Reports 1-1 of 1 Clear search Modify search
MIF (General)
kazuya.yokogawa - 2:00 Friday 02 November 2018 (6818) Print this report
[ALS PLL] The strange noise was the free-run frequency noise of the PROMETHEUS laser

[Izumi, Yokogawa]

Yesterday, it was confirmed that the broadband structure from 1 Hz to 100 Hz on the PLL feedback spectrum disappeared as we swiched off the KOACH air system (log6796), but there was flat noise after around 10 Hz.

We concluded that these noise is freerun frequency noise of the PROMETHEUS laser.


what we did:

  1. made a comparison of the relationship  between the measured PLL feedback spectrum and the spectrum obtained from the noise budget.
    • Here is the plot of these spectra. The pink line is the measured PLL feedback spectrum and the gray line is the spectrum obtained from the noise budget culculation. We assumed that the free-run frequency noise of the PROMETHEUS is 1e6/f as a typical value. These are matched at 1 to 10 Hz, but there are strange noises after 10 Hz.
  2. tried to do PLL by using the low noise VCO instead of the laser frequency and evalueted the sensing noise of PLLX loop.
    • We saw that there is no relationship between the strange noises and the sensing noise of PLL.
  3. made a comparison of the relationship between the PLL feedback spectrum and the PDH feedback spectrum
    • If the PLL works well, the free-run frequency noise of the PROMETHEUS laser and the sensing noise of PLL should be seen in the PDH feedback spectrum multiplied by 1/(1+G) and G/(1+G) respectivelly. (G is the olg of PLLX loop.) Here is the screenshot of these spectra. We indeed see that the PDH feedback spectrum is limited by the PLL sensing noise around 100 Hz and the free-run frequency noise of PROMEHTEUS laser is suppressed by the factor 1/(1+G) after 1 kHz. Therefore, the strange noise in the PLL feedback spectrum is the freerun frequency noise of the PROMETHEUS laser.
Comments to this report:
yutaro.enomoto - 11:05 Friday 02 November 2018 (6821) Print this report

Hi Yokogawa-kun,

Which channel did you use as a monitor of the feedback signal of PLLX? K1:ALS-PLLX_SLOW_DAQ_OUT(_DQ)?
If so, we haven't implemented an anti-whitening filter for this channel, which means that value of this channel is different from the voltage of the feedback signal in the real world, by a 2nd order whitening filter (two zeros at 10Hz and two poles at 100Hz, DC gain of 2).
We need an anti-whitening filter same as the one implemented in the filter module K1:ALSPDH_PDHX_SLOW_DAQ. 
(I implemented such filters only for K1:ALS-PDH{X|Y}_SLOW_DAQ_OUT. Sorry for this confusion.)

And my implementation of the anti-whitening filter by eyes to the spectrum gives me less-structured 1/f shape. This must be free-run noise! 

Images attached to this comment
kazuya.yokogawa - 12:33 Monday 05 November 2018 (6850) Print this report

Hi Enomoto-san,

I ploted the free-run frequency noise of the PROMETHEUS laser considering the effect of the anti-whitening.

The blue line is the frequency noise measured last Thuersday and the orange line is the frequency noise considering the anti-whetening. This plot looks like 1/f shape. Also, the green line is the free-run frequency noise of the PROMETHEUS laser measured on Oct. 11th with Enomoto-san.

yutaro.enomoto - 18:20 Monday 05 November 2018 (6852) Print this report

Great!
(By the way, the plot has disappeared at some point.)

To avoid further confusion, I implemented anti-whitening filters for CMSBs named as PLLX, PLLY, Summing node, and CARM servo.
Then updated the safe.snap of k1alspll

Search Help
×

Warning

×