Check lbw polB cal cables.
The polB cal value for lbw jumped on 23apr04
and again on
04dec05. (by 20%). In both cases the jump occurred
for both diode 1 and diode 2 feeding polB. This shows that the problem
was after the relay in the post amp chassis that selected diode 1 or diode
2 (i guess it could still be that relay).
On 06mar06 we tapped/wiggled the polB cable inside
the post amp chassis as well as the polB cables that goes from the
post amp chassis to the dewar. While the cables were moved, the hcorcal
was being turned on/off at a 1 second rate. The interim correlator
was set for 4 25 Mhz bands centered at 1385, 1405,1425,a nd 1445 Mhz. The
receiver was set for linear polarization's.
The computation was:
The cal deflection was computed for each 2 second calon,caloff.
The total power of the cal deflection was computed by taking the median
value of each Caldeflection spectra.
For each 25 Mhz band the calDeflection total power was normalized to the
median value for the entire 400 seconds of data. This emphasized the changes
in the cal deflection. Since the cal deflection is only 30% of Tsys, the
noise on the measurements will be that of (Tsys + cal).
The plots show:
fractional change in the cal deflection while the cables were moved (.ps)
The top 4 traces are polB, the bottom 4 traces are polA. Each trace is
color coded by frequency (black 1385, red 1405, green 1425, blue 1445 Mhz).
The polA, polB traces have been offsets for display purposes. The vertical
scale is in units of the median cal deflection.
The dashed vertical lines delimit when something was moved:
black: We tapped on the dewar
red: tapped on the polB cal cable where it connected to the dewar.
green: wiggled the dewar side of the polB cal cable that goes from
the dewar to the post amp chassis.
blue: tap the connector where the polB cal cable enters the post amp chassis.
purple: wiggle the post amp chassis side of the polB cal cable that runs
between the dewar and the post amp chassis.
dark red: tap the polB cal cable inside the post amp chassis that goes
from the bulkhead to the diode select relay.
When the dewar was tapped, the polB cal deflection for the 1385 Mhz band
started to move around. It continued to wander after the tapping stopped.
This band includes the 1394 Mhz resonance in polB. If the resonance is
coming from the omt, it is a bit strange that the cal deflection (calOn-calOff)
would be affected since the cal is injected after the omt.
The red dashed lines where we tapped the polB cal cable as it entered the
dewar shows a change for all 4 polB bands. The cal deflection decreased
by about 1%. Unfortunately it also looks like the polA cal deflection
(bottom traces) also dropped a little during this same time period.
There is definitely no 20% change in the cal caused by our moving the cal
cables that are external to the dewar. So maybe the problem is in the dewar.