How can I synchronise to RFSG with sawtooth FM mod sweep to the RFSA to perform fast tracking function

[MHR]

The RFSG can be made to do fast sweeps by using FM sawtooth modulation, with previouse drivers it was down to 1Hz, I find now that this is restricted to 2Hz due to the AWG minimum sampling rate being set to 1E+7, below results in error. At the moment this part may be workable, however I do want to synchronise the RFSA (5660) to the RFSG (5671). As far as I can see there are two main issues:

a) The RFSG begins sweep at its centre frequency, proceeds high util reaches highest extream, then imediately goes to it most negative extream value, then finally prceeding back to centre.

b) The RFSA begin freq sweep at Min freq and goes high.

I need to track these together without the time-overhead of reloading each for descreat frequencies increments.

The intended use is for user interactive adjustment that would normally be mode useing a network analyser for insertion loss of RF filters.

Any help will be appreciated

Thanks

Mike 

[AdnanZ]

Hi Mike,

Hope you are doing well today! I am escalating your question to the R&D team in the US and will reply to you once I hear back from them. Thanks for your patience!

[AdnanZ]

Hi Mike,

In general, applications that normally require a Vector Network Analyzer can not be accomplished with one of the RFSGs and one of the RFSAs. 

The requirements of producing a VNA include the ability to calculate the phase difference between multiple signals. Unfortunately, even multiple PXI-566xs are not able to reliably operate as a vector network analyzer because there is currently no mechanism to share a local oscillator (LO) between all three devices. Although each device is able to phase lock (PLL) to a common 10 MHz PXI system reference clock, this clock applies only to the digitizer and not to the downconverter. In fact, each incoming signal is mixed with a local oscillator in the downconversion (PXI-5600) stage. As a result, the phase difference between each RF signal will depend upon the phase difference between each local oscillator. Thus, multiple PXI-566x’s cannot be combined to produce a vector network analyzer because they cannot measure the phase difference between multiple channels.
 

Basically, the arb, upconverter, digitizer, and downconverter cannot PLL to the same clock. Let me know if this answers your query!

[MHR]

HI Adnan

My requirement does not include phase, I am only interested in insertion loss. On the face of it, it should be possible. according to one of my collegues we are not unique, I was hoping there was an existing example VI, but it seems not, therefore we will have to invest in the time and create one of our own. Also one this point, what has happen to the RFSG driver in recent times that has restricted the minimum sample rate of the AWG, which used to go down to 1E+6 now will only go down to 1E+7 with a default rate of 1E+8.

Thanks

Mike

[AdnanZ]

Hi Mike,

I would like to have some more information. How exactly are you trying to do your frequency sweeps? Have you written your own program?  How do you know the RFSA is beginning it's frequency sweep at the minimum frequency and then going high?

Also, what exactly do you mean when you put the minimum sample rate in terms of 1E+6, +7, or +8?  For example, the arb in the 5671 (a PXI-5441) has a maximum output rate of 100 MS/s, which would be 1E+5.  Is this a parameter you're putting into a VI? Are you getting any errors with your setup?

[MHR]

Hi Adnan

The original idea for the sweep was coded by myself some month ago for performing IF bandwidth tests, when initially coded we was running labview only and the routines worked.

The FM sweep was produced my modifing some NI example code, the essential features were to change the modulation signal type to Sawtooth wave and to use Export Signal with output selected to PFI0 between Write Arb Waveform and Initiate VI. The trigger in this case wat to start the Digitiser audio recording of the receivers audio line output. The only other issue was to run the AWG at a slower rate as possible, this was taken down th 1Hz, however this could not be done using using the defaultIQ rate S/s which you quite rightly said was 100Mb/s or (1E+8 as shown in the labview property). At the time this value could be set down to 1E+6, thus allowing a 1 Second sweep to take place.

The sweep starts a the Zero crossing point, then goes +, but this was not a problem as we were able to detect the abrupt change from + extreme to the - extreme and use that as a marker for the recovered response.

When were first wrote this application we had little experiance of Labview / test stand etc, We first tried to integrate this into Test stand 3.5 and were not able to run the RFSG and Audio digitiser simultaneously, late we found this to be due to our time VI settings, (it was not recursive only one instance could be run at any one time).

Later we tried to reserrect this into test stand 4.00 but during the intervening period we had driver upgrade, as a result of this the IQ sample rate cannot be set below 1 E+7. The sweep can be made to work but no lower than 1.5Hz which does produce a coersion warning. I also noticed some else that had changed, I cannot set an RFSG property node which show the same node name but one as an input but with the other as an output, two node with the same name can exist but, if you change one from a read to a write both change and vica verca.

The theory behind my immediate requirment, is to drive the RFSG in a similare way to the above, but to limit the number of samples (duration) to a half cycle, this will produce a frequency sweep in one direction only, this should be easy for the RFSA to follow it, both unit sould be linked together via the PFI0 lines to allow the RFSG to send the start sync to the RFSA to follow, clearly the sweep time and freq range of the RFSA must match the RFSG, and bandwidth of the RFSA should be suffiently wide to prevent tracking errors.

I hope this helps clarify what we are trying to do.

regards

Mike

[AdnanZ]

Hi Mike,

We added a property node to modify the IQ Rate (S/s) in the example "RFSG Frequency Sweep.vi" from our example finder. We were able to use a wide range of values from 100 up to 1+E8 for this property node without an error. As a result I can't be confident this is actually what is going on in your program.

Can you post the modified code or a screenshot of what you are doing?