Problem with response time measurement (around 1ms)

[Jenna]

Hi,
I am developing a system to measure the response time of our product, pneumatic valves, which can be turned on/off by input DC voltage. DAQ board: Keithley KPCI-3108. LabView version: 6.1. OS: Win2000 Pro.
My problem is that, I am getting very different results of response time from the Labview system between the current system we are using, which is only use hardware (e.g. scope, function generator, similar current-drive circuit board as Labview system). From the old system, if I get around 2ms response time, I will get around 6ms response time from Labview system.So I need help to figure out what causes the difference.
My labview system uses analog output channel 1 to get the DC signal and output to my e
xternal current-drive circuit, to turn on/off the unit under test. And it uses analog input channel 2 to read the pressure transducer. The response time is calculated from the valve loses power (AO signal goes to 0Vdc) to the valve really turns off (AI transducer voltage starts to drop). I also measure both signals from AO1 and AO2 channels on the ossiloscope, and I can get the same results from the scope and the labview graph. Also, I wired the AO output channel 1 back to AI channel 3, therefore, AI channel 3 is the reference of AI channel 2, I just wanted to graph AI 2 signal whenever AI 3 goes to 0Vdc. I was wondering why on the scope, AO 1 channels shows me the same signal as AI 3, but when I graph AO1 (before buffer) and AO3 (after reading from buffer), I get non-simultanous signals.
I attached my application VI here, please help me to figure out what I did wrong, or what could possibly cause the problem (if it is a problem). I do appreciate any help!

[Russell]

Hello Jenna,

I suppose I could offer some troubleshooting suggestions. I don't know anything about the Keithley cards, but maybe I can give some insight. One test you could run is to use your hardware instruments to run the test. In addition to connecting your scope to the two analog signals at the pressure transducer, also use your DAQ board to read these two channels. This would at least help narrow things down to determine if the difference is due to the AI. If this reads exactly the same as the scope, the difference might lie somewhere in the AO side of things.

Anyway, my personal hunch is this. Again, I don't know about the Keithly boards, but the NI E Series boards don't actually sample all their channels simultaneously. All of the channels are ac
tually read through a multiplexer so there will be a slight phase difference when comparing channels to each other. I suspect the Keithley board may act the same. You might look at it's user manual or even contact someone from their technical support for more information on this.

The second thing would be to increase the scan rate the DAQ board is sampling at and see if that changes anything.

I guess my last piece of wisdom would just be to make double sure you are using the same comparison criteria in your LabVIEW VI that you are when looking at your scope. (Verify the comparion portion of your code with some known numbers. Maybe make it into a subVI and run a couple of known waveforms through it.)

I hope this helps

Russell
Applications Engineer
National Instruments
http://www.ni.com/support