Multifunction DAQ

cancel
Showing results for 
Search instead for 
Did you mean: 

DAQ Card sampling clock voltage disagreement

Solved!
Go to solution

Hi, I am using a NI PCI-6281 trying to acquire signals in accordance with a sample clock. 

I am trying to acquire signals on two channels within this DAQ card, however the amplitudes that the DAQ card is measuring do not match the amplitude as measured by an oscilloscope and I am not sure what the problem could be (I've gone through the documentation and based my set-up on what I have read there)

I have included diagrams of my signals, my sampling clock, and my trigger as measured by an oscilloscope.

 

The first image, "BothSignals1M.jpg" Illustrates the magnitude of both signal as measured by an osciloscope, using highZ input impedance (1mega-Ohm). One signal is in yellow and the other in blue (sorry about the overlap). It can be seen here that the yellow signal (hereafter called the Reference) peaks at a value of about 1.7V, and that the blue signal (hereafter called the Probe) peaks at a value of about 1.4V

 

Just to make sure, I have done the same for both signals using 50 Ohm Input impedance into the oscilloscope, and the signals in this case are shown in "BothSignals50Ohm.jpg". In this case the Probe (blue) peaks at about 0.7V and the Pump (yellow) at about 0.85V.

 

The sampling clock is a TTL signal, which is shown in Green in "Sampling ClockTTL Overlap with Signals.jpg". In this case, everything is measured with highZ impedance. It can be seen that the sampling clock is timed such that a sample is collected at the signal peaks.

 

The trigger I am using is shown in Magenta in "Trigger with Signals and sampling clock.jpg" and is also a TTL pulse. It can be seen that the trigger is a little bit after the peaks.

 

 

Ok, so, assuming this setup, we now switch to using the DAQ instead of the oscilloscope. I have connected the reference and pump into a pair of analog inputs in the daq card, ai0 and ai1. When using internal sampling (instead of a sampling clock) I can observe that the signals are present, and so can conclude that these channels have been connected correctly. Next I have connected the trigger TTL signal into the PFI3 input of the daq card (corresponding to pins 42 and 9) and I have connected the sampling clock TTL signal into the PFI2 input of the daq card (corresponding to pins 43 and 44).

 

Now, when acquiring a signal, I should expect to see only the peak values. However, when I run a test, the daq card is not recording accurate values. The recorded values by the daq card are shown in "Recorded values on the NI DAQ.jpg". My reference beam (this time shown in blue) is read by the DAQ card to be peaking at a value of around 1V, and my probe beam (in red) is shown to be peaking at a value of 0.4V. These do not match the peak values as recorded by the oscilloscope, and I don't know what I'm doing wrong. I thought maybe the sampling clock is not recording at the peaks or maybe the trigger signal is affecting when the DAQ card takes a measurment but I don't really have a very good idea.

 

Can someone suggest possible problems with what I have done or otherwise explain why the values my DAQ card records don't match the values that my oscilloscope is recording? Thank you so much!

 

 

 

0 Kudos
Message 1 of 13
(3,947 Views)

Can you post a picture of how your hardware is setup? Also can you post your code? It is hard to say what could be going wrong at this point.

Maggie M.
Application Engineer
National Instruments.
http://www.ni.com/support
0 Kudos
Message 2 of 13
(3,893 Views)

Hi, thanks for responding!

I have since looked into some papers using this DAQ card and I believe that the problem is because of the sampling rate of the daq card. My signal is much narrower than the sampling rate of the DAQ card and so therefore the DAQ card is not able to precisely detect a peak that is finer than 1.6us. I am pretty sure that that is the problem, does that make sense?

 

Thanks again!

 

 

EDIT: whoops this is from my other account! I am OP

0 Kudos
Message 3 of 13
(3,889 Views)

The timescale in the diagrams is 400ns per horizontal square so that corresponds to a much higher samplign rate than kHz

0 Kudos
Message 4 of 13
(3,877 Views)

Looks like a counter signal, do what to make energy spectra? 

In the old days of slower sampling possibilities, the solution was a pulse shaper  or a fast analog peak hold.

While the peak hold tries to preserve the peak value, the pulse shaper try to preserve a ratio of the energy, so you sample more of the pulse and and use again the peak value or the integral of hole peak... there must to tons of papers covering that 😉 

And no, that's no my field of experience, theres just a scintilation PMT at home waiting for hobby spare time, so I kept some information in mind 😄

 

Next thing: There is a time gap between the trigger and the first value captured (trigger delay), more in the spec...

Greetings from Germany
Henrik

LV since v3.1

“ground” is a convenient fantasy

'˙˙˙˙uıɐƃɐ lɐıp puɐ °06 ǝuoɥd ɹnoʎ uɹnʇ ǝsɐǝld 'ʎɹɐuıƃɐɯı sı pǝlɐıp ǝʌɐɥ noʎ ɹǝqɯnu ǝɥʇ'


Message 5 of 13
(3,868 Views)

I found a couple of white papers on pulse shaping that may be helpful.

 

http://www.ni.com/white-paper/14849/en/

http://www.ni.com/white-paper/3876/en/

 

Can you provide more details on your sampling rate and the rate of your pulse?

Maggie M.
Application Engineer
National Instruments.
http://www.ni.com/support
Message 6 of 13
(3,840 Views)

Thank you magpie15,

My sampling rate is 625k in the DAQ card. The pulse is a 7ps Gaussian (in time) laser pulse, which has a repetition rate of 400 pulses per second. We are detecting the laser pulses with thorlabs PDA100a photodiodes, https://www.thorlabs.com/thorproduct.cfm?partnumber=PDA100A, which have a rise time of about 145ns. The voltage signal is then sent through a preamplifier and then to the daq card. We are trying to replicate the experiment done in this paper.

0 Kudos
Message 7 of 13
(3,836 Views)
Solution
Accepted by topic author hex12345

With a pulse of 7 ps, you are probably right that you are not sampling fast enough.

Maggie M.
Application Engineer
National Instruments.
http://www.ni.com/support
Message 8 of 13
(3,805 Views)

can't you read out the Tek scope?  OK, 8bit resolution, but better than nothing 😉

 

Greetings from Germany
Henrik

LV since v3.1

“ground” is a convenient fantasy

'˙˙˙˙uıɐƃɐ lɐıp puɐ °06 ǝuoɥd ɹnoʎ uɹnʇ ǝsɐǝld 'ʎɹɐuıƃɐɯı sı pǝlɐıp ǝʌɐɥ noʎ ɹǝqɯnu ǝɥʇ'


Message 9 of 13
(3,794 Views)

In addition to what Henrik_Volkers said, have you ever been able to read the pulse? 

Maggie M.
Application Engineer
National Instruments.
http://www.ni.com/support
0 Kudos
Message 10 of 13
(3,773 Views)