Continous acquisition on 800MHz, W_XP, with 6025E - has a problem on 200kS/s

[pawel]

Hello

I try to continously acquire data from two channels of PCI-6025E in Win XP, stand alone LV7.1 application.
It works well on 100kS/s, but it got real difficulty with going to maximum speed 200kS.

I set buffer to 1M, which is enough for 100kS, but soon is getting overloaded in case of 200kS and some often software triggering. But even without triggering it gets problematic. I found out that the 10-30kS per iteration is the optimal setting.

The question is whether it is my application fault, and this setup should easily run on this PC, or I am close to the limit.

Thank you in advance
Pawel

[BiG-C]

Hi Pawel,

You should be able to run 2 channels @ 200kS total (100kS each) since that it the max rate of the card. Your card is a multiplexed card, which means that the max sampling rate (200kS in this case) is divided by the number of channels you are sampling to obtain the max sampling rate per channel. I dont know if you are trying to run 200Ks per channel, but it should have given you an error stating that the sampling rate is not supported. I have done +200kS on slower pcs and laptops, so i dont think the issue is with your CPU speed.

What do you mean when you say overloaded? Does a -10846 software buffer overflow error occur? If so, you may want to consider the amount of RAM you have. Because you are running WinXP if there isnt a decent amount of RAM left ov
er after WinXP has taken its fair chunk of it, your buffer will be working slower due to OS file swaps. My favourite motto: you can never have enough RAM.

Hope that helps.

[pawel]

Hello Big-C.
I know I can sample (200kS/s/No.of Channels) per channel. I do not have a problem here.

What I mean is I have a software triggering since 6025E does not support hardware analog trigger. That means that I am running a detection procedure each time I get a new block of data from memory.

I looks like this procedure is taking up most of the processor time, and this is a limmiting factor.

I tried to acquire data without this trigger procedure, and get possitive result. I can read every 3k Samples, which come to the memory buffer.

It looks like I need to work on my treshold detection procedure. I have already found some creazy stuff going on, but that is a subject for a new thread.

Thank you for your comment.

kind regards
Pawel