New sbn Tx horn

may06

Links to plots:
system performance with all of the data combined (old, new horns, all sources) (.ps) (.pdf)
compare the gain and beam width of the horns (.ps) (.pdf)
compare the pointing errors for the runs (.ps) (.pdf)
compare the focus curves for the two horns (.ps) (.pdf)

Sections:
Intro
Calibration runs
Focus curves
Conclusions

Intro

A new sbn transmit horn was installed on 04may06. It was designed to have a smaller edge taper on the tertiary (to increase the gain). The increase in temperature from the spill over does not matter for the transmit horn (although you don't want to melt anything with the spill over of the transmitter!). To test the horn, a room temperature receiver was placed on the sbntx system before and after the horn was replaced. A cal noise diode was included for calibration. Data was taken on the following dates:

25apr06 focus curve sbnrcv (old horn) with B0745+101
28apr06 focus curve sbnTx (old horn) with B2251+158 (3C454.3)
29apr06 calibration run using sbnTx (old horn).
30apr06 calibration run using sbnRcv (old horn).
04may06 calibration run using sbnTx (new horn).
05may06 calibration run using sbn Tx (new horn). Repeated 04may06 but no cal fired.
10may06 focus curve sbn Tx (new horn) with B2251+158.

Calibration runs.

Data was taken on 3 sources. The data taking started at 16:00 to 16:30 and went to about 20:00.
The noise diode for the sbnTx measurements had an ENR of 15.6 db. It was coupled into the system with a 22.4 db coupler. These number would give a cal temperature of 62K. For the computations i used 60 K. Since we were comparing before an after measurements the important thing was that the cal was constant between the measurements (it's absolute value was not that important). The receiver for sbtx had to be installed and removed a number of times during this measurement period (there were ongoing sband runs). During the removal, the cal cable was not disconnected.

The new horn was designed to decrease the edge tape of the illumination pattern on the tertiary. This would increase the gain, decrease the beam width, and increase the Tsys (more spill over for the receiver). The gain should increase by the square of the beam width decrease (g proportional to 1/beamwidth^2). The system temperature change would not be obvious with the receiver we were using. The Tsys was about 110K while the temperature in the dome is 250-300 K. With a cooled receiver this would stand out better.
The plots show the results of the measurements:

The first set of plots show the system performance with all of the data combined (old, new horns, all sources) (.ps) (.pdf):

Page 1: gain, Tsys, Sefd, and Beamwidth. The sbTx (black,grey +, purple box, -) gain and Tsys may be a little high (since the cal value is not that accurate). The Tsys for sbTx is about 105K. The sefd's do not depend on the cals.
Page 2: The beam efficiencies depend on the cals so the sbtx values can be off.
Page 3: The pointing errors.

The second set of plots compare the gain and beam width of the horns (.ps) (.pdf):

Page 1 Gain: Each panel is a separate source. Black is rcv Horn, red is old tx horn, green is new tx horn. The offset between the tx and rcv horns is probably coming from the tx cal value being a little off. The increase is gain between txOld and txNew is real (since the same cal used for both). The may05 txNew data was not included since the cal diode didn't fire.
Page 2 beamWidths: This includes the 05may txNew horn data. The beam width decreased by about 5 asecs from the oldHorn to the new horn. This is the average beam width in the az, za direction. I checked the beamwidth difference (in the az,za directions). This remained the same before and after the horn change.
Page 3 Ratios: The txNew horn data was interpolated to the azimuth positions of the txOld horn points and then the ratios were computed. The top plot is the ratio GainNewTxHorn/GainOldTxHorn. The increase is 3 to 4 %. The bottom plot is the inverse ratio of the beam widths (BeamWidthOldHorn/BeamWidthNewHorn). There is a 3-4% change in the beam width. The 04may05 txNewHorn measurements were repeated on 05may05. The beam widths from the two txNew horn days track pretty well.

The gain should have increased as the square of the beam width. It looks like it is linear.

The 3rd set of plots compare the pointing errors for the runs (.ps) (.pdf):

Page 1-2 ZaErr: page 1 is the za Error vs za while page 2 is za Error vs az.

The Txnew horn has a 10 asec offset from the txOld horn for B0758+143 rising. They agree pretty well setting. The second source B1040+123 only had a 5 asec difference. The difference was also there on the 2nd day of the sbTxNew measurements. I checked the tiedown tensions and they didn't look like they lost tensions. It may be that the txOld horn data is the one that stands out (even if it is closer to 0).
There is a ramp in za error vs azimuth. This could be caused by a platform tilt or a time error. I checked the clocks and they looked ok.

Page 3-4 azErr: The azimuth error vs za and azimuth.
Page 5-6: A vector plot of pointing errors on the dish. The length of each arrow is proportional to the pointing error (1 tick mark is 5 asecs). The direction is the direction of the error. Each circle is a 5 degree step in za.

Focus curves

Focus curves were done using the old sbn tx horn (28apr06) and with the new sbtx horn (10may06).
The plots compare the focus curves for the two horns (.ps) (.pdf):

Top: Platform focus height for the old horn (red) and the new horn (black). The sold lines are the source rising. The dashed lines are the source setting. The vertical scale is in feet.
Bottom: The newHornFocus - oldHornFocus distance. The vertical scale is in inches. The focus height for the new horn was interpolated to the measured azimuths of the old horn and then the difference was taken. The large difference at za=2 is probably due to the rapidly changing focus error as the azimuth swings around at transit. The interpolation probably didn't work too well here. The new horn looks like it positioned .2 to .4 inches closer to the dish than the old horn (it needs to move farther from the dish to focus).

Conclusions:

The average beam width decreased by 3-4 % when the new txHorn was installed.
The gain increased by 3-4% when the new horn was installed. If the sidelobes remained constant then the gain increase should have been closer to 10 %.
The first sidelobe height did look a little larger for the new tx horn.
The new Tx pointing error for the source B0758+143 rising had a za error of about 10-15 asecs. This was not seen in the Tx old horn. It was also not this large for the rcv horn.
We may have a platform tilt (relative to when the last model was made).
The new horn is focussing .2 to .4 inches farther from the dish than the old horn. This means that it may be positioned closer to the dish than the old horn.

processing:x101/sbn/newhornmay06/comparehorns.pro, compareFoucCurve.pro

home_~phil