Telescope pointing
jan02
Links:
Description of the model parameters
The model parameter values by receiver.
Pointing models and measurements.
Steps in making the model
History.
Misc pointing info
Compute pointing error
for various platform motions
platform
motions.
motion limits:
The model
parameters: (top)
The model correction is a function of azimuth and
zenith angle. It has 13 coefficients in azimuth and 13
coefficients in zenith angle. There is also a zenith angle
encoder table (for azimuth and zenith angle) that is used to correct
for irregularities in the zenith angle tracks. To use the
model correction you:
 Convert from the astronomical coordinate system (ra,dec) to
az, za.
 Evaluate the model correction function at this az, za.
 Add the model correction to the az,za computed in 1. Since the
model correction is in great circles, you must divide the
azimuth model correction by the sine of the zenith angle
before adding it to the azimuth value computed in step 1.
 Evaluate the encoder tables for az,za at the current za and
add these values to that in 3.
The model parameters definitions are listed below.
There is a separate set for azimuth and za. The values are great
circle arc seconds.
param name 
An 
constant offset 
A0 
cos(az) 
A1 
sin(az) 
A2 
sin(za) 
A3 
sin(za)^2 
A4 
cos(3az) 
A5 
sin(3az) 
A6 
sin(zabalance)*cos(3az) ** 
A7 
sin(zabalance)*sin(3az) ** 
A8 
cos(2az) 
A9 
sin(2az) 
A10 
cos(6az) 
A11 
sin(6az) 
A12 
** the balance value is 9.15012 deg. This is the greg za that
balances the feed arm when the carriage house is at the stow
position (8.183 deg).
The model
parameter values by receiver. (top)
The values for the model parameters for each receiver are
listed in the files below. The file format has the azimuth
correction in the left most column, the zenith angle correction in
the middle column and then the parameter label in the 3rd column.
The top of the file contains the most recent file with older models
following. A # in column 1 are comments and should be ignored.
327
430dome
430ch
610
lbw
(1.2 to 1.75 Ghz)
alfa
(1.22 to 1.52 Ghz)
sbw
(1.8 to 3.1 Ghz)
sbn
(2.3 to 2.4)
sbh
(3 to 4 )
cb
(4 to 6)
cbh
(6 to 8)
xb
(8 to 10)
ao observatory position
Pointing models
and measurements: (top)
190918:
pointing errors for summer 2019 from platform motions.
190812:
az enc1 jumps by 1 tooth. Fixed on 11sep19
171122: model 19 after hurricane
maria.
170414: az encoder 1 fails, replaced(water in encoder), reset with
pointer
170411: az encoder 1 fails, reset, reset with pointer
01jan17: leap second added . used data 01jan17 thru 01dec17.
02oct16: azimuth encoder 1 fails, replaced. aligned with pointer at
270 degrees.
25may16:
controlling the platform height with the tiedowns and distomats.
12jan16: az encoder 1 fails. replaced. aligned with pointer at 270
degrees.
07sep15:
using 3 distomats to measure the platform height.
09aug15: pointing errors at noon (for venus run)
22jun15: refit polynomial for utcut1
after leap second. 01jul15 > 30sep15 and 01oct15>15jun16
> be sure pointing program gets restarted
sometime after 01oct15.
16jul14:
refit polynomial for utcut1
12mar14:
calibration runs made after the m84 cable repair. checked
pointing.
21oct13:
refit polynomial for utcut1
10sep13: updated sbw,cb models
after reinstalltion following ao19 removal.
05jul13:
refit polynomial for utcut1
13jan13:
updated pointing offsets using juldec12 data.
18jun12:
Pointing error as we move the platform vertically,.
06jun12: pointing
error as we tilt the platform using td12.
05mar12: model 18 after beam repair
(a year late!)
28feb12: refit polynomial for utcut1
mar12 thru jun12
15,16feb12: tilt sensor azswings
to verify "new..11may11" td positions still good.
24oct11: fitting the
C0,1az,3az errors for alfa.
08sep11:
sbn,alfa
pointing check after alfa installed.
30jun11: pointing corrections
needed after leveling the platform.
12may11: compute new td positions to
level platform.
19apr11: refit polynomial for utcut1
apr11 thru apr12
14mar11: correcting pointing
offsets of 800 Mhz rcvr
09feb11: pointing errors after the
beam reinforcement completed.
22nov10: pointing errors octNov10
with 1,2,3 az fits (after u2u2l3 repair)
03jun10: updated all models to 1,2,3 az
fits.mar10jun10 data.
08feb10: refit polynomial for utcut1
feb10 thru feb11
12dec08:
updated utcut1 fit for leap second.
20jun08: fit 1az and 3az term to
pointing errors to update model.
18jan08: pointing model
gradients.
16dec07:
model
17 after the painting project.
03may07:
modeling the azimuth pointing errors.
30mar07:
pointing
checks
during
the painting setup.
11dec06:
fit
1
az
and sin(za) terms to pnt errors and update the models.
10dec06:
refit
UT1

UTC (dec06 thru jun07).
28nov06:
nov06 pointing checks after az encoder failure.
27jan06:
check
the
azimuth
error after the nov05 encoder failure.
15mar05:
Model16
for the carriage house made.
27may04
:
pointing
error
as we lose tension in the tiedowns.
15may04
:
Model15
after
kevlar cables, alfa installed, and the dome side rollers
tightened.
14oct03
:
azimuth
encoder
jumps.
mar03
:Model14
after
shimming
and dome move.
15sep02
:
azimuth
encoder
jumped.
jul02
:
measuring
dome
tilt,roll while the tiedowns track the average platform height.
feb02
:How
many
samples
are needed to transfer the model between receivers.
jan02
:Model13
after
reflector
adjustments.
01aug01
:pointing
and
gain
changes from tertiary motor installation.
24may01:
pointing error, gain before and after teritiary platform
installed.
feb01
:
Model12
after
1st reflector adjusted.
apr01
:
platform
motion
as we pull on the tiedowns.
jan00
:
azimuth
encoder
rack gear error.
Pointing references pages:
apr07
Converting
platform
rotation
to sky rotation
apr07:
platform
motion,
gr
pointing error as we move ch 2 to 20 degs za
jul02 : tiedown ramps
while tracking do not change pitch, roll.
 model 0 oct97 
 used 4 term model: offset, cosaz,sinaz,sinza.
 sband tx receiver with turret at 60.54 degrees.
 Sources from AO catalog (chris salter).
 model 1 dec97 
 10 terms. added sin(za)^2, sin(3az), cos(3az),
sin(zabal)*cos(3az) , sin(zabal)*sin(3az), sin(za*57).
 model 7 jan98 
 13 terms. removed sin(za*57), added sin(2az), cos(2az),
sin(6az), cos(6az).
 model 8 apr98 
 switched to sband rx receiver (sband narrow).turret position
60.65 > 75.0.
 model 9 may98 
 model 9 sept98 
 hurricane Georges hits the observatory.
 model 10 jan99 
 dome realignment.
 sband turret position moved 75 > 76.6 degrees.
 model 11 feb00 
 rails reshimmed.
 Horn positions in turret resurveyed by lynn baker.
Sband horn moves 76.6>76.83.
 Use tiedown tracking of the temperature (.2 inches/degF)
during model.
 dec00 
 broken wheels on dome. Dome jacked up and wheels replaced.
Jacking beam on azimuth arm 20 feet from main bearing (weight
1000>2000 lbs ??).
 model 12 ,feb01
 switched to idl for processing.
 Sampling went 100ms > 10ms so time constant doesn't
affect measurements as much.
 Model data taken only during the day (10am to 6pm).
 The reflector surface above the main cables had just been
adjusted.
 Started including NVSS sources.
 Tiedown tracking of average platform height (using the
distomats) while the model was taken.
 Pointing preprocessor running to smooth out motions
 This model was never installed and used.
 19/21sep01 sband
wide receiver back on turret.
Sbw receiver back up on turret after over a year of "not
working". Did 2 strips on B1442+101 using turret scans and
heiles calibration patterns. see
plot
Changed za model offset 70 >90 asecs (now the same
as the sbn).
(see x101/010921/doit.pro)
jan02, model13
 Model made using sband calibration data (heiles scans) from
Sep01 thru jan02 and from turret scans in dec01 and jan02.
 The reflector adjustments finished in sep01.
 The model was made relative to model 11 (model12 was never
installed).
 The tiedowns were running for the entire time.
feb02 model13 installed for xband pointing
22mar02 model13 installed for sbh with sbn offsets.
aug02 felipe tightened the small rollers on
the dome right side. now about 1/32" play. was 1/4 to 1/2"
15sep02 az encoder jumped. added offset of 1446 to the az
term of the sinza coef. did it for all models except carriage
house.
02dec02 the new trolleys weigh about 3000 lbs more than the old
ones. Pointing is changing vs za because of it.
01mar03 model 14
data taken 01mar03>03mar03 after shimming
elevation rails, and moving dome up 1.65 inches.
04mar03 model 14 installed with old horn offsets.
7mar03. verify of other horns, offsets added to model14 for
each receiver 18:00.(except 327,610,430 needs to be done).
14may04: model 15 completed installed sbn. begin verify other
horns.
27dec04: cbh. model offsets (az,za) incremented by (.28,
3.8) asecs. from x102 data.
15mar05: model 16 for carriage house made.
16dec07: model 17 installed sbn.
apr08: alluminum work platform reinstalled top chord azimuth,
ch side 131 feet from center
 16apr08: work platform 500 lbs
 23apr08: access ladder for platform 200 lbs
 20may08: ladder cage 80 lbs.
20jun08: updated pointing model with fits to 1az, 3az terms.
01jul08: subtracted 4 asecs from xb za constant term for
bania's run... (other recievers not modified).
12dec08: updated utcut1 fit for
leap second.
14mar11: subtracted [5,25] (az,za) from 800 mhz model to get
rid of large offsets in za.
17jun11: sbn model switched to one with 1az,3az updates need
to use tietrktest to get correct td positions
21jun11: sbh model switched to one with 1az,3az updates. need
to use tietrktest to get correct td positions.
26jul11: modelSB, modelSBH back to models without 1az,3az fits
(since tiedowns not working so can't move to tietrktest. (new
model updates in modelSBHnew,modelSBnew).
24oct11: updated
constant offset for alfa (after reinstallation in sep11).
28feb12: refit polynomial for utcut1
mar12 thru jun12
05mar12: model 18 built.. but
not yet installed
03may12: updated offsets for alfa after reinstallation.
13jan13:
updated pointing offsets using juldec12 data.
05jul13:
refit polynomial for utcut1
10sep13: updated za
offsets forsbw,cband after reinstallation following ao19
test.
21oct13:
refit polynomial for utcut1
16jul14:
refit polynomial for utcut1
20sep17: hurricane maria
30nov17: model19 installed.
190812:
az enc1 jumps by 1 tooth. Fixed on 11sep19
model
pointing residuals [asecs]
model

azerr rms
noEncTbl

zaerr rms
noEncTbl

total rms
noEncTbl

azerr rms

zaerr rms

total rms

# of points

oct97 mod 0








dec97 mod 1








jan98 mod 7








apr98 mod 8








may98 mod 9

8.02

12.43

14.79

3.82

2.37

4.49


jan99 mod 10

8.86

12.59

15.39

5.40

3.05

6.20


feb00 mod 11

7.78

6.21

9.96

5.68

3.45

6.65


feb01 mod 12

7.57

5.77

9.52

5.61

3.65

6.69

1603

mar03 mod 14

8.98

6.65

11.18

5.38

3.48

6.41

1371

may04 mod15 
7.08

6.07

9.33

4.20

2.41

4.84

501

mar05
mod16 (ch) 
16.33

13.63

21.27

7.52

8.79

11.57

312

dec07 mod17

8.27

6.30

10.40

4.39

2.99

5.31

480

feb12 mod 18

7.46

5.80

9.45

4.18

2.13

4.69

557

nov17 model 19

7.23

6.02

9.41

4.76

3.06

5.65

435

comments:
 Model 9 to model 10 rms jumped up by almost 2 asecs. Hurricane
georges was between these two date. The encoder tables don't
seem to fit as well. The roller bearings that guide the dome
up/down the elevation rails were readjusted. I need to find out
if this occured during this period.
 Model 10 to model 11 no encoder table values jumped down in za
by about 6 asecs. This was because the reshimming of the
elevation rails brought them closer to the model fit. The
overall fit with the encoder table included actually got worse.
 Model 12 was never installed
 Model 13 was used for a few newere receivers
 Model 14 was made after shimming, and moving the dome.
 Model 15 after kevlar cables and alfa installed. The roller
bearings for the dome had been tightened on one side to force
the dome to track along that line (the other side was left
loose). It looks like this brought us back to the rms that we
orignally had in may98 when both roller bearing sides were
tight.
 16dec07 pointing model 17. finished receiver offsets in jan08.
 21jun08: updated pointing model from 1az, 3az fits to pointing
runs.
 21mar10: software limits 3.5 to 12 for broken beam.
 22oct10: software limits 1.06 to 16 after first round of
reinforcement.
 30nov17: model19 made after hurricane maria (20sep17). The
dish had not yet been repaired.
Misc pointing
info (top)
crossing
midnight at the end of the year and leap seconds..
The telescope model.
motions
and angles in the focal plane.
Fitting the data
Verifying the model
A
description of the pointing program and the structure it loads in
scramnet.
home_~phil
G