plots:
the measured pointing errors before and after the failure (.ps) (.pdf):

The Failure.  (top)

---

Checking the pointing:

• page 1 Data by Date: Black is B0340+048, red is B0518+165. This shows when the data was taken.
• 1st Hour of day (AST) when data was taken. There was no data during the hot part of the day when the tiedowns could go slack.
• 2nd Receiver used. This shows which receivers were used on each date.  The receiver numbers are: 5-lbw, 8-sbhi, 12-sbn.
• 3rd za errors vs date. The error spread for each day is the variation of the error with az,za.  The large values for B0518+165 occur at low za (where the model is not very good).
• 4th az errors vs date.  The rightmost values for B0340+048 (14nov06) have a positive offset.
• Page 2 Sky pointing errors vs az and za: Black is B0340+048 after encoder failure. Green is B0518+165 after the encoder failure. Red is B0340+048 before the encoder failure. Blue is B0518+165 before the encoder failure.
• The top 2 plots are za error vs za and az. The za error for B0340+048 has a little more of a horse shoe opening vs za than the pre data but the average value is about the same.
• The bottom two plots are az error vs za and az. The Azimuth error for B0340+048 has shifted by about 20 arc seconds from before the failure.
• Page 3 az,za encoder pointing errors vs az,za: I converted the errors on the previous page to encoder errors. The za errors remain the same. The az errors get divided by the sine of the za. I wanted to see if the az error was an encoder offset or an offset on the sky. It's a little hard to tell from the plots. The az error vs za for after the failure now has more slope vs za than the pre values.
• Page 4  az,za sky errors after correcting az encoder by  -80": I subtracted 80asecs*sin(za) from each of the azimuth errors after the encoder failure and then replotted the data. The high za portion of the az errors for B0340+048 do not line up very well.
• Page 4 az,za sky errors after correcting for -20" sky error: I removed 20" from each az error after the failure. The after az errors line up better for B0340+048 but the az errors for B0518+165 at low za  are now off by 20".

---

Summary:

• B0340+048 has an az error that is 20 asecs larger than before the encoder failure. B0518+165 is near the center of the previous errors.
• Changing the sky azimuth position for B0340+048 by -20" does a better job of aligning the After az errors with the pre az errors than a -80" azimuth encoder offset.
• Changing the sky az position by -20" moves the After az errors for B0518+165 about 20" away from the pre failure errors.
• The az encoder is about 60 feet from the center bearing. An 80" azimuth encoder offset is .28 inches (7 mm). The scribe mark, pointer alignment  should be able to align the azimuth better than this.
• We need to take some more data to decide whether it is really an encoder offset or a sky azimuth offset. We should also go upstairs and verify that the pointer/scribe mark is aligned.