Jennie, Sheila, Keita, Oli, Daniel, Camilla
After observed temperature dependence 78703, and ALS X beatnote intermittency dropping 78806, 78745 and this mornings PSL fiber realignment 78839, the beatnote was -30dB.
Jenne, then Sheila, then myself went onto ISCTEX (photo) and measured:
- Power out of fiber (before BS or PBS) = 0.5mW
- Power in laser path = 62mW
- Fiber is around 0.8% of lser beam power, agrees with when we blocked the fiber beam, ALS-X_FIBER_A_DC_POWER dropped 0.9% from 2.056 to 2.037 plot here
We started with a -60dB peak on the spectrum analyzer, using the amplifier's "-1dB Mon" channel. We plugged the spectrum analyzer straight into the BBPD and saw a very similar signal.
We then plugged into the amplifier's "+13dB " channel that's being used as the output. Beatnote ws clearer and around -40dB, photo.
Jenne and Sheila improved this by touching ALS-M1 and ALS-M2 (layout D1800270) we needed ot walk these two mirrors to make improvements. We also adjusted ALS-M3 as the DC power started to reduce on the BBPD. Once we got to -20dB on the spectrum anaylizer (in progress photo here), we plugged back in the main cable and using the ndscope beatnote, got back to -17dB. Although we are not happy, this was much better than the -30dB we started with and the PLL can lock. Leaving here for the day. Expect that if we had plugged back into the MON channel or straight into the BBPD we'd have seen -40 to -35dB, but we didn't...
We had a spare BBPD (EE shop ISC cabinet) but didn't try it. Daniel's view is that the coming and going bad signals is a sign or an amplifier or other electronics failing.
Reduced beatnote minimum to -40 so we shouldn't have overnight issues.
There is a power adjustment stage after the beam that measured 62mW of laser power. The nominal power on the broadband PD should be around 2mW. This is what the FIBER_A_DC_POWER is indicating as well. If 0.5mW exits the fiber, about 0.2 to 0.25mW should make it to the broadband PD. This would result in a 10-12% jump, when the fiber is blocked.
The expected beat note is sqrt(2mW * 0.2mW)=0.6mW, if the mode overlap is maximum. With a ~2K transimpedance gain and a 0.09A/W efficiency, we should see a beat note strength of -6dBm minus the overlap mismatch. The RF preamplifier nominally adds +13dB of gain. However the expected +7dBm would be about 2dB above the 1dB compression point of the RF amplifier.