Frequency error when generating/measuring a swept sine with PXI-5412/5124

[MS08]

I would like to generate a swept sine wave, with frequencies ranging from 1kHz to 6kHz.  The signal will be generated with a PXI-5412 Arb, fed through a bandpass filter, and measured by a PXI-5124 Scope.  I am curious to know what kind of error I can expect from both the generating and measuring devices.  If I am generating a 1kHz sine wave in the arbitrary waveform mode, can I expect this to be +/- 1Hz, 100 mHz, 10 mHz, 1 mHz?  I would also like to know the same thing for the digitizer.  I cannot find this in the specs, though it seems like it would be a function of the sampling rate error and A/D or D/A errors.  I was hoping that this would be a common question that someone else has encountered.

Thanks,

Mike

[Aaron_W.]

Mike,

Thank you for posting on the NI Forums. The spec that you looking for is the Frequency Accuracy for the PXI-5412. The Frequency Accuracy for the ARB is spec'd at +/- 25 PPM (Parts per Million) or if you are generating a signal at 1 MHz you should expect an error of +/- 25 Hz. So for your case of 1 kHz to 6 kHz you could expect an error anywhere between +/- 25 mHz @ 1 kHz, to +/- 150 mHz @  6 kHz. 

With the PXI-5124 you are looking for the Timebase Accuracy spec. This spec is listed as 25 PPM when Not Phase-Locked to the reference clock of your input device. If you do not Phase-Lock the reference clocks together you could get an error of +/- 50 mHz @ 1kHz to +/- 299 mHz. The math is not as simple as adding them together. For instance to find the worse case scenario the math would be as follows: [1 - (Accuracy Error for ARB @ Rate)] X [1 - (Accuracy Error for Scope @ Rate)] or [1-(.000150)] X [1 - (.000150)] = .9997000225 then 1 - .9997000225 = .0002999775 or approximately 300 ppm. This is the absolute worse case for the scope not being phase-locked to the reference clock of the ARB. If you do phase-lock the reference clocks together your timebase accuracy for the scope is equal to that of the reference clock (or ARB in this case), so it would be +/- 25 mHz @ 1 kHz, to +/- 150 mHz @  6 kHz. 

 I have linked the detailed specifications for both the PXI-5412 and PXI-5124. For the PXI-5412 the spec for Frequency Accuracy is on page 11. The Timebase Accuracy spec for the PXI-5124 is on page 14.

NI PXI/PCI-5412 Specifications

NI PCI-5124 Specifications

NI PXI-5124 Specifications

Hope this helps, let me know if you have any other questions. 

[MS08]

Aaron,

Thanks for the quick and thorough response.  This certainly addresses the possible timing errors for each instrument, but is there any contribution from the A/D or D/A conversions?  I'm guessing that any "noise" generated during the D/A (arb) or A/D (digitizer) conversions would likely appear as wideband or high frequency noise.  My main concern is any error in detecting the fundamental frequency in kHz - amplitudes are a distant second.  I believe your first answer should suffice, but I was hoping you could confirm my followup posting.

Regards,

Mike

[Aaron_W.]

Mike,

It appas as though the AC Amplitude Accuracy for the scope is spec'd at 0.7% of the signal amplitude at 50 kHz, so I would expect this to be similar at the range of 1 kHz-6 kHz. As far as the ARB goes it is spec'd at 0.006% of the amplitude range. I wouldn't expect these errors to give you much of a problem though, but it would depend on what the amplitude of your signal is going to be.