DAQ self-test, buffer checks, diagnostics, ... anything?

[ngio]

Hello all,

I have previously been monitoring some high voltage components as well as using a NI9401 board and two NI9201 boards (on a NIcDAQ-9178 chassis) and would like to know if there is a way to conduct a self test on the DAQ hardware to ensure I havn't broken any of the internals.  I ask this because I am receiving some real strange results when using an incremental encoder and converting the A,B, and Z phases to angular position.  I have included a screen pic to show what's going on (the shaft is continuously rotating, therefore the blue line should look like the pink line just with a different period). I have replaced the encoder and optical wheel to ensure that it is not the cause.

Please if anyone has another reason for this kind of behaviour I would be very interested to know about it, I have been trying to figure it out for a week now and have come up with no other reasons.

Any help is appreciated,

Nick

[AnalogKid2DigitalMan]

Can you run any self-tests through Measurement and Automation Explorer (MAX)?

-AK2DM

[ngio]

Thank you for the quick response, I will perform some tests and respond tomorrow morning. I only found one article related to this (http://zone.ni.com/devzone/cda/tut/p/id/4638) do you have any other suggested tests I should run?

I should add that I am still able to generate a delayed digital output pulse from CTR1 (which is delayed by counting pulses on CTR0 - the counter setup to monitor the encoder) to trigger external hardware.  I don't know how the internals of the DAQ board work although is it possible that since I'm rounting the digital output line to an analog input line (on the 9201) as well as external hardware, that the current draw is too high and this results in a malfunction of the other digital lines? (I was lead to believe there is circuit protection for stuff like this but maybe not)

Thanks again.

[John_P1]

Hi ngio,

Let's take another look at the graph in your first post (with some crudely drawn lines added courtesy of MS Paint):

Here are my two observations (which you've probably noticed as well, but I just wanted to call them out):

1. The slope is consistent for a majority of the time.  If you extend this slope, it goes from 0 to -360 between each reset (which I'm assuming corresponds to the z index).  This should be the expected result, so it's worth noting that your counter is functioning properly during the majority of the time.

2.  The actual measured value deviates from the expected value somewhere in the middle of the transition from 0 to -360 degrees (it looks like this disturbance occurs within about ±15 degrees of -180).  This likely corresponds to missing counts (less counts per time = more gradual slope), especially considering that one full rotation (between Z indexes) is not adding up to the full 360 degrees.  After the missed edges occur, the encoder continues to behave as normal (it resets to 0 at the Z index, then continues to count correctly until it reaches approximately the same point and misses counts again).

It's interesting that the samples are seemingly being missed at a consistent phase (at least based off of the two periods shown).  Something interesting must be happening to the signal at that point--have you tried scoping the A and B outputs of the encoder? I would suspect the 9401 is functioning properly given that it seems capable of counting the encoder as well as consistently registering the z index, but there seems to be something unexpected happening to the signal as you get close to -180 degrees.  I know you've already mentioned replacing the encoder and optical wheel so I think the best course of action is to actually look at the signal coming from the encoder to make sure it corresopnds to what you would expect.

Best Regards,

[ngio]

Hi John P,

Thank you very much for the post, I did however find the culprit of the issue (I.e an off-center mount which causes deflections up to 1-2mm, this resulted in the encoder sensor missing the wheel signals only at certain angles.) Finally It's fixed!!


On the other hand, just out curiosity I suppose, but if a certain section of the counters buffer were exposed to high voltage accidently would it destroy the entire counter or would those "slots" become inoperable. I'm thinking it's based on the design of the hardware, of which I know very little about. Any input is gladly accepted.

Thanks again,
Nick