Adjusting the RA worm blocks on an older Losmandy G11

[AMcD]

An older Losmandy G11 mount can develop an oscillation of 76 seconds in its periodic error.  Having posted a PHD2 guide log on the PHD2 forum in an effort to refine my guiding, and had it analysed by Brian Valente no less, it transpires that my trusty old G11 has this dreaded 76 second oscillation.

Because this 76 second oscillation is not an integer multiple of the RA worm cycle it cannot successfully be guided out (although Brian did kindly suggest some settings on PHD2 that he described as the “brute force” solution).  In the circumstances, any effort to reduce permanently this non-integer oscillation in the PE has to include the mechanical element.

Adjusting the standard RA worm blocks on the older G11 is difficult because of its design.  The original worm block design has many elements, making it easy for those elements, and in particular the worm blocks, to become misaligned in any or all of their axes.  Whilst an option is to buy an after market one piece worm block from Losmandy that eliminates the issue, these are expensive and there is currently a waiting list. 

So, before getting in the queue and spending money on the after market solution, I decided to try and improve things by adjusting my RA worm blocks, having scoured the Web for advice and procedures.

As with all my tinkering, I only attempted this because my mount is now aging and was already second hand when I got it in the mid 2000s.  Attempting such adjustments on a newer more expensive mount is a different proposition altogether.   I used the following process.

First, measure the periodic error to ensure you have a reference against which to measure the results of the adjustments you make.  As I say, I sent one of my PHD2 logs to the PHD2 Forum for assistance with my guiding and received advice on what the periodic error looked like.  This showed the 76 second oscillation:

 

Ensure the mount is balanced in RA (the end of the process a slight East side bias to the RA balance can be re-introduced to aid more reliable tracking and guiding).

 

Measure the side to side play in the RA axis at the end of the counter weight shaft.  There should be a small amount of play to prevent the RA gear binding.  The play at the end of the counterweight shaft should be slightly less than or equal to 1mm.  My play measured a little over 2mm.

Remove the the cover from the RA worm by backing off the two tiny hex screws.

 

Remove the drive motor so you can turn the worm by hand in order to check the free play. 

 

Examine the orientation of the worm blocks in each axis.  The gap between the back of the motor side worm block and the RA gear housing on my G11 was 0.84mm but the gap between the back of the other worm block the R.A. gear housing was 0.98mm, so we’re well out of alignment in addition to there being a lot of play in the RA axis.

 

Also check whether the worm blocks sit flush on the RA worm base plate.  If they do not then they may require shimming.  There is a good guide on how to make that adjustment if you need to before undertaking the steps below at http://www.wilmslowastro.com/tips/g11gemini.htm.

Clamp the worm blocks together using a small toolmaker's clamp so that they are held together when you loosen the hex bolts holding the blocks in order to adjust them.

 

Loosen the large hex screws under the worm blocks.

 

Using an appropriate feeler gauge placed between the back of the worm block and the RA gear housing, adjust the spacing between the back of each of the loosened worm blocks and the RA gear housing using following method.

With thumb pressure press each worm block against the chosen feeler gauge before re-tightening the large hex bolt. Take care when tightening the hex bolt as this action can move the worm ever so slightly back into the RA gear, potentially resulting in binding.  Re-tighten in stages with the feeler gauge in place.

 

Test the rotation of the RA worm by hand.  The aim is to make the fit between the worm and the RA gear as tight as possible (as indicated by the amount of play in the R.A. axis at the end of the counterweight shaft) whilst still being able to turn the worm effortlessly by hand on the balanced mount.

Now that you have adjusted the worm blocks to a known tolerance (as indicated by the chosen feeler gauge) you can continue to make iterative adjustments, using thicker or thinner gauges, until you achieve the optimal adjustment for fit and effortless turning.  

I started with a gap of 0.75mm for each worm block (at which the worm could not be turned by hand at all and there was no play in the R.A. axis at the end of the counterweight shaft) and finished at 0.88mm for each block (when the worm could be turned effortlessly by hand and there was less than 1mm of play in the RA axis at the end of the counterweight shaft).

Once you have achieved optimal adjustment make sure the hex bolts are properly tightened.  Again, tighten in stages with the feeler gauge in place to avoid unintentionally moving the worm block.

When reassembling the motor connection there is an option to install a thin stainless washer between the motor-side worm block and the Oldham coupler that connects the worm shaft to the motor, to eliminate the free play that exists at that point.

 

Now check that there is no binding  for the whole of the RA gear.  Just because you can effortlessly turn the worm at the point you are at on the RA gear does not mean this will be the case on the whole circumference of an aging RA gear. The only really reliable way to do this is to reassemble the motor and run an RA slew from RA limit to RA limit, but it is better to find out at this point that the worm and RA gear are binding than during an imaging run.  I found my RA gear was indeed binding at one spot so I removed the motor and, again using the feeler gauges, repeated the process above to back both the worm blocks off the RA gear housing by a further 0.02mm to 0.90mm.  This solved the binding.

 

When you replace the worm cover ensure that you have it properly aligned before tightening the set screws. 

As a result of this work I have managed to reduce the play in the R.A. axis from over 2mm to under 1mm.  Now I have to wait for a clear night to see whether I have also achieved any improvement in the non-integer oscillation as a result of correcting the mis-alignment of the worm blocks, or at least an improvement in guiding performance more generally.  More news in due course...

[RafalT73]

Hi, thank you very much for sharing this material. I fight with proper alignment many times but I have never used feeler gauge. It is so simple way and helps me a lot!!! thank you

[AMcD]

7 hours ago, RafalT73 said:

Hi, thank you very much for sharing this material. I fight with proper alignment many times but I have never used feeler gauge. It is so simple way and helps me a lot!!! thank you

No problem, glad it helped.  This method has improved guiding on my mount significantly.  

[markse68]

Hi AMcD, where does this PE come from then? is it the mount chattering on the backlash?

Regarding your spacing measurements, on my GM8 the motor side worm block fits in a tight fitting hole but the other side is in a larger hole allowing the whole assembly to be swivelled into mesh so the gaps between blocks and the gear housing wouldn't be the same. Isn't the G11 the same?

The other thing I noticed was that the mesh is heavily affected by the tightness of the clutches on my GM8 (think it's the same mechanism for the G11?) so good idea to set it at their normal use tightness as backlash increases as you loosen them.

I use a parallel to clamp the 2 worm blocks for setting the mesh and then adjust the end play out with brass tipped grub screws I fitted to the far worm blocks. Takes a lot of the hassle out of the process.

Mark

Edited September 28, 2021 by markse68

[AMcD]

36 minutes ago, markse68 said:

Hi AMcD, where does this PE come from then? is it the mount chattering on the backlash?

Regarding your spacing measurements, on my GM8 the motor side worm block fits in a tight fitting hole but the other side is in a larger hole allowing the whole assembly to be swivelled into mesh so the gaps between blocks and the gear housing wouldn't be the same. Isn't the G11 the same?

The other thing I noticed was that the mesh is heavily affected by the tightness of the clutches on my GM8 (think it's the same mechanism for the G11?) so good idea to set it at their normal use tightness as backlash increases as you loosen them.

I use a parallel to clamp the 2 worm blocks for setting the mesh and then adjust the end play out with brass tipped grub screws I fitted to the far worm blocks. Takes a lot of the hassle out of the process.

Mark

 

 

Oops - I have just noticed that those graphs in the post reflect the pre-adjustment situation!  I am not sure why multiple copies made it onto the end of the post  as well as appearing at the beginning. 

Following the adjustments I made per the post, the amplitude of the 76 second oscillation was reduced from the nearly 8.3" shown in those graphs to an amplitude of a little over 0.3" as measured by the Frequency Analysis in PHD2.  As a result, my RMS of 5.9" has been reduced so that I now regularly achieve an RMS of between 0.55-0.80" when guiding which, whilst not matching some higher end mounts, is much, much better than it was and well below the very best nights of seeing we get here.

The G11 is not the same as the GM8 as far as I am aware - with both holes allowing movement.  Part of the problem with the older G11, and the source of some of the PE pre-adjustment (the rest coming, I think, from inherent variations in the worm gear and mesh) is that so many of the components that go to make up the worm gear / block assembly in the G11 can move relative to each other.  Normal tightness on the clutches also helps on the G11.

I do like your elegant solution of the parallel and the brass tipped grub screws - I must remember that for the future 😀