PHD2 - generating graph without guiding

[Tommohawk]

Hi all.

I want to improve guiding - don't we all - but initially Id like to have a closer look at whats going on with my mount without guiding switched on. If I could generate a guiding graph with guiding off that would be helpful - point being if the drive is horribly nasty and all over the place its going to tricky to get it right, where if its basically smooth without guiding I'd have a fighting chance. Make sense?

I can see how to turn Dec off, but I cant see how to switch RA off. Maybe this isn't even possible, or there is some other way of generating the graph basically to show how the guide star deviates without guiding.

My guiding is fair, usually sub arc-second, but I want to improve further with a view to using longer FL. I'm using an ADM dual mount so pretty sure I don't have any flexure issues.

I'd be grateful for any input.

[Craney]

Hi there, don't know if you have seen this.

New version of Phd2  2.6.7  has a thing called a Guiding Assistant, which does pretty much what you have described. 

https://astrobackyard.com/phd2-guiding/

check out the video, which is live podcast of a knowledgeable chap explaining how it works .  I'm working through it tonight as my guiding has been very poor recently. 

 

Sean.

 

 

 

 

[Tommohawk]

Hi Sean, thanks for that, guiding assistant looks just the job. I'll check it out tonight.

 

[Tommohawk]

Ok, given the guiding assistant a go, and it's really good. Well... it shows how awful the mount is unguided. It's a miracle I can get sub arcsecond. 

Not sure if this is down to the stepper motors or bearings or what, but I think trying to tweak PHD 2 setting further is a lost cause.

One other comment. The DEC backslash assessment isn't really measuring the total backlash unless it's perfectly balanced. If the weight is biased one way as it usually would be there won't be significant measurable backlash. But overall a great tool. 

[michael8554]

4 hours ago, Tommohawk said:

The DEC backslash assessment isn't really measuring the total backlash unless it's perfectly balanced. If the weight is biased one way as it usually would be there won't be significant measurable backlash

The PHD2 test moves the mount in Dec a goodly number of steps.

If it was balanced to have the backlash taken up, then when it reverses and starts to move back it will encounter Backlash, and will measure how long it takes to start moving back to the start.

If balanced the other way, it will encounter Backlash at the start of the test. 

Michael

Edited April 20, 2020 by michael8554

[LongJohn54]

I ran the guiding assistant a couple of days ago and have still to work out what's going on.   I don't like 12 seconds backlash but everything else seem ok.   Can you post a screenshot of you result?

Edited April 20, 2020 by LongJohn54

[Tommohawk]

7 hours ago, michael8554 said:

The PHD2 test moves the mount in Dec a goodly number of steps.

If it was balanced to have the backlash taken up, then when it reverses and starts to move back it will encounter Backlash, and will measure how long it takes to start moving back to the start.

If balanced the other way, it will encounter Backlash at the start of the test. 

Michael

I'd like to believe that, but I cant see how that can be so.

If DEC is imbalanced, then at the outset the mount will move immediately whichever way Dec is driven - either the drive will move it against gravity, or gravity assist will move it the other way. The same is true at the point of reversal. The whole point of being imbalanced is that when guiding you could have a mile of backlash, but this will never become apparent.

By way of example I ran this test with my rig set as it normally is - and the report said I have no significant / measurable backlash - or words to that effect!. But if I remove the scope and just grab the puck and tweak it I can feel backlash pretty obviously.

The issue with DEC backlash when guiding is that if you have to drive/guide in both directions - unless you have the perfect mount. You can effectively remove the backlash with imbalance... BUT when you do the drive will only be smooth when working against gravity - when driving with the weight behind it tends to be notchy. 

This is why we tend to work East heavy in RA - you could eliminate RA backlash by being West heavy, but then you'd be driving with the weight behind you so  to speak and its more notchy.

Maybe my explanation isn't very eloquent - I did something similar here which may be better worded.  

..................................................................................................................................................................................................................

In any event, the guiding assistant was very useful - here's the graph with guiding off - it's horrible!! what the heck is going on in RA. ??? The PA isn't great but that cant be poor PA because it recovers after a while and drifts the other way. 

[michael8554]

Let's say that in the test the mount is balanced so that the worm is already preloaded, is engaged on one face. 

At the end of the North moves the motor reverses. 

The motor now pushes on the other face of the worm to move south. 

Unless the worm is perfect, the move from pushing on one face of the worm  to the other face of the gear involves overcoming backlash. 

And the balance still wants to engage on the other face, but the motor has overridden that load. 

In the Guide Assistant RA drive is turned off, so you're seeing the RA Periodic Error that most mounts have. It's about 20arcsecs peak to peak which isn't  too bad. 

Use the PHD2 Permanent PEC Algorithm, or investigate recording PEC in your mount. 

Michael 

Edited April 20, 2020 by michael8554

[Tommohawk]

1 hour ago, michael8554 said:

It's about 20arcsecs peak to peak which isn't  too bad. 

OK thanks thats reassuring.

However: 

1 hour ago, michael8554 said:

The motor now pushes on the other face of the worm to move south. 

I cant agree with this! If there is a load in one direction on the worm gear because of imbalance the worm will only ever engage on one face of the worm gear. If for example you dismantled the mount and just removed the worm gear, the Dec would spin because of the imbalance. The backlash only becomes apparent if there is perfect balance - this is true whether guiding for real or using the guide assistant.

Edited April 20, 2020 by Tommohawk

[newbie alert]

4 hours ago, Tommohawk said:

By way of example I ran this test with my rig set as it normally is - and the report said I have no significant / measurable backlash - or words to that effect!. But if I remove the scope and just grab the puck and tweak it I can feel backlash pretty obviously.

You can iron this out by adjusting the worm, all you need are 2 Allen keys and a few mins or several hours of wow that was easy or swearing at yourself for touching it

[Tommohawk]

3 hours ago, newbie alert said:

You can iron this out by adjusting the worm, all you need are 2 Allen keys and a few mins or several hours of wow that was easy or swearing at yourself for touching it

Thanks for that. As it happens I have stripped and rebuilt a few mounts and am well aware of the adjustment to minimise backlash.

My position on backlash is that we shouldn't obsess over it, because - as the guiding assistant demonstrated - provided there is slight imbalance it will be undetectable and in practice of minimal consequence.

For a more detailed explanation I would encourage you to read my other thread on this referred to above. 

Edit -  sorry just re-read this and it sounds a bit snotty. Wasn't meant to be!  😉

Edited April 20, 2020 by Tommohawk

[michael8554]

Hi Tom

Hmm, I'm beginning to understand what you're saying.

Talk me through what happens when you're balanced for guiding, say, north only, and a dither takes the graph the other side of the axis?

Michael

[newbie alert]

1 hour ago, Tommohawk said:

Thanks for that. As it happens I have stripped and rebuilt a few mounts and am well aware of the adjustment to minimise backlash.

My position on backlash is that we shouldn't obsess over it, because - as the guiding assistant demonstrated - provided there is slight imbalance it will be undetectable and in practice of minimal consequence.

For a more detailed explanation I would encourage you to read my other thread on this referred to above. 

Edit -  sorry just re-read this and it sounds a bit snotty. Wasn't meant to be!  😉

Oooh.. I can get snotty too but I refrain ..

[Tommohawk]

2 hours ago, michael8554 said:

Hi Tom

Hmm, I'm beginning to understand what you're saying.

Talk me through what happens when you're balanced for guiding, say, north only, and a dither takes the graph the other side of the axis?

Michael

That would depend on how you have the imbalance set. 

Case 1. You are guiding North only, perhaps because your PA is slightly off without guiding and its drifting slightly south, and lets say the weight of the imbalance is also pushing to the North. In this case the imbalance force is always pushing North, but the engagement of the worm wont allow that movement unless the drive turns that way, as it will when North guide pulses are made. This is not ideal, because the drive is notchy when the force is in that direction. 

When dithering north, pulses allow the north movement in the same way as the  normal guide pulses. When dithering South, the pulses drive against the imbalance force to make the South movement.

Case 2. Would be where the imbalance is set the other way, with the imbalance pushing South. This is better for guiding North only, because now when making North corrections in normal guiding, the motor is working against the south force of the imbalance. No South pulse commands are given when guiding of course, but when dithering South pulses allow the imbalance force to move the scope southwards.

Case 3. DEC is perfectly balanced. Backlash may become apparent on the initial North commands, but once backlash is taken up should be OK. If dithering backlash will become apparent if dithering to South - although obviously this wont matter, except it will take a moment to take up backlash when guiding north again. 

The issue with the jerkiness of the drive when driving with the imbalance force behind it seems common to worm drives on Synta mounts - I'm confident of this having stripped a few. In all cases if you remove the intermediary gears (Rowan belts in my case) and drive the mount manually by rotating the worm drive, it's smooth when driven against a force but notchy with force behind ie "helping" it. I can't speak for other mounts though.

The practical consequence of this is twofold. For RA you want to guide East heavy so you are always driving against this force. Not too much of course. For DEC you have to be a bit careful. In theory if your PA is slightly off as in the above case, you can guide in one direction only - but ideally you need to balance the mount against this force. If you dont have imbalance and are guiding both N and S, then your backlash will become apparent whenever DEC pulses switch from N to S. I think the idea of the "resist switch" guide algorithm is to help with this, on the assumption that with good PA DEC shouldn't need too much correction. 

The reason I make a big deal of all this is really to encourage folk not to obsess about backlash - especially if you end up with tight spots as a result. 

Having said all that...... !!!!  I don't find massive differences in practice if I'm West heavy - eg if I forget to reset after meridian flip. Also if I try and guide in DEC in one direction only this also makes little difference.

TBH its a miracle that PHD2 can get these mounts to sub-arcsecond performance! 

Having seen the raw data on the GA plot above, and bearing in mind what you've said about my PEC being about normal, I think I conclude that the mount is probably preforming about right - on a 90 minute imaging run last night I only got 1.19" RMS, but it was a bit windy.

I still think the PHD2 help guide should make the point about Backlash assessment needing neutral balance though.  

Edited April 20, 2020 by Tommohawk

[michael8554]

I wasn't precise enough in my scenario.

I'm case 2, front of scope heavy pushing south, PA slightly off (5 arcmins) heading south, guiding north.

Thanks

Michael

[Tommohawk]

1 minute ago, michael8554 said:

I wasn't precise enough in my scenario.

I'm case 2, front of scope heavy pushing south, PA slightly off (5 arcmins) heading south, guiding north.

Thanks

Michael

Should be spot on then!

[masjstovel]

There are some points here worth mentioning i think regarding the GA and how it can be useful. Especially if you have beltmodded the mount. Ive been beltmodding and tuning my EQ6 - new bearings, regreasing etc. I struggled like a mad man on DEC backlash for days and days. Was 33000ms now down to 2000 and some ms - so still not perfect.

I was tuning the worm setscrews for every microscopic nanometer possible i believe, but still backlash was huge according to PHD2 guiding assistant.  

 

Im a newbie and all that, so things could have gone smoother, but my point is i think you can use the Guiding Assistant to your favor. You can use it to troubleshoot the backlash source. If you imbalance DEC and do GA and still get alot backlash - then you "know" your backlash problem is not the wormdrive. Then you can check for example the guidescope arrangement for looseness, and redo the test. If the backlash is still consistent then you can maybe try readjusting the belt (if the mount is beltmodded), and so on.

[Tommohawk]

16 minutes ago, masjstovel said:

There are some points here worth mentioning i think regarding the GA and how it can be useful. Especially if you have beltmodded the mount. Ive been beltmodding and tuning my EQ6 - new bearings, regreasing etc. I struggled like a mad man on DEC backlash for days and days. Was 33000ms now down to 2000 and some ms - so still not perfect.

I was tuning the worm setscrews for every microscopic nanometer possible i believe, but still backlash was huge according to PHD2 guiding assistant.  

 

Im a newbie and all that, so things could have gone smoother, but my point is i think you can use the Guiding Assistant to your favor. You can use it to troubleshoot the backlash source. If you imbalance DEC and do GA and still get alot backlash - then you "know" your backlash problem is not the wormdrive. Then you can check for example the guidescope arrangement for looseness, and redo the test. If the backlash is still consistent then you can maybe try readjusting the belt (if the mount is beltmodded), and so on.

You make a couple of good points. Backlash as you say could be in the gears/belt, so it isn't necessarily the worm that's the issue.

If the GA consistently shows a big backlash even when DEC is imbalanced then I think the friction in the bearings must be greater then the imbalance force. As you suggest an insecure guide camera could cause apparent backlash, but this would presumably be erratic. 

Like you I obsessed over the adjustments for ages before I realised that it really isn't necessary. With RA especially you could have miles of backlash but so long as you're East heavy it wont matter at all.

 

[michael8554]

Good point Tom about the effect of friction in the bearings - Stiction will look much like Backlash in PHD2 graphs. 

Michael 

Edited April 21, 2020 by michael8554