---

Intro

    On 26jun22 crosses were done on casA with the xband receiver. This was to check things after the receiver was reinstalled following the QRFH testing. While monitoring the data, a dip was seen in the total power trace of polA. It lasted for a little more than 5 minutes. It dropped about 4% and then returned  to the original value.

---

Setup and data taking

• cross pointing pattern:
• 10 sec cal On
• 10 sec cal Off
• 120 sec azimuth strip of .6 deg (great circle)
• 120 sec el strip of .6  deg
• Data taking:
• 7x172MHz bands with 1024 channels each
• 10hz sampling
  

---

Plotting the total power

    The   plots show the measured total power (.ps) (.pdf) :

• Page 1 has data from the 8222 MHz band
• Page 2 had data from the 8504 MHz band
• Black is polA, red is polB
• The narrow positive spikes are the 10 second calOn
• This is followed by a 10 sec calOff , the azstrip over the source, and then the el strip over the source.
  
• top Frame: Total power vs time.
• Data was normalized to the first 10sec caloff.
• Around 540 seconds polA starts to decrease, hits a minimum, and the comes back up.
  
• the dip lasted for about 400 secs (actually a little  longer since i've left out the move time from azstrip to el strip).
• There was a cal measurement at 740 seconds (during the dip).
• Middle frame: blowup of the vertical axis.
• The blue line marks the minimum of the dip (.966 8222 band, .960 8504MHz band)
• Bottom frame: cal Deflection normalized to first calOn - calOff
• The * mark the location and amplitude of the cal deflection.
• The cal diode is temperature stabilized so its output power should be constant.

---

Summary

    The dip could be a gain change, a noise level change, or both.

• with only a noise level change, the cal deflection should remain constant
• It doesn't
• A gain only variation should see the cal deflection change by the same amount as the noise floor.
• the cal deflection changes by a little more than the noise floor.
• Looking at the cal deflection during the dip, the noise floor is increasing during the calOn,calOff. This causes the cal deflection to be a little less than the noise floor deflection.
• The cal deflections show that both polA and polB gains are changing with time:
• polA: 2% in 1300 seconds
• polB: 3% in 1300 seconds
• these changes are probably electronic gain changes with temperature.
• Not sure what would cause  the 4% gain dip and then recovery after 400 seconds.
• The dip occurred in multiple frequency bands:
• The problem occurs before the 2nd mixing stage if1 -> if2
• It can be in the : lna,cables, postamp chassis, fiber optics, or udc..
  

processing:x101/220624/220626_xbpolAdip.pro