How to make RF frequency match dynamically with SRF cavity resonant frequency change

[zyb1003]

Hello guys, 

Currently, I have problems with the RF sent to SRF cavity. 

For the reason that the SRF cavity's resonant frequency will change slightly (maybe 1HZ) and frequently (maybe 100HZ) around its marked resonant frequency in super conducting environment.  so we need to make the RF frequency match with the resonant frequency change. Now in my point of view, the way should be monitoring the amplitude of RF feedback signal, if the amplitude reaches largest, then we can say that the RF frequency matches the resonant frequency of the cavity. But how to make the RF frequency change automatically to follow that?....Also, we need the system response time very quick (at least 0.1s). 

Now we have NI 5651 RF generation card and LabVIEW. 

Anyone has a good idea for this problem? 

Thanks

[AgentAstronaut]

Hi Zyb,

It looks like you are trying to do some kind of autotuning with your system. I couldn't find any reference material to look at, but I know you won't be able to do it with just an RF Sig Gen. You will need some instrument that can sense when when the cavity has reached his resonance frequency. I think checking the power is a good idea.

After you determine a method of detecting where there is a peak, you have to know how to change the frequency of your signal. I can think of three major methods:

1. Use an arbitrary waveform and change the frequency of the signal you are generating in real time. This will probably be fastest, but it is hardest to program. It will also have the LO in the signal.

2. Use the standard waveform mode and change the frequency of the signal you are generating. This is simple to do, will be slightly slower then above, but will still have the LO

3. Only output the LO, change the LO frequency. This will be slow, relatively easy to program, but since the LO is what you are using as your CW, you don't have to worry about LO leakage.

This is where you may have to do some benchmarking in order to understand which method will work best for you.

[zyb1003]

Hello, 

We also have other NI cards, such as Flexrio and its adapter, RF modulator and demodulators. We can detect the voltage of the pick up signal from the cavity and can manually change the LO frequency to match with the resonant frequency. 

The problem is how to autotune the frequency and let it match its dynamic resonant frequency very fast (within 0.1s). The solution you mentioned is like manually solutions I think. 

I believe someone ever done this before, because if cavity is superconducting, you have to use the FM loop. 

Thanks

[AgentAstronaut]

You should be able to automate the method I described. There are property nodes to change the frequency. I don't know how long it would take though. 

[zyb1003]

I know you use property node to change the frequency. I mean the automate process, since the resonant frequency might drift to either side (higher or lower), I can detect the amplitude of the pick up signal every cycle, but then? How to make it automate follow the resonant frequency point (with the largest pick up signal voltage)

[Henrik_Volkers]

Tuning a cavity or other resonances is a 'common' problem (Cs -clocks 😉 and ...)

Since it is hard to tune on the top of peak, mostly you tune at one slope near the peak.. (and swap the slope side from time to time to match the peak) ... should be tons of papers on that ...

another way would be to FM modulate the RF generator around the peak and monitor the output amplitude. Without going deeper (I bet there are fantastik papers on that with a lot of math :D) 

if you fm modulate with 1kHz around the peak, the AM ouput is 2kHz and the phase of the 2kHz output (in relation to the 1kHz signal) tell you the position (= frequency) of the peak. Now you 'just' have to add a DC offset to the  keep the peak around the center ... 

And I'm shure there are many more (and more clever) ways to do it, since the use of cavity resonace shift for sensors is a very active field of investigations 

[Henrik_Volkers]

SRF cavitys are cool stuff ... just had a look at that paper fig.27 and the text .....

I have the impression that you will need some cool equipment to run a SRF...

and that you will need some good clock sources 😉