Adjust Lock Position - Is this what I think it might be?

[PartlyC]

Hello!

I have finally been working on moving the star into the slit (seems to be popular today), and am looking for some suggestions or validation on what I am doing.  I am just testing to get the limitations and some best practices for this.  I am working in "manual" mode using the GUI (by manual, I mean using the GUI and not separate software talking to your server)  I have re-read the instructions on the Adjust Lock Position, and I'll be working this again tonight...  The mount is an AP1100AEL.  Scope is at a remote site, and is an open 10 inch GSO RC Truss (2034 focal length), so those may have to wait for the site people to put the primary cover on for me.  I think lack of darks should not be too big an issue at this point (hope)

First, Logs from last night.  Skip to the very end if viewing in the "log viewer", 28, 26 and 25 etc. are probably the best examples - I was trying a lot of things last night and even switched to the the 174 guide camera in the spectrometer early (thought closing and re-opening would start a new log file, but was wrong):

  

https://openphdguiding.org/logs/dl/PHD2_logs_Phne.zip

Tonight I do hope to be able to get darks for the guider, but not sure if I'll be able to. I have done them before, but it has been a while. 

My finder is getting the target pretty close to the slit with the plate solves.  I'll need to tighten that before this is finished.  That said, I was testing using some stars that were very near the slit.

Note that the system seems to be doing OK guiding when not messing with the lock position.  It shows that PHD2 is calibrated, but does not have darks.

So, the issue:  I can't  tell if adjusting the star position is even working, and it kind of looks like it is not actually adjusting for the new lock position.  The big crosshair is showing the change when I set the new position.  I wait, and the box and target star don't seem to move.  I make additional  small moves, and though the star stays in the box, tracking goes away, as if it has stopped.  Large changes will cause an almost immediate loss of lock, of course.  Either way, even small changes don't seem to move it or quickly seem to cause guiding to stop or drift (RA and Dec just go out and never come back). 

The AP does not let me change the guiding rate (it is always 1x with the encoders).  Suggestions offered there are to change the "Aggressiveness" settings.  I have not tried this  yet.

I did try enlarging the "box" a bit (set it back to the default 15 pixels).

Any other thoughts?  Am I missing something obvious?

Any suggestions would be appreciated.

Thanks!  

-Mike

[Richard Scott]

Hi Mike,

I use the Adjust Lock Position feature for my stellar intensity interferometer and it's a fabulous feature of PHD2. I use it to move a star onto a fiber optic cable and then to fine tune the position for maximum photon count rate. Once centered properly, I save the lock position and PHD2 will move the star onto the fiber cable for following runs. The speed that PHD2 will move the star depends on the exposure time you're using, so I use shorts exposures (typically 0.1 sec) during centering and then back to long exposures for guiding.

Rick

[Bruce Waddington]

It looks like you're sort of banging away at this thing willy-nilly.  The adjust lock position window lets you specify how far you want to shift the lock point and then shows you the numerical values of the new lock-point.  If those numbers change, the tool is working.  But the actual shift of the target star to the new target requires some number of guide corrections to complete, so if you're just whacking away without waiting for the guiding process to do its job, you won't see much happening.  Of course, you have to be actively guiding - not just looping, not calibrating, only guiding.  Since you're using 2.5 sec exposures, each guide correction will take at least that much time to complete and you can't assume that the shift will be completed in one guiding cycle.  Finally, it looks like you were chipping away at this process using incremental amounts of only a few tenths of a pixel even though your guider image scale is only 0.59 arc-sec pixel. Again, you're not going to see anything happening using such tiny adjustments.  I think if you slow down and just work with this a bit on a test target, you'll start to get good results.

Bruce

On Saturday, May 30, 2026 at 11:25:23 AM UTC-7 PartlyC wrote:

[PartlyC]

Rick,

Thanks for the additional tips.. I do know this is a great feature, and that somehow I was doing something obvious wrong... based on Bruce's message, I was right <g>  Th faster update rate is a good idea....  Thank you!

-Mike

[PartlyC]

Bruce,

As always, many thinks...  I have no doubt your main points are the culprits, but I want to be sure I understand an item or two.  I tried a lot of options trying to narrow down my issues (last night and tonight are dedicated to this) based on documentation and (other) dumb mistakes.The test run was all to get me a feel for what I didn't know on this subject, so no real targets were involved<g>.  

So, based on your message (which was mostly how I understood it to work) and a couple of "duh" moments....  

 

1) Main issue:  I was using the "default" Adjust Lock Position move amount (pressing the buttons) of .33 pixels most of the time (when in doubt...). And was not relating that properly to the guiding being in angles vs pixels(!) as you point out.  The 2.5 second update rate does indeed make that a lot worse.  That would definitely make this REALLY slow (or perhaps not move?).  I did look at the numbers once and I had done  fixed change tests  (over 100 pixel moves) and a test or two with higher increments earlier, and I got guiding lost alerts pretty quickly if I recall.   In your docs it warns you that too much of a move can cause loss of lock, so I went back to small increments for most of my tests (overcompensated).  Yes, I should have caught this one... How does it decide how much to move on each update, however?  Max move?  I  would assume that "min move" would keep the move from happening if it were too small (I probably saw a lot of that). Does it accumulate? 

2) Let me make sure I understand the second part, which when added to 1 above... When it is moving to the new lock position, is regular guiding still actually happening?  My most perplexing issue was that from when I started the move and suspected it was moving, the target/guide star would start moving away from the target position as if guiding was not really active. RA and Dec errors would shoot up.  Eventually it would move completely out of  the image and/or l would lose lock if I didn't stop it and re-acquire and guide.  Is this what you expect when the move is too slow (and perhaps too far)?  That would explain part two of what I saw - the target star being tracked pretty well, then on setting the adjustment watching it roll of the screen.  This is visible in the guide logs at the end of the file... RA and DEC spread and never re-converge.  Or am I missing something else obvious here...  It would make sense if this were the case.  It is not a slew, but simulating one with guiding pulses...

FYI, I would often stop guiding, re-start looping and re-select for the next test.  And ya, forget to restart guiding.  I watched for that and restarted the process when I noticed it.  At times, I did an auto-star select mostly to make sure the star I was selecting was a PHD2 "Preferred" star, and not something that just looked good to me.

I will work with "1" tonight, and hopefully that will show how PHD2 does the move in "2", or my question will become moot. 

Great help, as always... This is a great feature!!

Thank you!!

-Mike

[Bruce Waddington]

Trying experimenting with it this way:

1.  Make sure you are guiding normally with the guider exposure time set to 1 second.  Make a note of the time of day when you go to step 2.

2.  In the Adjust Lock tool window, set the 'step' slider to 3 or 4px.

3.  Click once on the 'left' button, then leave it alone until the guiding stabilizes

4.  Click once on the 'up' button, then leave it alone until guiding stabilizes...

5.  Repeat steps 3-4 for 'right' and 'down' directions

This should bring you back roughly to where you started.  I'd rather you did this and reported back with any new difficulties before trying to dig around in your earlier log file trying to figure out what happened.

Good luck,

Bruce