Help with improving RA tracking

[Magnus Larsson]

Hi!

I have a CGE-Pro that I upgraded to OnStep about a year ago. I used to have good RA tracking - down to 0.5" when seeing allowed it (no PEC). 

Some 2 months ago, I had problems with a ball bearing crashing in the RA drive block. So now I have taken down the whole RA system, cleaned, lubricated, replaced ball bearnings (twice) and so on. Still, however, I can not get back to the state I was in before. I'd appreciate all comments and ideas. 

Right now, the drive train consists of the worm gear with with a period of 337.9 secs (the big wheel has 255 teeth if I am correct - not much information to find on this). On the worm gear (the "screw") sits a GT2 pulley with 60 teeth. This is driven by a GT2-pulley with 12 teeth, via a belt. And in the end a stepper motor. 

Attached is a PHD2 log with a Guiding Assistant run as section 13. The frequency analaysis looks like this

Beyond the 1 frequency (the worm gear) all peaks are on fractions of this. The 5th is biggest, and that resonates with the drive belt (60 tooth - 12 tooth). But there are more peaks. 

My guiding is now around 1" at best (as can be seen in the section 14 of the log file), so despite most of these frequences being relatively slow, I seem not to be able to compensate for them well enough. And before, the GA curve was in fact much more smooth. 

Any ideas on what I should check or test here?

Magnus

[pollya...@gmail.com]

Hello Magnus,
I'm an LX200 Classic- PHD2 learner too, but running with the original drives. I see you've converted to the OnStep system and noticed the guidelog says the RA drive speed is well above siderial rate. I suspect running the RA at 25 arcsec/sec is likely to introduce serious oscillation issues. I'm sure the experts will advise more when they return.
- Jack T

[Brian Valente]

Magnus

the most meaningful feedback from your guidelog is in the 1 hour run, this is the residual error. Your main PE is handled well by guiding, but the smaller errors are not corrected. specifically around 33.8, 48.5, and 67.8 seconds. I suspect the 33 and 67 are related. All of these are a frequency below the worm. it sounds like you have a good handle on the mechanical, so i think you have some more work to do on that belt drive.. the error at those frequencies isn't sudden, so it doesn't look like the belt is jumping or meshing improperly

--
You received this message because you are subscribed to the Google Groups "Open PHD Guiding" group.
To unsubscribe from this group and stop receiving emails from it, send an email to open-phd-guidi...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/b9823591-c797-4610-bbc8-62eac19efaden%40googlegroups.com.

--

Brian 

Brian Valente

astronomy portfolio https://www.astrobin.com/users/bvalente/ 

[Magnus Larsson]

Hi!

Thanks a lot for looking at this and trying to think about it!

Just wondering: is there anything in particular that makes you suspect the belt drive system, rather than, for instance, the ball bearings? 

Moreover, if you look at the GA, at the "top" of the curve, there is like a double peak. That has been there for a while, while I have been tinkering with different ball bearings and the distance between worm and wheel etc. Are those "double peaks" indicative of anything in particular?

I'll keep experimenting. Unfortunately, really bad weather stops me from moving forward with more tests. I wish there was a way to test these things without depending on stars.... A way to spend cloudy nights :)

Magnus

[Brian Valente]

>>>Just wondering: is there anything in particular that makes you suspect the belt drive system, rather than, for instance, the ball bearings? 

just what you said:

you replaced the ballbearings twice, right? did you see any difference

>>>The 5th is biggest, and that resonates with the drive belt (60 tooth - 12 tooth). But there are more peaks.

you pointed out its the belt drive

It's possible it's something else, but it's certainly mechanical and certainly between the worm and motor. You might ask celestron (or celestron user group) what the 5x error would come from if you're not sure

To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/03b1aed3-1b8d-460a-8683-bcc5986ae45an%40googlegroups.com.

[Michael Waring]

Hi Magnus.

I see plenty of opportunities for harmonics to occur in a belted drive system that gears-down.

I see very little opportunity in a very slowly moving ball bearing.

Michael

Wiltshire UK

[Magnus Larsson]

Hi!

Thanks for your thoughts and suggestions! So now I am directing my attention to the belt drive, leaving to the side (for the moment) obsession with ball bearings and adjustment of distance to the wheel. 

So what I have done so far is to replace the small pulley (12 tooth) with a new one, and tighten the belt a bit more (not sure how tight it should be (driving with teeth rather than tension, but sufficiently tight). I think there is clear improvement. 

Attached is a guide log from last night. I was observing a lot of Mira variables, each observation run is 9-10 minutes. There is also a GA run, segment 3. The GA is now more smooth, but there are two interesting (I think) inverse peaks repeated with about 1:45 min distance at the low point of the curve. However, just two - not a repeating pattern of 1:45 min period. 

 

Some of the guiding runs are terrible, I know. Some are quite good (below RMS = 0.8"). Stars are all over the sky, some thin clouds came in at times, so I am not worried by the varying quality. I am more intrested in the best guiding sessions (or is this the wrong approach?). In some of these, like 14, 17, 25, I can see the same inverse peaks about 1:45 apart. 

Now, what I have is: 

small pulley: 12 tooth

large pulley: 60 tooth

belt: 200 tooth

Interestingly, 60/200 is very close to 105/338 (the distance between the inverse peaks in seconds/worm period). Is it reasonable to take this as indication of a problem with the belt?

Other ideas on these results?

New belts and new 60 tooth pulleys are ordered but not yet arrived. 

Magnus

[Brian Valente]

Magnus

i wouldn't say those 0.8" RMS results are particularly good. your RA is 2x Dec, so you will have star elongation. 

the unguided results still show some pretty significant jumps even without guiding. the notable one is around 48.8 seconds (also pointing out the primary PE if around 8" p-p, which isn't bad

looking at your guided runs and the residual error, it's that same period that is limiting your RA performance (note there are others: you have quite a bit of error >0.4" below that period, so once you find and resolve that 48 second error, there is likely more to be done):

You might get some additional mechanical tips from users who also have the same mount setup, but i think your time is better served by working with celestron or a mount user community to find the exact causes. With PHD logs, we can only point out where the error is, what is causing it does not come through in the logs. Borrowing a recent comment from Bruce "The PHD2 logs quickly reach a point of diminishing value when trying to understand the details of mechanical problems.  For the most part, we are just making semi-educated guesses about what's going on"

>>> leaving to the side (for the moment) obsession with ball bearings 

i suspect this is residual from your prior mount

Brian

To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/86dbc524-b5c4-4062-a8be-429eb3f76975n%40googlegroups.com.

[Magnus Larsson]

Hi!

Thanks! Yes, I do not mean to say 0.8" is good. I mean this at least on the right planet - not 2" or something... I'll keep working. 

A bit of the problem for me is that the "user community" is not easily found. Not many use this mount anymore (if there ever was) and certainly not with the OnStep conversion - and the OnStep people knows too little about the mount. But that will not stop me, the mechanics are not that complicated. 

I'm working on chasing down that 48 second thing as a top priority now. 

Thanks. 

Magnus

[Brian Valente]

regarding the 0.8" - that value can actually can be 'okay', but the main issue is RA is 2x the rms of Dec, so you will get star elongation (vs. just general star blobbiness)

Good luck hunting this down - you definitely picked a mount and mod that is has a limited user base 

To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/95e6cd8b-cc1c-4542-9dd3-9de0366e9acan%40googlegroups.com.