update: 19apr21.. the glitches were created by the sband portion of the signal overdriving things.

---

Intro

   On 09apr21 9 crosses were done on the crab nebula. The setup was:

• 7  172  MHz Bands covering xband were taken (8100-9200 MHz).
  
• The crosses consisted of 1.2 deg long strips done in azimuth (great circle) and then elevation.
• The beamwidth at xb is about 10 Amin (FWHM)
• there were 7.2 beams across a strip
  
• Data was sampled at 10Hz giving 1220samples/strip and 166 samples across a beam
• PolA,B  data was taken separately. The two pols were combined when fitting the data.
• No cals were taken to scale the two polarizations
  

• The receiver chain still  had glitches and baseline variations, so the results should not be take as final.
  
• The crab is  an extended 7'x5'  while the beamwidth at xband is about 10'. No attempt was made to correct for the source extent.
• At xband the crab flux is about 612 Jy.
  

Processing the crosses

    For each cross a 2d gaussian was fit to the total power data for each 172MHz Freq band.

•  PolA and PolB where  averaged (no cals were taken to scale the data).
• The total power was taken across the entire band.
•  No attempt was made to remove rfi.
• bands with only partial rfi power (1st and last band)  still computed the total power over the entire band.
  

The fit used was:

•  z= A0 + A1*exp[-xp^2/sigxp^2 -  yp^2/sigyp2] + a7*deltaElevation
• xp,yp have been rotated to align the axis with the major  and minor axis of the ellipse (assuming the beam is ellipitical)
• Let x,y be the az,el great circle offsets from the center of the strip.
• x0,y0  be the fitted offset from the center (pointing error).
  
• xm=(x - x0), ym=(y-y0)
• xp=xm*cos(th) + ym*sin(th)
• yp=-xm*sin(th) +ym*cos(th)
• where th  rotated the axis to the major ,minor axis of the ellipse.
• A7 measures the slope of the baseline with respect to the change in the elevation encoder
• Note  thate x,y  are offsets from the center of the  strip. For the az strip , there is no change in the elevation offset.
  
• But during the az strip, the elevation encoder changes since the source is moving, so A7 needs to multiply the actual encoder value and not the  eloffset from the center of the source.
  
• The actual encoder used for the A7  coef. is the elEncoder - ElEncoderAtcenterOfmap

---

Plotting the results

Over plot  the azimuth strips for the 7 frequency bands from 1 cross (.ps) (.pdf)

• Plot the total power  vs azimuth offset for  1 cross
• top frame polA
• bottom frame polB
• The colors show the strips for the 7 172 MHz freq bands
• the dropouts are occurring over all freq bands at the same time.
• The dropouts in polB are much larger than polA
• PolB also has a baseline variation during the 120 seconds of data.

 Plot the azstrips data and fit at one freq band for all  9 patterns. (.ps)  (.pdf210409_crosses_crab_2.pdf)

• The previous plots showed that all freq bands have the glitch, baseline  variation pattern.
• Each pattern has 2 frames
• frame 1: tp polA (black) polB (red)  total power vs az offset
• frame 2: tp =(mean(polA+polB) green .  Blue line is the 2d gaussian fit evaluated for the az strip
• The  first few  patterns have a larger  polB baseline variation.
• pattens 1-2 do not fit very well because of  the baseline variation  of polB.

Plot the fit parameters vs elevation (.ps) (.pdf)

• Each color is a different  172 MHz freq Band.
• The rightmost + is the first cross, the leftmost + is the 9th cross.
  
• Page 1 Amplitude and pointing offsets
• Top: src deflection. The different freq bands are pretty close.
• The src amplitude has been divided by the fit constant value  (Tsys)
• The first 2 patterns (on the right) did not fit  well. Their amplitude varies more
• If  the crab  is 612Jy (and we ignore that it is extended) then SEFD=6120 Jy  (this is not a good measurement of the sefd since the source is extended),
• Middle: azimuth pointing error.
• There is lots of variation (probably because of the baseline variation
• The median value is .326 arc minutes
• Bottom:elevation pointing error.
• this data was taken with model 3 (2015) installed.
• This data was taken with a .35 deg offset always add to the elevation. Looks like things have changed since 2015.
• the median value is .213 Amin
• Page 2: Beam widths and rotation angle
• Top: azimuth beam width  (after rotation of the ellipse). About 10 Amin FWHM
  
• 2nd:  elevation beam width (after rotation). Again about 10 Amin FWHM.
• 3rd:  BeamWidthAz/BeamwidthEl  .. to see if the beam is elliptical.
• We will have to wait till we stabilize the receiver before saying anything about this.
• Bottom: rotation angle to rotate x,y axis to the major minor axis of the ellipse
• Values cluster about 0 , but the fits are not stable enough to draw any conclusion
  

---