what is the single sample for a single channel acquisition signal?

[Identcor]

We have a NI system composed by:

            PXI-8196

            PXI-6251

            SCXI-1141

            SCXI-1304

We need to implement a system that acquires a signal and ininterruptly transmits, with no lost of data, all the samples to a remot (serial port) to a laptop where the signal will be analysed .

We need to acquire one sample at a time and transmit it, that is, before acquiring another signal sampling I need to have transmitted and stored the last one. It has to be done in up to 4 ms at least, that is my sample frequency will be 250 Hz minimum.

We are using LabView 7.1. In order to check these timings we implement the AcqTiming.vi in attchment. We sourced it with a senoidal signal (100 mVp, 10Hz) and configurated 1000 Hz to the cutoff frequency filter, then we measured the timing it takes to perform one single sampled. We´re disappointed when saw it taked more than 30 ms to do it. This way we could not use LabView to do this. Are we doing anything wrong?

If anyone help us we would thank you.

Sincerely,

Celia

[Dennis_Knutson]

In your for loop, you are continuously starting and stopping the task. These two functions take quite a bit of time and there is no reason that you have to do either more than once in a program. Moving these to their correct locations will significantly improve your timing. Whether it will be fast enough, I don't know. You are using software timing with this technique and that is going to be highly variable depending on your OS and type of pc.

[Norbert_B]

Celia,

if i may ask: are you creating some kind of control?
If not, why the hack do you want to acquire the samples as single points??

Regarding performance: Some of the 30 ms are wasted due to your programming. You have to move start and stop task out of the for-loop.

hope this helps,
Norbert B.

[Identcor]

Thank you for your answers. In a matter of fact I´ve already tryed to put start and stop out of the loop, but I forgot that I had changed from the initial "while loop" to the "for loop" I put to measure the acquisition time. So, when I tested start and stop out of the loop with 30 loops the waveform was bad and I thought it was not working. When you sugest this I tried it again and I noted what was happening. The single sample time is around 18 microsegundos.Thank you very much again.

Hugs,

Celia