03-19-2009 12:02 PM
One more thing:
You tell me : Accuracy is less
You tell me: Observed precision is reasonable (which is not good)
Then why this card is being marketed at such a heavy cost?
Alok Damle.
03-19-2009 12:36 PM - edited 03-19-2009 12:41 PM
Alok Damle wrote:One more thing:
You tell me : Accuracy is less
You tell me: Observed precision is reasonable (which is not good)
Then why this card is being marketed at such a heavy cost?
Alok Damle.
I understand your frustration.
The accuracy of this card may be less from the your application's point of view. That doesn't mean it is a useless card. It funtions well within its limitation (which is mentioned in the specification).
It will be a good idea to discuss this issue your local NI office and see if it is really a card issue. May be to discount the possibility of a faulty card, you can try with another 6259 card and see whether you get the same issue.
I guess the cost question you raised is a rhetorical one and if it is not then don't expect to get a satisfactory answer because perceived value is highly subjective.
03-19-2009 01:04 PM
Alok, I would be interested in seeing a couple of things, if they're not too hard to produce:
1. Can you show the drift just as a DC voltage vs. time? The presence of the ramp makes it hard to tell if the problem is random wandering or linear drift. Also, it would make it easier to reproduce the problem.
2. Can you do it again, except looping back to the 6289 instead of the 4461. I suspect the noise problem you mentioned could be solved with averaging. Also there is a hardware filter you can enable on the 6289, which would help a little. But if your sample rate is very low, then you'll need to sample fast and average to implement some supplemental digital filtering.
Using the 6289 for both source and measurement eliminates the voltage reference as a source of error, since it is common to both sections.
I'll go ahead and take some wrap-back measurements on a 6289 here and post what I find.
Chris
03-19-2009 02:00 PM
Here's about 5 minutes worth of drift of my 6289 wrapped back to its input internally:
Output voltage was about 9.5V. Acquisition was averaging 1000 samples taken at 500kS/s for each data point. The hardware filter was on.
Chris
03-19-2009 04:12 PM
Hey Chris,
Can you post that image again?
Thanks
03-19-2009 04:19 PM
03-20-2009 02:05 AM
Hi Chris,
I am travelling to france at the moment. would get back to our R&D center in UK by this weekend. will execute the same experiment as that of yours and will post the result.
Looking at the results that you got from your card, card fault can not be ruled out with the card that we have here. but rather than jumping to conclusion directly i will get the scan data and will tell you the result.
Alok Damle.
03-20-2009 02:15 AM
By the way Chris, thanks for trying out the test at your end. Looking at your results, the card seems to be suitable for application.
Regards,
Alok Damle.
03-25-2009 06:51 AM
Hi All,
After a series of tests with combinations of options, like changing offset, range reference etc with the 6289 AO we acquired the data to check the drifts. however what we came through is a drift ranging from 100 - 200 uV with 6289 which is not good for our application.
however now we have managed to solve the problem by modifying the technique used. now the voltage drift is very much within 10uV (peak to peak over 20 mins).
Many thanks for your help and time. We really appreciate it.
Regards,
Alok Damle.