Adjusting the RA Worm Block on a G11

[bazza]

Well, my PE is way too much so I'm going to have to adjust the RA Worm block on the G11. Can anyone guide me through the steps.....just the thought of it........it's frightening me to death!!

All help, gratefully received.

Barry.

[spacemonkey]

I can't, but I know a man who can...

http://tech.groups.yahoo.com/group/Losmandy_users/

Regards, Dave

[coatesg]

Make sure you don't make it too tight or it will bind and possibly damage the motors. It should be just loose enough to be able to turn the mount fully loaded and balanced up through the full 360 degrees of ra just by turning the drive gear with your fingers.

RA worm can actually be really loose as long as the mount is weighted on the east side (ie whatever is on the east of the mount has a slightly larger moment than what ever is on the west). This causes the worm to always be in contact with the gear and means that (when guiding at < 1x siderial rate) there's no "swimming" around in the backlash of the RA gear.

When you do the spacing, use a feeler gauge to get the gap the same both sides.

And finally, make sure you know what the PE is before and after to see what change you have made to it!

Cheers,

[bazza]

Thanks for those tips - I had already been to the Losmandy users Yahoo site, though theres loads of discussions about the RA drive, I can't find the specific instructions on 'how to.'

Graeme,

Thanks for the advice - Using Pempro, having done 6 cycles with PEC off, then creating a PEC curve, it comes out at 11.41 arc secs. I have uploaded this to the mount.

I did then do a run with PEC on, but when I try to 'refine' this in Pempro (ie. adding or subtracting curve 1 (PEC off) from /to curve 2 (PEC on) it comes out at more than 11.41 arc secs...so this appears the best I can get.......therefore my reasoning to adjust the RA Worm. Would you agree?

Or do you think at 11.41, I should be able to guide that out?

Thanks in advance.

Barry.

[ottUp]

It's really simple - nothing to fear. Take the cover off (two tiny hex screws). Loosen the large hex screws under the worm blocks, adjust the spacing with finger pressure, re-tighten.

Remove the drive motor first so you can turn the worm by hand to check the free play - the idea is to make the fit as tight as possible without affecting the ability to turn the worm effortlessly by hand. (Effortlessly assuming the mount is balanced in RA.)

While you're in there you might want to put a thin stainless washer between the motor-side worm block and the oldham coupler, since there is a bit of free play in there. That will reduce the ability of the worm to slip side-to-side in the block, greatly reducing backlash. Two very thin stainless washers did the trick for me.

Coatesg already mentioned feeler gauges: I just want to pipe in to agree strongly. Get a set of automotive feeler gauges, put a suitable combination in the gap, so you can press the worm block firmly against the gauge & tighten. Then test play and backlash. Having pressed it firmly against a known thickness, you can now do controlled adjustments - loosen the block, then tighten it pressed securely against a different gauge thickness. Without the feelers it is difficult to do methodical controlled adjustments.

When you replace the milled worm cover after the adjustments, be careful that you have the rear flange aligned before tightening the set screws. If you don't there can be an air gap big enough to let you turn the set screw all the way through the cover, when it will fall onto your driveway, amidst the set-screw-sized pebbles, providing great entertainment in the search.

When you have the cover off, look at the worm. Is it silver (stainless steel) or yellowish (brass)? The newer mounts have brass worms which are milled to higher precision, and people are reporting great improvements just by replacing their old worm with the new brass one. For $50 from Losmandy, it may be worth it (I've ordered two - one each for two mounts; you don't need to do Dec, just RA).

So, step by step:

1. (optional) buy new worm gear

2. remove worm cover

3. remove RA drive motor & gearbox

4. (optional) shim worm drive with a washer under the coupler

5. Loosen worm block screws

6. Adjust spacing with feeler gauges

7. Turn worm by hand to check play

8. Repeat #5-7 until as tight as possible without causing reduction in free play

9. Replace gearbox & motor

10. Replace cover

11. Search for lost cover set screw on the driveway

Total time, minus step #11, is about 20 minutes.

Regards

Richard

[bazza]

Richard,

Many, many thanks for your lengthy and informative reply.

I feel encouraged by your description! I'll give it a go and let you know what I achieve.

Barry.

[anthony]

Hi Richard,

That was very helpful - dont suppose you can put up a picture of the oldham coupler, (never heard of one but wikipedia has!) and your suggested washer location? I think I now know where you mean but would like to be sure.

Many thanks

Anthony

[ottUp]

That was very helpful - dont suppose you can put up a picture of the oldham coupler, (never heard of one but wikipedia has!) and your suggested washer location? I think I now know where you mean but would like to be sure.

Pleased to, but it'll be one or two evenings before I can. Meanwhile, this page:

http://www.astro.uni-bonn.de/~mischa/mounts/g11_tuning.html

has a photo, about halfway down the page.

- Richard

[coatesg]

I did then do a run with PEC on, but when I try to 'refine' this in Pempro (ie. adding or subtracting curve 1 (PEC off) from /to curve 2 (PEC on) it comes out at more than 11.41 arc secs...so this appears the best I can get.......therefore my reasoning to adjust the RA Worm. Would you agree?

Or do you think at 11.41, I should be able to guide that out?

Did you try flipping the curve in PemPro and reuploading? (sometimes it's needed for the Geminis, as sometimes it doubles the PE rather than reducing it! My one needed this.)

Also, 11.something arcsec (peak-to-peak I assume?) is actually quite good for raw PE! My Titan only manages 13" peak-to-peak. If it's +/-11" (ie 22" P-to-P) then that's a bit higher than normal - however, if the error is smooth, then you should be able to easily guide it out. The problem comes where you have a big jump in the tracking.

Also, worth noting for the G11, if you have a non-harmonic error (ie - if you look at the frequency spectrum in PemPro it doesn't lie on an integer line) and it has frequency 76 seconds, then this is due to the worm bearings and can't be corrected with PEC - this is where the shimming/alignment process really helps. (info here: http://overton2.tamu.edu/rdb/sbig/76-second.htm)

Also, just to add on the feeler gauges - be careful when you tighten the worm block back up - at least on the Titan, as you tighten up, the worm moves back into the gear a little (a few thousandths maybe - but enough to cause a bind if unlucky!) - do it iteratively, until it just starts to bind - then back off slightly.

Cheers

GC

[bazza]

Thanks again Graeme and Richard,

Graeme - I did invert the curve and uploaded it to the Gemini.

Now when I try to refine I get the following:

Adding Curve 1 and 2 still increases the PE - as does subtracting, but I averaged the two curves and with a bit of shifting of one of the curves have uploaded a PE of 6.7 Arc Secs back to the mount. This is the best I have got up to now.

My only concern is that, nowhere in the manual for Pempro, can I find information on 'averaging' the two curves, so it's a bit of an experiment really.

It was reassuring though that you mentioned flipping the curve - it certainly stopped the PE doubling itself!

Next clear night oppurtunity (laugh), I'll try it out at this.

Many thanks for all the advice and guidance.

Barry.

[coatesg]

Thanks again Graeme and Richard,

Graeme - I did invert the curve and uploaded it to the Gemini.

Now when I try to refine I get the following:

Adding Curve 1 and 2 still increases the PE - as does subtracting, but I averaged the two curves and with a bit of shifting of one of the curves have uploaded a PE of 6.7 Arc Secs back to the mount. This is the best I have got up to now.

I haven't used it much myself (only a brief play really - but got down to +/-3 arcsec very quickly after I figured I needed to invert the curve when the star went shooting off on a wander!), but it looks like it can be fiddly if you start playing! You need to be careful that you don't go chasing the seeing (so everything you do needs to be averaged over a fair few worm revs (so 6 should be fine) - and check out the 76 second error in case you're seeing that. Any attempted correction for the 76sec non-harmonic will actually make things worse. Bear in mind, if you have a 76 sec error, then you're ultimately going to be limited by it (and the seeing), unless you can remove it by worm adjustment.

I'm not sure how averaging curves will help you (unless you are correcting for a non-existant error) - I think it'll could effectively land up reducing the "aggressiveness" (oooh - that's a horrible word) of the PE corrections. Though, if you are over correcting already then maybe it's going to help a bit.

If you do land up with 6 arcsec peak to peak with no big "bumps" in the tracking, then it's probably pretty good. The key thing is smoothness - if you autoguide, then as long as your guide exposures aren't *very* long (I mean many seconds - 5 to 10 upwards), and the mount is not drifting by a large enough amount in the time between corrections to be noticeable in a subimage, then you're probably fine. So depends on scale, but for 1 pixel per 5 sec say - typically that may be 1-2 arcsec per 5 sec max drift (at 1-2arcsec/pixel resolution with 5 sec guide exposures).

This is where imaging at long focal length becomes hard - the rate of drift naturally increases, so you need more frequent corrections (or smoother tracking) - but you need to find a suitable guide star and avoid chasing the seeing!

[bazza]

Graeme - thanks, really good info.

Dare I say it, but the mount shows no trace of the dreaded '76 sec error' - it's second hand, so maybe the previous owner sorted that. Hence my reluctance to physically adjust the worm block!

I'll let you know how I get on.

Thanks again,

Barry.

[anthony]

Hi Richard -

That link demonstrates it nicely - no need for a picture but thanks for the offer. That was the link I was looking for but could not find.

Graeme - as ever, excellent information from your end. (Stratford 1/2 next Sunday?)

Anthony

[coatesg]

Graeme - as ever, excellent information from your end. (Stratford 1/2 next Sunday?)

Oooh. Crikey no. Extremely unfit at present (++beer!!), though have just entered London next year, and back on the cycle to work (50min each way), so hopefully back to proper fitness again soon...!!!

Cheers,

[bazza]

Well, averaging the values of curve 1 & curve 2 didn't really achieve anything as you said would happen Graeme. But you know what's happened?

We mentioned that dreaded '76 sec error' word!!!!

The first time I ran Pempro last week with PEC off, so just checking the RAW PEC of the mount, I got no 76 sec error and a PEC of roughly 13 arc secs peak to peak, probably what I expected.

Well, with no adjustments on the worm at all and nothing changed within the set up, I ran a RAW PEC run again last night after all the faffing of trying to establish a curve - thought I'd start from scratch again.

Damn and blast, it's shot up to about 19 arc secs and there's that nasty 76 sec error reared it's head! Don't know why for the life of me 'cos I haven't touched it!

Anyway, got the feeler gauges out and I've adjusted the worm block from 0.588mm to 0.564mm between both bearing blocks and the mount.

There doesn't seem to be any backlash to take up though -I've checked I can turn the worm OK - it's not too tight but there doesn't seem much more movement available for adjustment on the bearing blocks.

The motors run no problem with no error messages.

Hopefully, I'll try to give a run of the RAW PE tonight.

Barry.

[coatesg]

Take it you ran both tests pointing at the celestial equator? (you get less PE at higher (or lower) declinations - makes imaging near the pole easier on tracking - but you need to be better polar aligned... )

The 76 second error tends (though depends a lot on the worm and blocks) to come from one of the blocks being slightly twisted, or too high (same kinda error really - just in a different dimension). It's tricky to reduce, but I think one of the sites above gives an indication on how to reduce it.

Of course, the trouble is that you need a clear night to try and sort it out!

[bazza]

Graeme, I was using a star near the celestial equator as you said.

I suspect the 76 sec error might have come from some play in the Oldham coupler...not sure how it got there, but hey hum!

After working night shift for two nights and the third night up until 4am with Pempro...well, last night I was shattered, so didn't get out.

No rush though, I will sort it......if it kills me!!

You know, I can see a buisness oppurtunity here....you become a freelance troubleshooter of peoples equipment!!! Trouble is, clear nights are rare, so you'd have to charge the earth to mke ends meet! Then of course, the customer would have to pay travelling costs and put you up for the night (or two), yep...there's an opening there that needs developing!!

Barry.

[bazza]

Last night was the first night that's been clear since I adjusted the worm block last week.

I'm very pleased. The overall PE is down to about 6 Arc Secs peak to peak. There is still some error evident at 76 secs, however, it's dropped dramatically - from 3.85 Arc secs to 1.05 Arc secs. I should be able to improve on this.

I'm guiding the errors out though, so up to 5 minute exposures seem easily within grasp.

Trouble is, when I've made such a drastic improvement, I'm lothe to adjust it again for fear of making it worse! As long as I keep a running record of adjustments though - it should be OK.

Thanks to everyone who has helped me with this, especially Graeme and Richard.

Barry.

[coatesg]

6 arcsec P-to-P is pretty impressive 8)