pxi-5105 clock and trigger

[Aaron_W.]

Douglas,

There will be a  "safety factor" on the input clock. If you look at the specifications for the PCI/PXI 5105 the specification for "Duty Cycle Tolerance" on page 8 is this factor. The actual clock's Duty Cycle can vary between 45% and 55% of the user input clock rate. We do have customers as if you can vary your sample clock as you are acquiring with an external signal, and the answer is that you can, but the duty cycle of one period of your clock to the next cannot vary out of the range of 45-55%. 

Overall, if you estimate your sample clock and your input is within those limits you will not have a problem. 

Thanks!

[Douglasg610]

Unfortunately, the CFD output, which we are using as the clock replacement, is random.  A cursory check using Ni-Scope Express indicates that the phase-locked-loop of the clock does not work with the signal, shown on channel 1 (the positive black signal).  The effect, shown in images 2 and 3, is that the clock pulses would be only randomly spaced, making a stable rate of sampling difficult.  The error shown in image 1 indicates this is the case.  IS there anything I can do about this?

Do I need a system that does NOT use phase-locked loops?

[Douglasg610]

One of my colleagues produced this as an alternative code, but it uses DAQmx modules, which are different.  I am unable to make a DAQmx "task" using NiScope components (MXI-4 link, PXI-1042 chassis, two PXI-5105 digitizers).  Is there a way to re-write the code using NiScope modules?

[John_P1]

Hi Douglas,

From what I've read here it sounds like you want to give a random clock signal to sample your data.  This is not possible on the 5105, since it requires a free-running sample clock.  What you can do on the 5105 is trigger multiple records off of this external random signal.  In this case, you would use the internal clock of the 5105 as your sample clock, set minimum record length to 1, and re-trigger multiple records off of your external random signal using a reference trigger.  Keep in mind the re-arm time is 2.4 u seconds (416 kHz).

If this is indeed what you want to do, a better solution would be to use a Multifunction DAQ device (which does not require a free-running clock).  So, instead of using the signal as a trigger you could use it as a sample clock and acquire up to the maximum rate of the card.  For example, the 6250 can acquire up to 1.25 MHz on a single channel.

So, it sounds like the 5105 might work using a reference trigger depending on the frequency of your random pulses (if they are too fast you will miss triggers during the re-arm time).  The requirement of a free-running clock makes the 5105 less than ideal for this application.  I hope this helps,

-John