issue with PPS signal to trigger AI start

[TiTou]

Thank you for your suggestions

---

@mcduff wrote:

Try to explicitly set the reference clock rate after setting the source. For some X-Cards I believe it is 100MHz, for your DSA card it should be 10MHz. See if that helps.

---

The PXIe-4468 will only accept 100 MHz, so I set it explicitly but it didn't change anything.

---

@mcduff wrote:

Lastly, how are you measuring the phase drift?

---

The analog output generated by the PXIe-4468 goes on a cRIO based PMU (Phasor Measurement Unit).
The PMU is gps synchronized (same GPS antenna connected to the PXI-6683) and it sends via UDP (using C37.118 standard communication) the phase and amplitude measured.

As I compensate the AO filter delay and make sure that the 50Hz pure tone starts on a rollover second, I expect the phase to be pretty close to 0° - and it is the case - but I also expect the phase to be somewhat constant through time - and it's not not.

---

@mcduff wrote:

One more thing to try, query the output sample rate after setting it; you may not be at exactly 20kHz like you assume.

---

Good point, I checked and it returns 20kHz.

[mcduff]

Let's take a step or two back:

• Initially the problem was with t0 when you use Analog Input. I assume that has been corrected now. Is this true?

New issue with Analog Out, sample clock seems to drift from your measurements, is that correct?

• According to the specifications GPS has ±40 ns, <8 ns standard deviation synchronization performance. Assuming worst case performance of 40ns, my quick calculation(please check) gives a max phase uncertainty of ~0.29 degree for 20kHz wave. Not sure what you have on your plot, but are you within specs?
• I still don't understand the phase delay on the Analog Out, a simulated instrument gives an error when trying to read that property node.

Do you have another GPS, like a Trimble? If so, output a PPS from both the Trimble and 6683H and digitize both on your card. See if both stay in sync for a long period. Doing that, I discovered an unexpectedly long trigger delay of about 400ns when using the PXI trigger line and a NI-5172 Scope Card.

The 6683H should rephase its clock every PPS, but there can be drift in that 1 second time frame where you are doing the measurement. Not sure if you are seeing that.

[TiTou]

---

@mcduff wrote:

Let's take a step or two back:

• Initially the problem was with t0 when you use Analog Input. I assume that has been corrected now. Is this true?

---

Yes.

---

@mcduff wrote:

New issue with Analog Out, sample clock seems to drift from your measurements, is that correct?

• According to the specifications GPS has ±40 ns, <8 ns standard deviation synchronization performance. Assuming worst case performance of 40ns, my quick calculation(please check) gives a max phase uncertainty of ~0.29 degree for 20kHz wave. Not sure what you have on your plot, but are you within specs?

---

Yes.

---

@mcduff wrote:

• I still don't understand the phase delay on the Analog Out, a simulated instrument gives an error when trying to read that property node.

Do you have another GPS, like a Trimble? If so, output a PPS from both the Trimble and 6683H and digitize both on your card. See if both stay in sync for a long period. Doing that, I discovered an unexpectedly long trigger delay of about 400ns when using the PXI trigger line and a NI-5172 Scope Card.

The 6683H should rephase its clock every PPS, but there can be drift in that 1 second time frame where you are doing the measurement. Not sure if you are seeing that.

---

I might be able to borrow a PXI-6289 (multi-function DAQ) to help with that.