09-16-2025 01:32 PM
Hi Team
I am using PXIE-5451 Fgen instrument to generate sine wave and PXIE-6363DAQmx (AI) to capture the sine wave ,but here i have a concern that these two instrument reference clocks are PXI_CLK10 and PXIE_100Mhz respectively and the chassis is PXIE-1095.
Please guide me to synchronise both the instruments.
Regards,
Sagar
09-16-2025 08:18 PM
According to PXIe-1095 User Manual
The backplane uses a 100 MHz Voltage-Controlled Crystal Oscillator (VCXO) to directly create PXIe_CLK100 and does a divide-by-10 to create PXI_CLK10.
For synchronization, you can refer to community example Sync a FGEN, HSDIO and M-Series card with a Timing and Sync Card Using LabVIEW with DAQmx
09-17-2025 02:01 AM
Hi Zyong,
Thanks for the effort.
From the shared examples it appears that a timing module card is required to synchronise instruments.
is there any other way to synchronise mentioned instruments with out using timing card?
Regards,
Sagar
09-17-2025 06:53 AM
What is your intention to synchronize? to match the start of some generation and capture or sync to the sample level?
09-17-2025 07:04 AM
Very generally, the more you try to approach perfect synchronization, the more you get mired in diminishing returns. So it's pretty important to define exactly what degree of sync you really need. And why -- because if you can't answer the why, you don't actually know the what. Things like a timing module and careful use of Star Trigger signal paths can produce a rather extraordinary level of sync. But again, do you need it?
To help you find the appropriate bang-for-the-buck level of sync, please define the degree of sync you're trying to achieve and the consequence for your system or analysis of not achieving it. For example, suppose you were aiming for 5 nanosec skew and/or uncertainty but can only manage something like 25. How would that affect the thing you're trying to accomplish?
-Kevin P
09-25-2025 07:01 AM
Inside the PXIe Chassis the 10 MHz and 100 MHz Refclocks are always synchronised. Both of your PXIe cards can be synchronised to one of these clocks. (use one of these backplane clocks as reference clock). This done, even if you run for days and weeks at different samplerates, the samplerate ratio stay constant. The timing card usually has better internal reference clocks (less ppms to ppbs absolute and less phase noise) and allow more methods to synchronise different tasks and triggers.
As Kevin sated, with some more coding you can deternime and shift the phase relation of these two clocks in very precise (small) quantities. (Finally restricted by the residual phasenoise of the individual clocks and remaining jitter). Also a common start (if needed with defined delays for each card) within some tiny uncertainty (depends on actual hardware) can be realized.
But keep in mind every meter of cable has about 4-5 ns delay 😉