Issue:  Centering performance in Declination has deteriorated over recent sessions

Issue is evidenced by

a) Centering Chart showing final pointing errors with a clear overall bias to the north.



and

b) specific examples such as NGC 7501 (Az 190, Alt 40) with pointing errors of  0.48' 2.68' 2.28' 1.93' from the 4 locate frames used, with the pointing error after slew (W -0.32', N -0.36') being smaller than the pointing errors after any of the 3 subsequent centering attempts including the last attempt (W -0.12', N 1.92').  Pointing error in Declination jumped from S 0.36' on initial slew to N 2.62' following an initial 0.36' N centering jog.  Pointing error in Declination only came down slightly to N 2.21' following a S 2.62' jog,  and only came down to N 1.92'  following a further S 2.21' jog.  

The problems are in keeping with observation that certain targets including NGC 7501 which had very poor declination guiding with Southward Pulse Guides being largely ineffective in keepin the scope held on the lock position.  The following Guide Performance Chart for NGC 7501 , show that South jogs are ineffective in holding the telescope onto lock position for some 13 minutes, before eventually moving to lock position for the final 2 minutes of the guiding run.

 

Update 2021-12-16 Centering Issue has been fixed by 2021-12-16 by calling a new routine (JogUsingPulseGuide) that employ PulseGuides to move the scope when the required correction in a given direction (east/west or north/south) is <2 arc mins instead of the using TheSky6.Telescope.Jog method.  (Note: using Pulse Guide scope takes about 8.5s (8500ms) to move 2 arc mins with the guide rate used (x0.90).

Plan for improving performance of the Observatory's 12" LX200GPS/R scope is listed below.   The actions are designed to improve centering and guiding results.

• Backlash. Measure backlash in Dec (using new  Dec Motor Unit) and compare with previous Backlash measurements (using old Dec Motor)
  Done 2021-11-21
   
• Drive Training.   Train Ra/Dec Drives using a fixed terrestrial reference point
  Done 2021-11-22
   
• Residual Backlash.   Measure residual backlash after retraining RA drive. 
  Done 2021-11-23
   
• Polar Alignment.  Improve Polar Alignment, in particular improve alignment in altitude from c 19' error to < 1' error.  Impact of poor Dec guiding is lessened if star drift is lessened.  Procedure to improve polar alignment has been defined earlier (see Proposed Polar Alignment Change, 2021-10-25)
  Done 2021-11-23  (polar axis lowered by 17.7')
   
• Telescope Balance.   Check/rebalance telescope.   Poor balance might be producing the poor Dec guiding performance.
    Done 2021-12-05 
   
• TPoint Model.  Map new pointing data following Polar Alignment changes and build a fresh TPoint model.  This will be necessary for accurate pointing as key terms will have changed, and will also allow the improvement in Polar Alignment to be precisely measured.
    Done, Adapted Exitsing Model 2021-11-23
   
• Centering Method.  MExamine existing centering model and identify opportunities to improve it.  As a result the CorrectiveJog/ApplyJogs/Scope.Jog routine(s) where modified to call a new routine (JogUsingPulseGuide) that employ PulseGuides to move the scope when the required correction in a given direction (east/west or north/south) is <2 arc mins instead of the using the existing TheSky6.Telescope.Jog method. Centering performance is considerably better (faster and more accurate)
    Done, Improved Method 2021-12-16

A series of measurements of Declination Backlash measurements were made using PHD2 Guiding Assistent with a LX200 guide rate of x0.90 sidereal. 
For the tests the scope's anti-backlash setting for the Dec/Alt axis was changed to 0%.

Results

No.	Az (deg)	Alt (deg)	Dec Drift Rate (arcsec/min	Polar Alignment Error (arc min)	Dec Backlash (ms)	Dec Backlash (arc sec)
1 (S)	180	42	0.41	1.6	3294	41.2
2	142	41	-1.55	6.1	3325	41.7
3	116	40	-2.45	10.1	4001	47.9
4 (E)	99	18	-2.59	10.0	3729	44.0
5 (W)	274	38	3.19	14.5	4031	47.0
6	190	40	0.75	2.9	2936	36
7 (W)	270	7	-4.09  (*)	15.7	2667	30.3
8 (NW)	322	31	2.97	18.7	> 5606	65.7
9 (NE)	32	32	-2.26	15.5	3331	41.4
10 (N)	3	27	-0.63	4.6	> 5144	64.7
11 (0%)	130	64	-1.05	5.0	3981	49.0
12 (100%)	139	65			3850	44.4
13 (130%)	143	66			2717	32.5

 * The negative Dec Drift is unexplainable ! The screen capture data that was capture does suggest that star was drifting in an opposite direction to that expected for a azimuth position that lies West of Meridian.   Is this a refraction effect due to the star's low altitude (7 deg).


Analysis
Previous measurements of Dec backlash (2019-12-10) gave an average backlash of 4896 ms at guide speed x0.66 (range 4736 to 5076 ms) .  This is equivalent to 3590 ms at x0.90 guide speed (3457 to 3705 ms).

The average Dec backlash from the first 4 new measurements is 3587 ms at  x0.90 guide speed (range 3294 to 4001 ms) .

This suggests there has been no significant change in Dec backlash with the newly installed Dec Motor (2021-10-12) despite Dec guiding observations (2021-11-09) suggesting the contrary.   At x1.00 guide speed this corresponds to 3328 ms (3.3s).  DocG's notes reported that normal tolerance for backlash on a well tuned LX200 is between 2 and 4s.

Measurement points 2 & 3 are close to the positions where significant unexpected jumps in Dec Position of around 5" occurred with just a single Dec guide pulse during guiding on a couple of targets in session S932 (2021-10-09 , suggesting that the 130% antibacklash setting for the Dec axis in the scope is too extreme.  The setting was reduced to 80% on 2021-10-10.  The same issue occurred in Session S934 (2021-11-16) suggesting that the 80% Anti-Backlash setting itself was also too extreme and should be reduced further, and as a trial the setting was reduced to just 20% ahead of session S935 (2021-11-19).

Tests 11, 12 & 13 (made with anti-backlash settings of 0, 100% and 130%) showed a reduction in residual backlash with increasing anti-backlash use,  but even with 130% anti-backlash a residual backlash of 2717ms (32.5 arcsec) is still measured, suggesting that an even higher antibacklash should be used (a 150% setting was previously used with some success - see AntiBacklash Re-Tried, 2019-12-16 ) , or more probably that scope's backlash setting needs to be updated by 'Re-Training' its Dec Drive.  The later action was taken the following day (see Ra/Dec Drive Training, 2021-11-22).

Polar Alignment

Polar Alignment Errors indicated by PHD2 Guiding Assistent vary with position as anticipated. Since Polar Alignment Error is mainly in Elevation rather than Azmiuth, the alignment error seen by a single Guide Assistent run is low for positions near due South (1.6') and due north (4.6'), positions and greatest for positions in Western Sky (14.5', 15.7', 18.7') and Eastern Sky  (10.0', 14.5').

PHD2 measurement are broadly consistent with the Polar Alignment Error of 19.5' determined from most recent TPoint Mapping run that comprised  Az/Alt errors of  -19.5' ME and +0.7' MA  (see TPoint Mapping Run & Model, 2021-10-19).

Drive Training

 

Method
RA and Dec Drives on the Observatory's LX200GPS/R scope were trained today during daylight hours using AutoStar II's Setup/Telescope/Train Drives facility and the apex of a particular house's gable end, located some 400m to the SSW of the Observatory, as a Terrestrial Reference point. 

With the Telescope Tracking set to off (i.e. Terrestrial mode),  the apex of the distant house was centered in the GuideScope view using a ZWO 178MC Camera connected to SharpCap and Autostar II's Drive Training initiated.

Autostar II instructions were followed precisely and absolute care taken to not overrun the target upon centering since the operation depends on not reversing the direction of travel.

The whole operation took just a few minutes. 

Starting with the target centered, the order of operations after initiating the RA/Az training was :-

   1) scope slews left  - right arrow key used to recenter on the target (then Enter key)

   2) scope slews right  -  left arrow key used to recenter on target (then Enter key)

Starting with the target centered, the order of operations after initiating the Dec/Alt training was :-

   3) scope slews down   - up arrow key used to recenter on the target (then Enter key)

  4) scope slews up  - down arrow key used to recenter on target (then Enter key)

Notes:
- The centre of the guidescope view has previously been set to be very close to the pointing of the main scope
  (the residual difference (about 1 arc min) is considered irrelevant to Drive Training given that the
   same reference point / same camera view was used throughout the training)
- A x8 guide speed was used for recentering the scope after each telescope slew.  This provided a good balance between speed and control.
- A reticle was turned on in SharpCap to allow precision centering.
 
- Only after compiling this note was it noticed that the reticle in the SharpCap screen capture doesn't seem to be centered in the camera's field of view
  (not sure why it wasn't centerally located, it shouldn't affect the drive training provided that the reticle stayed in same position throughout the training.   Later experimentation showed that the reticle can be dragged and dropped to a new position, this must have happened inadvertently when setting up the reticle over a remote computer connection before physically re-locating to the observatory for the training itself)

Train Drive Results
'Edit Az Train' and 'Edit Alt Train' options within the AutoStar II's 'Setup/Telescope/Train Drive' menu was used to see the backlash values determined by the Train Drive operation.   The values for RA (Az) and Dec (Alt) drives were +3 and +44 respectively.

I've read in on-line forum postings that 'for well adjusted LX200s (the Dec backlash value) comes out between 60 and 80. '
https://www.cloudynights.com/topic/562518-getting-an-lx200gps-14-to-autoguide-smoothly/page-2

It's not absolutely clear what the units these backlash values have. They are probably in arc secs, but are they values which are independant of guide speed or not ?  . Moving 44" moving at x0.90 guide speed (13.537"/s) would correspond to 3250ms, which is not too far off the backlash value of 3587 ms determined by PHD2's Guiding Assistent (see Dec Backlash Measurements (2021-11-21) )

After completing the RA/Dec drive training the Anti-Backlash setting on the Dec Axis was set to 100%.
(this will be tweaked once new Backlash measurements are taken)

It will be interesting to see if the drive retraining will :-

- improve target centering
- improve guiding performance (especially in Dec)
- tighten up future TPoint Models

Other Work
Opportuntity was also taken to

 - Align the Red-Dot Finder with the Telescope pointing
   (red-dot finder is a fallback mechanism for re-aligning the scope should scope pointing be lost and
    it is becomes impossible to recover using (wide search) plate solution)

 - Slightly adjust the weights below telescope tube so there was a little more weight toward the front tube end
   (purpose of this slightly change was to make it slightly easier for the scope to guide south where it seems to have increased difficulty).

Update 2021-11-26
Residual Dec Backlash Measurements made during S936 session (2021-11-23) using PHD2 Guide Assistent shows that a 100% Anti-Backlash setting is slightly too aggressive for certain telescope positions, but not sufficiently aggressive for other telescope positions. See Residual Dec Backlash Measurements (2021-11-23).  To prevent overcompensation at any telescope position it would seem that a Dec Anti-Backlash setting of 85-90% would be probably be most appropriate setting to use.

Following charts compare centering results from the two sessions (S934 & S935) immediately prior to the 2021-11-22 Drive Training with those from the two sessions (S936 & S937) immediately after the Drive Training.    Target Centering (in Dec) seems to be better in the S937 session, but the S936 session doesn't seem to be parrticularly better than the two sessions prior to Drive Training.