record after trigger, then wait for next trigger: digitizer

[weizbox]

Im working on a program that requires me to look at a channel and wait for a pulse to arrive (triggering off the edge). When a pulse arrives I need to record for 5 microseconds (the largest of the pulse widths that I will be getting). After this I need to go straight back to looking at the channel for more pulses that may arrive and do the same thing. The problem right now is that I cant seem to get the speed of this operation as fast as I would like it. I would like to go from triggering the pulse, recording the pulse and go back to waiting for a trigger all within 20 microseonds or so. Is there someway I can store the pulses immediatly to the buffer and then read them later so the process can go as fast as pos
sible? After I get these pulses I would like to check them for pulse width/amplitude and eventualy match up a frequency once I get enough pulses in. The reason I can no just read for a long time and then get frequency from that is that I will possibly have more then one pulse coming in at a time, useing differnt frequencys, pulse widths, and pulse amplitudes, and everyhting premade in labview is set to look at one pulse at a time, but i could be getting a lot more. The triggering/recording idea is what Ive come up with so far, and if you could help me accomplish it I thank you very much. But if you have any other ideas on how to go about this problem, please let me know!

Thank you very much!
-Mark

What Im running:
LabView 7.1
PXI-5102 Digitizer
PXI-DAQ 6025E
PXI-8176
PXI-1002
Windows XP Pro
512 MB Ram
1.2GHz

[Jack_A.]

I'm assuming that currently you are using a While Loop or For Loop in order to acquire each of your records. Inside the loop, you configure the acquisition, read the data, and then save the data. Is this close?

You will be able to save quite a bit of time by storing the data in memory instead of writing to a disk. You can do this in LabVIEW by simply using a shift register or index your data as it comes out of your loop structure. You will want to make sure that your PC has enough RAM to store all of that data and still run your programs at the same time.

Once your acquistion is done, you can do any kind of post processing that you want and finally save to disk.

20 us is very fast and I'm doubtful that you'll be able to setup the 5102 in software quick
enough, but I haven't personally tried. If the 20 us is a hard and fast requirement, you might want to consider something like the 5122 which can do a multi-record acquisition, has a fast trigger rearm time, and can have up to 256MB per channel of onboard memory .

Good luck,
Jack

[weizbox]

I would rather use a method of saving it to memory, but this program needs to always be running, not just for a little while, and then process, its continuous. Im about as low a level that I know of for the time being, working at the .dlls that came with Scope, and still seem to not be able to get the speeds that I would like. With the card running at 20MHz I thought this wouldnt be much of a problem, but its turning out to be one. For the processing needs that I require I think I may go with an FPGA/ADC combo, unfortunetly NI doesnt make one with the speed that is needed. Still trying to get it working tho. Thanks for the help!

[weizbox]

I would rather use a method of saving it to memory, but this program needs to always be running, not just for a little while, and then process, its continuous. Im about as low a level that I know of for the time being, working at the .dlls that came with Scope, and still seem to not be able to get the speeds that I would like. With the card running at 20MHz I thought this wouldnt be much of a problem, but its turning out to be one. For the processing needs that I require I think I may go with an FPGA/ADC combo, unfortunetly NI doesnt make one with the speed that is needed. Still trying to get it working tho. Thanks for the help!