Need fast continuous acquisition with FieldPoint modules

[Lowpass]

Hi
I have a FP-1601 with two FP-AI-100 which are used for continuous acquisition (all 16 channels). The program runs under LabVIEW 7.0 Express with FieldPoint 4.0.1.
I tried several ways to realize that: I used the FieldPoint Express VI, then the FPRead.vi (standalone, which is possible now) and in third case the "classic" way: FP Open.vi --> FP Create Tag.vi --> in while loop with "Wait Until Next ms Multiple": FP Read.vi--> out of loop: FP Close.vi
In all of the programs there runs a waveform chart and the Express-VI for saving the measurement data. Sampling rates higher than 7Hz aren´t possible. For a test I removed the waveform chart. For this case a sampling rate of 20Hz worked.
Only the 2 FP-AI-100 are c
onnceted to the FP-1601; the FP-1601 is connected to the PC via crossover cable.
Now my question: Can you give a hint how to make the continuous acquisition work faster / as fast as possible? I red in the forum that with this combination sampling rates of 80 till 90Hz are possible.
Thank you for your help.

Best regards,
Marc

[Lowpass]

I just modified the program. I think the Express-VI for saving measurement data was the problem. Now I use "Open/Create/Replace File.vi", "Write File" and "Close File". Sampling rates at 20Hz are possible. There are several things running in the background parallel to the acquisition so I think this result is quite well.

[nieroniero]

If anybody knows something more about this topic I'm also very interested

[Chieudam]

Hi ALL,

My set up is cFP2020 with one DI-301, two AI-112, and two TC-120. I do a scan all channels then save it to compact flash using SGL write module every loop. The fastest time loop I can get is about 250mS per loop. Can You please tell me if it is possible to make it faster? (Thank you very much) I'm looking for 10Hz for my DAQ system. I don't know how fast is the limit.

Thanks alot
Dam Nguyen

[sergesl]

Hi,

To optimize your program for speed, you need not use express VIs for your acquisition. Express VIs are made for ease of use, but they do not perform very well speed wise. In addition, you could have two separate loops, one with high priority where you are acquiring data and putting in a queue and a second with a lower priority were you are logging the data. That way, the acquisition always runs faster.

[Modrep]

Thanks SergisL,
That's what I came up with, too. I'm logging RPM, pressure(mA), flow(mA) and valve position(mA) from a turbine with a cFP-2020/CTR-500/AI-110. The two loop trick got it going, but I still have a problem with keeping the time stamp straight. Seems the difference in execution speed of the loops reeks havoc with logging the right relative time. Any ideas ?

Thanks

[sergesl]

If you want to also keep the time stamp, you can enqueue both the data and the time stamp from the critical loop. That way, when you dequeue, you have the original time stamp.

[Modrep]

Thanks, Why didn't I think of that!!