Data acquistion channel sequence

[czur]

Hi all

Am having a PXI 8186 embedded controller with 6280 DAQ card to acquire 92 thermocouple channels. The 8 scxi 1125 sits on the 1052 chassis along with the DAQ and pxi controller. rest of the 4 SCXI modules are placed in a 1001 scxi chassis. when I acquire data from all the channels the data received in the read function of DAQmx is not matching with the channel sequence I had supplied to create task function. the data for nth channel comes in the n-1th position. i have cross checked all the sequence and the data and compared them several times, but the result is same. but if acquire only the channels from the second chassis that is data from only the 4 modules the sequence of data is proper.

has any one faced such problem?....what might be the cause of this?

Regards

Arun

[Otis]

Hi Arun,

How are you reading the data from the last channel?  Are you using some sort of index array or index waveform VI?  If so, keep in mind that all Arrays in LabVIEW are zero indexed. 

That means that if you have 92 thermocouple channels then that will be array elements 0-91.  So if you want to read the 56th channel, then you need to look at the data on row 55.

Regards,

[czur]

Hi

It's not the problem of the index of the array. the thing is when i acquire measurement from all the channels one of the channel is skipped and in place of it the data of the channel next to it is written. and the channel which is skipped is not at all in the output array of the DAQmx read function. And another thing that i observed just yesterday, this problem persists only when i acquire measures from all the channels, if i acquire the measure excluding the first module, the data sequence is fine. another thing that i observed is if i acquire data only from the first module the measures for some channels are totally different and also in range of the thermocouple reading. the measure is not floating and is perfectly stable (thermocouple is not open).

Regards

Arun

[Otis]

Is there any consistency in which channel in the scan list is the skipped channel?  Which DAQ device are you using to communicate with the SCXI Chassis?  Have you tried changing which module is the communicating module?  If you do that, does the problem stay with the offending module, or follow the module that's been connected to?

[czur]

After facing this problem i had deleted and reconfigured the chassis, and the problem has changed to a different one. the DAQ card am using is 6280 m series. If I acquire data from all the channels, the data from the first chassis (1052) are good for some channels and very high values for some other channels, although it does seem like the channel responds to increase and decrease in temperature. the measures from the second chassis are all bad. if i acquire data from only the second chassis the measures are fine. if i acquire measures  from only the first chassis the measures are similar to what i observed when acquiring all the channels.

We also have a another exactly similar data acquistion system which works perfectly fine and the configuration (in max) of both the working system and the bad one are same.

I tried replacing the 1001 chassis of the working system with the bad one and the acquiring all the channels, the measures seems to be fine in all the channels. also i tried to change the chassis communicator to different module but the problem persists.

Thanks for the reply.

Regards

Arun

[Otis]

Hi Arun,

It sounds like there may be one or more open Thermocouple channels in your chassis.  It could be that these channels are floating high and if you are reading a good channel directly afterwards then it could cause a faulty measurement.  Hence one chassis appearing to be bad, but the next appearing to be fine, but if you use both then everything looks bad.

The best way to test this theory is to increase the Convert Clock rate (increase the amount of time between sampling channels to decrease any potential error from the first chassis).  To do this in LabVIEW you will need to use a DAQmx Timing Property Node and increase the rate that way (the best rate will be 1/(sampling rate*number of channels) to make it more of a round robin style sampling).

You may also want to measure each channel individually from MAX to make sure that none of the channels are giving you faulty readings when read on their own.

Regards,