Project a3106 (run as a2934) (Barbara again...) reported baseline problems with lbw polB on 26mar17 .  On 28mar17 we took some data with lbw and the interim correlator to check the receiver stability.

26mar17 a3106 data

a3106 data with polb drifts (.ps) (.pdf)

• Setup:
• lbw, linear pol, freq around 1390.
• 5 min onoffs with cal using interim correlator
• 12.5 MHz Bw, 2048 channels.
• filters used:
• 1320 hipass  (lbfb 7)
  
• 1370 hipass with 1230-1470 bandpas (lbfb 1 3)
• Top frame: total power from 5 onoff pairs (file corfile.26mar17.a2932.2).
• black is polA, red is polB.
• the narrow jumps are the cal.
• On the start of the 5th pair (around sample 2490) polb started to wander around.
• Bottom frame: blowup of 5th on,off pair showing polB wandering around.

    The Big change from on/off pair 4 to pair 5 is that the lbw rf filters were switched:

• from 1320 hipass (lbfb 7)
• to   1370 hipass and 1230-1470 bandpass (lbfb 1,3)

    It looks like the filter switches are not making good contact when switching.

28mar17 interim cor test data

    On 28mar17 we took some data with lbw and the interim correlator.
The setup was:

• lbw , linear polarization, cfr 1397 MHz
  
• interim correlator , 1 sec integrations, 25 MHz bandwidth.
• Telescope was being used by maintenance. It was mostly stationary. toward the end they moved it a little.
• We setup a filter set, adjusted the power and then started taking data.
• During the scan, we switched to another set of filters (causing a glitch) and then back to the original set. If thing were working correctly, we should come back to the same power levels..

The plots show the total power vs time data during the filter switches (.ps) (.pdf)

• the data comes from the 4th scan of file corfile.28mar17.x101.3
• we started with  the 1370 MHz hipass and 1230-1470 band pass filters in (lbfb 1 3)
• the switch sequences flagged in purple were:
  
• 2190 : (1370hp,1230-1470bp) -> (1320 hp) -> (1370hp,1230-1470bp)
• 2730 : same sequence
• the level changed after each switch sequence for both polA and B.
• They too 10's of seconds to approach their original values.
• The other jumps were similar filter switches.
• the general baseline variation is probably from the maintenance crew moving the telescope.
  

We also tried switching between:

• 1230-1470bp -> 1120-1800bp -> 1230->1470bp
• this also showed a long recovery time for polb.

Summary:

• It looks like the switches in the lbw filter bank are occasionally not making good contacts (this could be the through switch in use, or one of the 50 ohm load switches of the unused filters).
  
• We need to bring the lbw rf filterbank down to the lab and test the switches.
  

processing: 170328/a3160_lbwjumps.pro,lbw_filters.pro