intro

• 10may11: az swings done to measure tilt. New tiedown offsets computed
• 11may11: New tiedown positions computed from 10may11 were used to remeasure the tilt
• 12may11: 4 az swings using new td positions. the 11may11 swings were repeated to see if there really was a difference between  cw and ccw swings (there wasn't).
• At table of all 1,3az tiltsensor fits can be found here.
  

10may11: azswings before leveling (top)

• az 260-640 at .35 deg/sec.
  
• za=10 (cw and then ccw)
• za=15 (cw and then ccw)

• Top: pitch data
• black are clockwise data, red are ccw data
• black,red,green,blue are the residuals for the fits
• the fits value are show. The important parameters are:
• amplitude,phase for sin(az..)  term
• amplitude,phase for the sin(3az..)
• the first black trace jumps around at 265 degrees. this was when the first swing was getting under way. The oscillations must not have damped down yet.
• bottom: roll data

• The plots and table show that the pitch amplitude for the first spin was different from the rest. This probably was residual shaking when the azimuth arm first started to move. The first spin was not used when computing the average ampl,phase.

• The black points with + are the tiedown positions needed to level the platform
• top,center,bottom are for td12,4,8
• The horizontal axis has data from  the 4 az swings
• When computing the new tiedown positions, only the pitch results were used.
• Mapping roll to platform motion is more difficult than mapping roll.
  
• The dashed lines are averages:
  
• green: the average td position using all 4  az swings
• blue: the average td position using the last 3 azswings (1 had the jumps at the start).
• red: the old tiedown position.

11may11: aswings after leveling the platform (top)

• turpos:75
• tiedowns:
• old  :14.363,14.611,14.384
• new:14.133 ,15.530,15.822
• azswings: .35 deg/sec
  
• za=10 az 260-640 and back
• za=15 az 260-640

• Top: pitch data
• black are clockwise data, red are ccw data
• black,red,green,blue are the residuals for the fits
• the fits value are show. The important parameters are:
• amplitude,phase for sin(az..)  term
• amplitude,phase for the sin(3az..)
• the first black trace jumps az=50 to 140 degrees. I'm not sure why this happened.
• bottom: roll data

• the cw terms (za=10,za=15) have a different 1az term than the ccw za=10
• Not sure what caused this.
  
• .0031 is a pretty good leveling. the .008 could probably be improved.
• At least the terms got smaller !.

12may11: aswings after leveling the platform (top

• turpos:75
• tiedowns:
• new:14.133 ,15.530,15.822
• azswings: .35 deg/sec
  
• za=10 az 260-640  and back for 4 swings.
  

• Top: pitch data
• black are clockwise data, red are ccw data
• black,red,green,blue are the residuals for the fits
• the fits value are show. The important parameters are:
• amplitude,phase for sin(az..)  term
• amplitude,phase for the sin(3az..)
• the first black trace jumps az=50 to 140 degrees. I'm not sure why this happened.
• bottom: roll data

• The first 1cw swing has a pitch that is different than the other 3.
• The average pitch around .002 is pretty good and repeats 3 times.
  

15,16Feb12: azswings with "new td positions"

    The tiedown positions computed to level the platform ( 10,11 may11) were never installed as new tiedown positions since it would change the pointing. In preparation for a new pointing model (Feb or mar 12) azswings were done with the 11may12 tiedown positions to verify that they still gave a small amplitude to the 1 azimuth term. The setups were: xyf

• td position used:td12,4,8:  14.133 15.530,15.822 (same as 11may11).
  
• az speed .35 deg/sec.tilt sensor sampling at 5 hz.
  
• 15Feb12
  
• az 260->640->260 , dome=10, ch=stow
• 16Feb12:
• az 260->640->260->640->260, dome=10,ch=stow
• at az=512deg of first CW spin the turret motor was shut off.

• 1,3 az fits to each az swing
• Top frame is pitch fit for the swings
• Bottom frame is roll fit for swings.
  
• 15Feb12 data (.ps) (.pdf)  2 swings.
• The 1st CW pitch data has jumps in it. The CCW doesn't. The Roll data looks ok. Because of the jumps, i repeated the swings on 16Feb12
  
• 16Feb12 data (.ps) (.pdf)  4 swings
• top frame (pitch) . the first (black) az swing has jumps for the first part of the swing similar to 15Feb12.
  
• I noticed that he turret position was jumping by .003 degrees every now and then. At az=516 (156 on plot) I shut off the turret motor so it  couldn't move. After this this the jumps stopped for the next 3 swings.
• Back in may11 i was also seeing occasional jumps in the pitch data. The turret has probably being doing this for awhile. The turret has a master and slave motor that are phased to get rid of backlash. It may be that the adjustment is a little off.

Notes:

• ignore the swings with the jumps caused by the turret.
  
• The pitch 1az is about .0033. this is a little larger than the .0022 we had back in 11may11.
  
• The phase is jumping around since the amplitude is not well defined.
• The roll 1az is .008 . This is larger than the .003 measured back in 11may11 but still pretty good.
• The differences 11may11 and Feb12 may be that the tiedown positions may have moved a bit (when  encoders reinstalled). The values are still small enough that it isn't worth trying to make it any better.
  

Summary:

• There is an .02 deg amplitude 1az pitch term with phase at 110 deg az
• zero pitch is at 110 deg.
• az > 110 increases pitch so dome moves higher.
• Platform tilt has td12 low, td4,8 high, rotated by about 20deg east of north.
• The beam repair was symmetric except for the beam that broke. It has extra weight.
  
• The broken beam is on corner 12 east of north so the tilt of the platform corresponds to where the extra weight was put.
• New tiedown postion:
• 14.1327,15.5300,15.822 (td12,td4,td8)
• relative offsets from td12:0,1.3973,1.6895 inches
• Azswings with new tiedown postions:
• pitch 1az: .002 deg, roll 1az: .003
• some of the swings looked to have a cw,ccw difference. repeating the swings showed that this was not the case. We probably were seeing some temperature,rain variations.
• --> 16Feb12 .. the differences or jumps in the pitch reoccured on 15Feb12, 16Feb12.. On 16Feb12 after the first swing i turned off the turret motor and the jumps stopped. While the motor was on, i  could see the turret moving by about .003 degrees every now and then. This must have been changing (shaking?) the tilt sensor.l
  
• The new tiedown positions have not yet been installed since we need to also figure out the pointing correction.
• Feb12:
• swings redone and td positions still do a good job of getting rid of 1az term.
• The jumps seen in pitch are probably from the turret moving a little during the measurement.

page up
home ~phil