Analog input errors when using multiple channels with different ranges

[Chris90265]

Has there been any progress on this issue? This is a serious bug for us and the requirement for extra channels is limiting how fast we can run the hardware.

[_kb]

Hi Chris,

We're researching long-term solutions.  There should be more information available next week.  For now, there are two workarounds, and the tradeoffs are maximum sampling rate and measurement sensitivity.  For applications where the sampling rate is more important, I'd recommend using a consistent input range across all channels.  Where there is more emphasis on sensitivity, using the extra channels is the way to go.

For your application, how many channels are used, what sampling rates are required, and how are your accuracy requirements defined?

[Henrik_Volkers]

Chriss, I don't know your needed uncertaincy, but to solve the issue in hardware (seems that you need to cover two ranges) : TI abd AD make nice high side current shunt amplifiers, that will bring all shunt voltages to the same (low impedance) level .

[Chris90265]

Thank you, Henrik. If we have a chance to redesign our board, the LTC6102 will work nicely.

[John_P1]

Any updates on CAR #483646?

Best Regards,

[Chris90265]

I have not heard anything. For me, this has turned out to not a big problem once identified. Still, I feel sorry for the poor suckers that are dealing with this bug thinking it's something wrong with their setup.

[John_P1]

Yeah, I just came across this other post and it seemed a little suspicious:
http://forums.ni.com/t5/Multifunction-DAQ/Single-DAQ-task-six-channels-RSE-and-Differential-modes-mi...

Now I'm also curious for my own sake--we use a good number of X Series cards in our own applications.

[_kb]

Hey John,

The CAR has been closed.  Our initial efforts to apply a software correction to the poor behavior hit a dead end when it was determined that a hardware constraint wouldn't allow for such a fix.  Our investigation of long-term solutions has made way for more auspicious corrective action, but there is no CAR tracking these changes.  Completion is still several months away.  We are in the process of updating documentation, so that any updates will be made clear with any future release of hardware or drivers.

I can assure you that this issue is limited to multi-range tasks run on 632x or 634x boards.  There is only cause for concern if you have applications where these criteria are met.  Please refer to this article for recommended workarounds.