 4095 input) that I get the 200546 error. Checking against the 1124 user manual, I assume that error is from exceeding the allowable calibration constant storage. I.e.,  0<BL<255 and 3840<BH<4095.  Using the formulas from the page 5-2 of the manual, in +/-10V mode I'd need BL= -9245 and BH=13442, and in 0-1V mode I'd need BL=0 and BH=21782. Also, not sure how it clips input values outside the range, but sending it a 20000 produces the same output as a 4095.

So, based on those numbers, I figure something has to be wrong with the board. I was just hoping it was something user correctable. 🙂