56 replies to this topic

[supraslow]

Recently started experiencing guiding issues, and I'm having trouble nailing down what's happening. I've been running this mount since January, and guiding has been relatively good but over the last few weeks I've been seeing start trails due to an RA correction that is creating star trails on 50%+ of my subs.

 

I started with the default settings on the ASIAIR+, and just changed the exposure settings from 2s to 1s or 0.5s to try and improve guiding. I recently tried with a different guide camera, but the issue still persists leading me to think it's something else.

 

I've made sure that everything is tightened down on the scope, guide camera, and cameras themselves. Cable management is good and isn't snagging on anything. Polar alignment is typically under 3" error based on the ASIAIR PA routine.

 

Equipment Details:

- AM5N

- ASIAIR+

- ASI2600MC Pro

- RedCat 71

- WO 32 Uniguide 120mm

- ASI120MM Mini (guided well until it started these issues)

- ASI715MC (new camera, tried to use this to see if there would be any improvement.

 

Typical Exposure Settings:

- 180s at 0 gain

- 300s at 100 gain (dual band filter)

 

Guiding Settings:

- Calibration Step (1800ms)

- Max DEC Duration (500ms)

- Max RA Duration (500ms)

- Auto Restore Calibration enabled (I did try to disable this for a while and force a calibration on every guiding, but didn't get improvements.)

- Guide Stability 2", 5s, 60s timeout

- Dither 5px, interval 2

- Dark Library enabled for ASI120MM

- Corrected trigger Acc. 0.1px

- RA Aggr - 70% (left this at the default)

- DEC Aggr - 100% (left this at the default)

 

The AM5N was set to 0.25x tracking rate, and I changed this to 0.5x to see if it would help but had the same issues.

 

Attached my guiding logs from a good session (2025-03-08 and 2025-03-12), and then from 2 sessions where it just got terrible (2025-03-14 and 2025-03-25)

 

https://drive.google...UBb?usp=sharing

 

Here's a screenshot from PHD2 Log Viewer with the RA showing some crazy changes.

 

 

[Kevin_A]

Try turning down the aggressiveness to about 45% on both axis as this was a known issue on some am5 mounts where a too high setting caused spikes and oscillations in Ra. Tracking rate should be 0.5X standard.

There is no default aggressiveness setting with these mounts and lower numbers are usually better… but not always. Try these settings with 1s exposures and ensure that PA is fairly good. Use about gain 68 on the 120mm mini guide camera too. 
 

Edited by Kevin_A, 26 March 2025 - 11:42 AM.

[rgsalinger]

You are losing the stars probably due to short exposures coupled with the use of a color camera sometimes. You can see that when the star mass in some situations declines, you get a guiding spike. Even the cheapest 120 or 220 cameras are likely to work much better.

 

In other cases, I think that you have some kind of cable drag or wind because both axes suddenly spike. That's not going to be a guiding error. So, check your cabling for possible problems. 

 

I think that you should be calibrating once at the celestial meridian. I know that it's acceptable in theory to calibrate at each target but when you get above DEC 50, it can lead to errors. I don't necessarily think that's the problem but if you plan to image more than one target a night, you are wasting time and not accomplishing very much. 

 

Set the min move in pixels to correspond to .5" not .25", that's going to lead to useless corrections. Set the max move to 2500ms so that the system can move rapidly to correct errors. 

 

I have no view on aggression settings for that mount. 

[supraslow]

Thank you! Gonna give it a shot tonight with lower aggressiveness and go back to the 120MM to see if that helps.

[supraslow]

Last night didn't seem to go well either, I changed the RA and DEC aggr to 45% and the issue still persists. Calibration performed only once as close as I could get to the meridian and celestial equator. I also disabled dithering at one point, hoping that maybe there was an issue with these commands being issued and not being properly done. I did notice in the logs that the RA correction seems to happen roughly every 5 minutes.

 

I was standing next to the mount, and saw the corrections happen in realtime as the mount pulsed and move the telescope for 5 seconds or so until it stabilized and then continued guiding. No wind, cables didn't touch the mount or get anywhere close to it. Outside of these corrections, the guiding was around 0.7 to 0.9" with the adjusted RA/DEC aggr, with some drops into 0.5".

 

I then tried using my laptop with NINA + PHD2, same settings and the issue followed with the corrections every 5 minutes. I'm at a loss now to what's going on, could it be an issue with the mount itself?

Attached Files

•  PHD2_GuideLog_2025-03-26_202022.txt   497.43KB   8 downloads

Edited by supraslow, 27 March 2025 - 07:37 AM.

[rgsalinger]

If the mount is belt driven check to see that the belts are tight. 

[PIEJr]

Is there a way to do a factory reset on one of these?

Takes the software back to like new. (When it worked good.)

[markalot63]

Last night didn't seem to go well either, I changed the RA and DEC aggr to 45% and the issue still persists. Calibration performed only once as close as I could get to the meridian and celestial equator. I also disabled dithering at one point, hoping that maybe there was an issue with these commands being issued and not being properly done. I did notice in the logs that the RA correction seems to happen roughly every 5 minutes.

 

I was standing next to the mount, and saw the corrections happen in realtime as the mount pulsed and move the telescope for 5 seconds or so until it stabilized and then continued guiding. No wind, cables didn't touch the mount or get anywhere close to it. Outside of these corrections, the guiding was around 0.7 to 0.9" with the adjusted RA/DEC aggr, with some drops into 0.5".

 

I then tried using my laptop with NINA + PHD2, same settings and the issue followed with the corrections every 5 minutes. I'm at a loss now to what's going on, could it be an issue with the mount itself?

 

I would stick with NINA while debugging and watch PHD2 to see if it's losing the stars.  Also make sure you are guiding using multiple stars and the primary star is not a hot pixel of bogus reflection.  You want to be watching live as the RA starts to drift.  Or if on a PC use a video capture app to record the screen.  

 

You want to be watching when this happens:

 

 

No way this is a hardware issue, someone bumped the mount.  Within what looks like 1 second the guide star jumped 30"?  Check your guidescope and see if it or the camera is loose?  Something moved.  A hardware issue would be a slow drift, same with a sticky cable.  IMO 

[rgsalinger]

I don't know what the term "1 second guide star means". I do know that, if you have a dragging cable, then you can get some pretty nasty jumps. 

 

In looking at the guide log, there is no five-minute periodic error. So, it's not a loose belt as I originally suspected. 

 

Since he's using an AA, it would, indeed, be useful to see how the guide scope is mounted on the OTA. 

 

This is one of those guiding setups that looks like anything could go wrong. The guiding image scale is 6.45 arc seconds. The calibration is at DEC 47. And, unless my trigonometry is all wrong, the calibration was no good. The AA is infamous in not detecting calibration problems. The star mass was also jumping around oddly and the polar alignment was too far off for good guiding. 

 

None of those issues should lead to a jump like the ones in that log . So, if it's not a dragging/releasing cable and not something loose in the overall system, then I'd be looking at perhaps power to the mount. 

[markalot63]

You're starting with the most unlikely first IMO.  That blue line is the RA position, if calibration is wrong there would be no good tracking.  So the mount is guiding at a  .48 arc seconds and then bam, the guide star jumps 20".  At this point, because PHD2 is (correctly IMO) set for 500ms pulse times it takes a number of pulses to get the mount back to tracking the guide star(s).  This is why watching the screen is so important.  If one witnesses the image jump then what? 

 

A power issue or cable dragging would have a mirror image of the graph.  As the mount stops moving the RA would start to drift.  No jumps, a constant drift as the earth turns.  As power is fixed or the cable breaks free the mount would then jump back to position.  

 

This is what is happening in the logs (mostly RA but also some dec):

 

 

Flipped this is more like what a sticky cable looks like:

 

 

 

We could come up with all kinds of wild theories (as we like to do) as to what would cause the mount to jump like that, but I think a loose guide scope or camera make the most sense.   Perhaps bad cable management and it IS a dragging cable from the guide camera, but causing the camera to move?

 

Debugging the issue could involve:

 

Check for loose parts.

Check cables to determine if one could drag and cause something to move

Change the power source temporarily to see if this is a power issue

 

I'm imagining watching the PHD2 guide screen and suddenly seeing all of the stars jump in the RA direction by 20".  That's a fairly extreme jump, and according to the logs similar to a dither.  If you've ever watched PHD2 during a dither you know what that jump looks like.  It can be commanded, but what is doing it?

 

Here's a zoomed out portion with dithers on either side.

 

Edited by markalot63, 27 March 2025 - 01:01 PM.

[PIEJr]

Your graph in your first post sure shows a rhythmic hop in it.

Maybe there is gravel in the gears? Or a missing tooth?

Try ZWO in New Jersey

[supraslow]

Guide stars look OK and it doesn't seem to be losing them when it makes these corrections. Multi-star guiding is enabled by default on the AA, and I enabled it in PHD2 as well. I did my polar alignment in AA to under 1.5", same with PHD (based on what TPPA was telling me). 

 

I've added weight to the tripod, about 25lbs to try and mitigate any movement but this is sitting on grass in my backyard. Power is coming from the house through an extension cord.

 

I was watching the guide screen live, when suddenly it just does the huge RA correction and as I mentioned, no wind or cables touching the mount and then it settles down and repeats after a bit. Everything is about as tight as I can make it on everything. Here's a picture of the cable management:

 

 

I don't believe these belts are user-serviceable on strain wave mounts unfortunately. If there is a way to do it without voiding any warranty, I'd be happy to try it.

 

Here's what I mean about the roughly 5 minute peaks outside of dithering, this is on the 03-26 log with the AA:

 

 

 

[supraslow]

Your graph in your first post sure shows a rhythmic hop in it.

Maybe there is gravel in the gears? Or a missing tooth?

Try ZWO in New Jersey

Opened a case with them this morning, hopefully they have a user-serviceable way to look at it and see if that's the issue.

[markalot63]

Guide stars look OK and it doesn't seem to be losing them when it makes these corrections. Multi-star guiding is enabled by default on the AA, and I enabled it in PHD2 as well. I did my polar alignment in AA to under 1.5", same with PHD (based on what TPPA was telling me). 

 

I've added weight to the tripod, about 25lbs to try and mitigate any movement but this is sitting on grass in my backyard. Power is coming from the house through an extension cord.

 

I was watching the guide screen live, when suddenly it just does the huge RA correction and as I mentioned, no wind or cables touching the mount and then it settles down and repeats after a bit. Everything is about as tight as I can make it on everything. Here's a picture of the cable management:

 

 

 

I don't believe these belts are user-serviceable on strain wave mounts unfortunately. If there is a way to do it without voiding any warranty, I'd be happy to try it.

 

Here's what I mean about the roughly 5 minute peaks outside of dithering, this is on the 03-26 log with the AA:

 

Screenshot 2025-03-27 134004.jpg

 

You setup in the grass?  In the spring?

 

I think it's important to understand that it's not suddenly correcting, there is no single correction pulse.  The scope suddenly moves and then PHD2 sends multiple correction pulses to recover.  It's important to understand these graphs.

 

 

The square rectangles are the pulses, the line is how far away the guide star is from centered in each axis.   As you can see above there is no large correction.  Let's compare this with a dither pulse...

 

 

See the large RA and DEC pulses to move the mount?  That's what a large 'correction' looks like.    

 

These settings:

 

- Max DEC Duration (500ms)

- Max RA Duration (500ms)

 

Control how big a regular pulse can be sent in each axis during guiding.  This is generally a good thing for the AM5 because you want to limit the correction size as it's not needed and generally any large correction the guider thinks it needs to make is not true (usually poor seeing).  In this case it takes a number of these short pulses to recover from the underground critter that moved your scope (this is a wild guess).     

 

It's really hard to pin down what the problem might be but you can add something else to things to try:

 

Put tripod feet on pads to see if one of the legs is sinking into the ground, or setup on a solid surface for testing.  

 

The other guesses in this thread might be true, but I'd tackle the simple solutions first before worrying about something worse.  These errors are very regular, almost too regular for a leg sinking, but only show up during one log period. I assume you turned off dithering here?

 

Edited by markalot63, 27 March 2025 - 02:34 PM.

[rgsalinger]

You're starting with the most unlikely first IMO.  That blue line is the RA position, if calibration is wrong there would be no good tracking.  So the mount is guiding at a  .48 arc seconds and then bam, the guide star jumps 20".  At this point, because PHD2 is (correctly IMO) set for 500ms pulse times it takes a number of pulses to get the mount back to tracking the guide star(s).  This is why watching the screen is so important.  If one witnesses the image jump then what? 

 

A power issue or cable dragging would have a mirror image of the graph.  As the mount stops moving the RA would start to drift.  No jumps, a constant drift as the earth turns.  As power is fixed or the cable breaks free the mount would then jump back to position.  

 

This is what is happening in the logs (mostly RA but also some dec):

 

 

Flipped this is more like what a sticky cable looks like:

 

 

 

We could come up with all kinds of wild theories (as we like to do) as to what would cause the mount to jump like that, but I think a loose guide scope or camera make the most sense.   Perhaps bad cable management and it IS a dragging cable from the guide camera, but causing the camera to move?

 

Debugging the issue could involve:

 

Check for loose parts.

Check cables to determine if one could drag and cause something to move

Change the power source temporarily to see if this is a power issue

 

I'm imagining watching the PHD2 guide screen and suddenly seeing all of the stars jump in the RA direction by 20".  That's a fairly extreme jump, and according to the logs similar to a dither.  If you've ever watched PHD2 during a dither you know what that jump looks like.  It can be commanded, but what is doing it?

 

Here's a zoomed out portion with dithers on either side.

 

You didn't read my post carefully. That's too bad for you. To be clear - I never said that the OP's poor practices would cause a spike like that. I said, to paraphrase, that there are best practices to be used if you want to get the best that you can out of your mount. I pointed out the issue with the polar alignment shifting which is consistent with either the mount moving or a momentary loss of power. I chose loss of power because the error is in the RA axis but shifting makes sense too. 

 

At the same time, as you can see the problem is likely caused by setting up on grass. It's humorous to think that you don't understand the importance of a rapid correction sequence as well.

 

I think that it's obvious that setting up on grass, particularly if neither spikes nor pavers are being used, is yet another bad practice.

 

There's no such thing as a 20" correction being incorrectly applied due to seeing at any time when you would want to be doing DSO AP. That's why experienced imagers, who allow for things like a passing cloud, want large corrections made quickly to get the mount back on track quickly. That's particularly true when imaging at lower aggressions where a 2.5s correction is much less than you might think if you don't understand how the algorithms and the aggression parameters actually work.  

Edited by rgsalinger, 27 March 2025 - 02:42 PM.

[Bob2023CL]

Which version of the mount firmware are you using?  I am sorry if I missed it above.

 

I am currently using v1.6.3

 

If the belts are too tight you would see slipping as the belts slide over the teeth.

 

It is fairly easy to check, and there should be like 2mm play in the belts when you press on them with an allen key.

Edited by Bob2023CL, 27 March 2025 - 02:50 PM.

[markalot63]

You didn't read my post carefully. That's too bad for you. To be clear - I never said that the OP's poor practices would cause a spike like that. I said, to paraphrase, that there are best practices to be used if you want to get the best that you can out of your mount. I pointed out the issue with the polar alignment shifting which is consistent with either the mount moving or a momentary loss of power. I chose loss of power because the error is in the RA axis but shifting makes sense too. 

 

At the same time, as you can see the problem is likely caused by setting up on grass. It's humorous to think that you don't understand the importance of a rapid correction sequence as well.

 

I think that it's obvious that setting up on grass, particularly if neither spikes nor pavers are being used, is yet another bad practice.

 

There's no such thing as a 20" correction being incorrectly applied due to seeing at any time when you would want to be doing DSO AP. That's why experienced imagers, who allow for things like a passing cloud, want large corrections made quickly to get the mount back on track quickly. That's particularly true when imaging at lower aggressions where a 2.5s correction is much less than you might think if you don't understand how the algorithms and the aggression parameters actually work.  

 

Ok.  I'm sorry.

 

I did forget one other simple test and that's to replace the cable to the mount and possibly the guide camera?  I had a guide camera cable go bad on me, put a line and other oddities in the picture.  Not sure how this would play out if it was a mount cable.  I am also assuming use of USB to connect the mount.  

Edited by markalot63, 27 March 2025 - 03:08 PM.

[w7ay]

 

...  These errors are very regular, almost too regular for a leg sinking, but only show up during one log period. I assume you turned off dithering here?

 

The periodicity of the large swings definitely appears to point at something more deterministic than something that is caused by cable drags, wind, tripod slippage, or something bumping into the tripod.  Occam's razor says it is most likely caused by the mount itself.

 

The period appears to be about 270 seconds.  I don't have that mount, and don't know what that number corresponds to (one period of the gearing?).  The OP might check what in the mount is repeating every ~270 seconds, and look closely at that.  My strain wave mount's RA period is 430.82 seconds, but it is possible that the OP's mount has a much shorter period.

 

The 30" error might provide another clue.  I.e., if there is a geared belt slippage, what is the distance of a tooth?

 

Software bugs can repeat periodically too.  But in this case, I'd wager that it is mechanical.

 

Chen

Edited by w7ay, 27 March 2025 - 03:23 PM.

[w7ay]

Another clue would be to notice if the errors appear when it is an extra cold night, and not on other nights.

 

Chen

[supraslow]

You setup in the grass?  In the spring?

 

I think it's important to understand that it's not suddenly correcting, there is no single correction pulse.  The scope suddenly moves and then PHD2 sends multiple correction pulses to recover.  It's important to understand these graphs.

 

 

The square rectangles are the pulses, the line is how far away the guide star is from centered in each axis.   As you can see above there is no large correction.  Let's compare this with a dither pulse...

 

 

See the large RA and DEC pulses to move the mount?  That's what a large 'correction' looks like.    

 

These settings:

 

- Max DEC Duration (500ms)

- Max RA Duration (500ms)

 

Control how big a regular pulse can be sent in each axis during guiding.  This is generally a good thing for the AM5 because you want to limit the correction size as it's not needed and generally any large correction the guider thinks it needs to make is not true (usually poor seeing).  In this case it takes a number of these short pulses to recover from the underground critter that moved your scope (this is a wild guess).     

 

It's really hard to pin down what the problem might be but you can add something else to things to try:

 

Put tripod feet on pads to see if one of the legs is sinking into the ground, or setup on a solid surface for testing.  

 

The other guesses in this thread might be true, but I'd tackle the simple solutions first before worrying about something worse.  These errors are very regular, almost too regular for a leg sinking, but only show up during one log period. I assume you turned off dithering here?

 

Yep - rookie mistake setting up on grass, it's been mid 60's dropping to 30's so I'm sure there's some settling happening on the tripod. I'll try on the driveway and putting some pavers on the grass to see if that helps at all.

 

Correct on the dithering, tried turning it off in case it was causing something wonky.

 

Which version of the mount firmware are you using?  I am sorry if I missed it above.

 

I am currently using v1.6.3

 

If the belts are too tight you would see slipping as the belts slide over the teeth.

 

It is fairly easy to check, and there should be like 2mm play in the belts when you press on them with an allen key.

The mount is on v1.6.3. I popped open the front panel and didn't notice any slack when I pressed slightly on the belt.

 

Ok.  I'm sorry.

 

I did forget one other simple test and that's to replace the cable to the mount and possibly the guide camera?  I had a guide camera cable go bad on me, put a line and other oddities in the picture.  Not sure how this would play out if it was a mount cable.  I am also assuming use of USB to connect the mount.  

I tried a different cable for the guide cam, didn't notice any difference. I did use a different cable from the laptop to the mount though for the standalone PHD2 test and got the same result.

 

Here's the PE report for this specific mount:

 

[supraslow]

Got the chance to do some more testing last night, and the issue persists, unfortunately. I tried to test as many different variables as I could. I only used the AA and didn't use my laptop.

 

- I had some large pavers sitting around, set the tripod on those and ensured that everything was completely level. Polar alignment under 1" using the AA routine, did this twice to ensure it didn't change.

- Performed a guiding calibration as close as I could to the meridian and celestial equator, although this is pretty far in DEC (+37.1) since there are a lot of trees on the south side of the telescope.

- First target was North + East of the meridian. Within the first 10 minutes, I got the -45" peak on the RA. I have a short video while I was watching the guiding session here. I switched targets and continued guiding, started getting the RA peaks after 15-20 minutes in (max -28.86" peak), video here.

- I then switched sides and pointed to a target North + West of the meridian, over an hour of guiding with no RA peaks over 5". Total RMS was 0.97".

- After I turned off dithering 2 nights ago, I didn't turn it on to ensure this wasn't causing any adverse effects.

 

At this point I went to bed but setup my plan for M101 which has been the focus of this current project. Calibration ran automatically (not sure why it's not honoring the restore calibration switch), but guiding ran pretty horribly with the max peak around -53.62". I noticed in the graphs that it took about 20 minutes for the first peak to happen and then it just continues with those peaks ~5 minutes. Although, eventually I can see during the session that I get 2-3 additional smaller peaks the longer it goes on. At 03:10 EST, it did the meridian flip and started on the East side of the pier, guided for over an hour at 0.74" RMS with no peaks.

 

I'm noticing that the issue seems isolated to the west side of the pier. My view from the backyard is strictly north, so my targets are within 2-3h either side of the meridian.

 

The guiding log is too large to add, but here's the link.

Edited by supraslow, 28 March 2025 - 08:55 AM.

[Bob2023CL]

Do you have an estimate of the mass of your telescope? I also am concerned about the fact there's no slack in the belts. If they are that stiff they are way too tight.

[supraslow]

Do you have an estimate of the mass of your telescope? I also am concerned about the fact there's no slack in the belts. If they are that stiff they are way too tight.

13.7lbs with camera, guider etc.

 

I've been running this setup since January with a Sony A7III (1.4lbs) and then moved to the ASI2600MC (1.5lbs). I didn't have any guiding issues for those first 2 months actually.

[markalot63]

This frame is crazy, 2 guide stars indicating this caught the jump.

 

 

What in the world would cause that?

 

A belt slipping would not cause a jump this large in 1 second would it? 

Edited by markalot63, 28 March 2025 - 11:55 AM.

[rgsalinger]

Well, your guiding image scale is 6.45 arc seconds/pixel. That means that a 20" deviation is just over 3 pixels. The default search region in PHD is usually 20 pixels.

 

For some reason the value you set was 50 pixels - over 2 arc minutes.

 

"Search region = 50 px, Star mass tolerance = 50.0%"

 

 

That means that it's going to be easy to get a second star in the frame. When that happens, PHD may not be able to accurately compute the weighted centroid.

 

Now what bothers me is that multi-star guiding which should prevent a single odd measurement from causing this effect. (I've never seen this before since I never use a short guide scope.) Are you using multi-star guiding or not? I can't find that parameter in the last log (3/23) that you provided.

 

https://bbs.zwoastro...ti-star-guiding shows how to get multi-star turned on using an AA. 

 

I would also point out that you really want to use 16 bit guiding, not 8 bits. You'll get a more accurate centroid calculation using the finer granularity provided by 16 bits. Again, that's mostly just a best practice thing, to me the key is that 50 pixel setting and possibly not using mult-star. You might also try to get a better focus on the guiding system. Your stars are almost 25" of FWHM. 

 

My guess is that this happens mostly on one side of the meridian because you are imaging in a star sparse region coupled with some degree of imprecise centering after the flip. That leads to it appearing that one side is much worse than the other. 

 

I've always hated the tiny guide scope/110guide camera combo. It just never sounded like a good idea. So, maybe this is all just confirmation bias.