Strange vertical 50Hz noise in one BPM

[Laura]

Hello,
I am struggling with this since July: we have one BPMs that sees some strange noise only in the vertical direction. The frequency of this noise is 50 Hz.
The problem is that this BPM is in the FOFB loop and the noise is then spread around the whole sector.

We have:
- Checked and changed the associated corrector (but the noise frequency was already telling us that it was not the guilty one)
- Changed the Libera (even if the noise was only in vertical)
These modification provided no results.

Moreover this noise is not always present. We are sure that it is seen by this particular BPM since we removed corrector and BPM from the FOFB loop and we could see the noise in the vertical BPMs data even when it was out from the feedback. The other BPMs of the sector did not show anything.

I can add that this BPM is located in the middle of a standard sector (no IDs or cavities close by).
I also want to stress that this noise is observed only in the vertical direction.

Has it happens to some of you? Do you have some suggestion.

Thanks in advance,

Laura

[Benoît Roche]

Hello Laura, I hope you are doing well.

I don't remember having a 50 Hz noise perturbation that was not a real perturbation on the beam, here.

Nevertheless, can you describe a little more your graphs? In particular we can see jumps on the second graph, it is not clear to me if those are related to the noise you have.

[Laura]

The second graph is what we see when both BPM and corrector were out of the FOFB loop: we could still see the noise and, since they were not in the loop, the jumps were not corrected (i guess). If you zoom at the jump you would see the same 50 Hz pattern (as in the third plot, unfortunately I don't have a zoom of this specific graph).
The first plot is how the noise appear usually when the BPM and the corrector are in the loop.
I think this is a noise that somehow enter in the BPM and then the FOFB try to correct and propagate it through the sector.
The last plot is an FFT of the noise. The 50Hz line appears broader.
What I don't understand is that it appears only in vertical... If it was some random noise leaking in the buttons/cables it would be in both planes.
I am quite lost 

[Benoît Roche]

It is indeed quite puzzling. Also it seems that it is not a pure 50 Hz sine wave perturbation, but rather a square signal with varying duty cycle. This suggests that it is not just a coupling with mains power supply.

Can you try to capture the signal seen by each channels (Va, Vb, Vc, Vd) to see if it is present on all of them, or just two? Does the sum signal also have this behavior?

[Ji-Gwang Hwang]

 Dear Laura

It is a bit different issue but I have observed the 50 Hz ground motion in our optical table with a vibration sensor. We found that the source was small chillers installed onto a vibration damped carriage. The 50 Hz noise disappeared (please see the attached figure) when the chillers are off for the summer shutdown.
It may possible to check the motion with a vibration sensor. Then, it would be clear that the motion is real or artefact.

Best regards

Ji-Gwang Hwang

[Laura]

Hello Ji-Gwang,
Indeed we also suspect some cooling system but we have not been able to identify it. We put some accelerometers in the tunnel on the block and on the girder and in both case we see this strange frequency but we could not identify the source, moreover the frequency is there either we observe or we do not observe vibrations in the BPM data. As you suggest we can try to do some tests with accelerometers during shutdown switching off each of the element to see if the movement measured with acceleromters stop in some case.
Thanks!
Laura

[Ji-Gwang Hwang]

 Hello Laura,

It is good to hear that you did the test with accelerometers. In general, the sensitivity of the accelerometers (depending on the model and frequency) is an order of magnitude better (less than 100 nm) than the BPMs (~ 1 um). This could explain the discrepancy between the two monitors.

If the vibration amplitude is changing daily (especially more in summer), the cooling system would be the source since the water flow/pressure is variating to fight against the temperature raising in the tunnel or circumstance. But I am not sure whether the cooling water could drive the 50 Hz noise. In my case, the compressor of the chiller was working with that frequency which is equivalent to the frequency of the main powerline in Germany.

Best regards

Ji-Gwang Hwang

[Laura]

Well, the step I observe with BPMs is 4 um so I should be able to see it with accelerometers!
Also, sometime it is a jump: it does not go back to the previous location...
I will keep you updated!

[Benoît Roche]

Just an idea: maybe it could be an "impurity" which is in the vaccum chamber between the button and the vaccum chamber a create a non-infinite resistance. Sometimes the impurity creates the short-circut, sometime it does not.

It does not explain everything (why the perturbation is only in the vertical plane? why 50 Hz?), but it is easy to check with only a multimeter (without beam).

[Laura]

Good idea, I will try next week (shutdown)!