Not logged in
Sun Apr 27 05:09:30 2008
MC WFS Whitening turned on
hardwired on the MC WFS whitening
The MAX333A switches which choose between whitening and bypass on that board were in the bypass position
because the Xycom220 connections are not there. So the control switch gets +15V but there is no pull
down to set it to the whitened mode.
The least invasive (easiest) change I could do was to tie all of those inputs to ground. This pulls a few mA
through the pull-down resistors but is otherwise innocuous. All of these control lines come in on the A-row
of the P1 connector, so I was able to solder a single wire across all of them to ground them all.
The WFS2 board had a
blown electrolytic capacitor on the -15 V line
and so there was probably some extra noise
getting in that way. I couldn't find any extra SMD to replace it so I cut the legs off of a 22 uF polarized
tantalum and stuck it in there. Its even close to being the same color. I checked out the other caps, and they were all
close to 68 uF as spec'd. This one had luckily blown open and so didn't suck down the Sorensen and destroy everything.
Plugged everything back in switched the WFS servos back on. Looks good. Took before and after spectra.
In the plot:
GREEN: Open loop dark noise before changes
RED: Open loop bright (MC locked but MCWFS off)
BLUE: Closed loop, MC locked
BLACK: Dark noise after whitening
ORANGE:Closed loop after whitening
The cursor is at 16.25 Hz, the SOS bounce mode.
The I ran the new
script which uses
set the UGFs of the 4 loops to 4.01 Hz
We have in the past had problems with high WFS gains causing
instabilities with the CM servo
around 10-30 Hz. If this happens we should
just lower the gain by a factor of ~5.