Spring loaded worms?

[Horwig]

1 hour ago, Rusted said:

I'm still doodling potential motor/worm mounting ideas to include the clever "flexor" hinges.
I need these "hinges" to be much nearer the worm housing and further away from the motor housing.

The belt drive can easily manage the tiny movements required for the worm to maintain mesh.
New, longer belts are readily available, if needed.

I'm moving the worm block and motor as one unit, the 'hinge' is quite some way back from the worm/wheel.

Rigidity is paramount. My dec assembly is quite sturdy, not so sure about the R.A. I've had to build up a 'spacer' from 2"x4" box section to put the worm at the edge of the wheel, not ideal.

Huw

[Rusted]

How about simple bracing between the base and the worm assembly?
Go kart track rods with built in turnbuckles and spherical joints?
If it still needs such stiffening?

 

[Horwig]

22 hours ago, Rusted said:

How about simple bracing between the base and the worm assembly?
Go kart track rods with built in turnbuckles and spherical joints?
If it still needs such stiffening?

Ahh, what's missing in that photo is a plate across the front, joining both side plates, just behind the box section. This solidified the mount enormously, maybe a further bracket from the worm block spacer box to that would make sense.

Ta

Huw

[Rusted]

Yes, it would make good sense to triangulate the open, Beacon Hill style frame as you have done.
A prism frame has no stability across the uprights unless it is triangulated by reinforcing plates on all three, square faces. [Stressed skin effect.]

The problem with most mountings is that the manufacturer has to allow for almost universal, PA altitude adjustment.
Which means that some (obscenely expensive) high end mountings abuse simple, blade-like, PA, support forks.

Because I had no interest in lightweight mobility I used a PA support fork in 20mm solid metal.
But, further, ensured the fork tines were braced upright by clamping across the massive, square section, PA housing.
This was only possible by offsetting the PA altitude, clamping stud to clear the PA shaft.

The Fullerscopes MkIV shot itself in both [fork] legs by insisting on a centrally placed, PA, altitude pivot. [Cosmetic handicapping!]
Which meant undersized, screw threads in the very soft, cast metal, PA cone.
This denied the application of a massive clamping force via a long, through stud. 
These silly screw threads are probably all as badly stripped as the silly, little, altitude locking screws by now.
The only way a MkIV [and many other mountings] makes any sense is with an altitude adjustment turnbuckle.

I considered drilling a large hole in the underside of the MkIV's PA casting.
To allow the insertion of large [hex socket] PA altitude bolts to face outwards through the fork tines.
"Fishing" such bolts through such an arrangement would be a nightmare. So I did not proceed.
The MkIV's PA fork "ears" are weak metal and hollow section to boot. So easily crushed unless stuffed with packing washers.

I should say that I am grateful to the MkIV for teaching me to avoid all of its serious pitfalls as I designed my own mounting in my head.
Parallel reinforcement by multi-tasking every component proved to be the way ahead at lowest cost.
Where I could not afford castings and very large, box section profiles, I used multiple compression studs in all 3 planes.
Each component leans on all the others to become greater than the sum of their parts.

Unfortunately I abandoned my own principles when it came to supporting the [now glaringly inadequate] Beacon Hill, worm housings.
I badly underestimated the linear and torque loading involved on the motor/worm housing.
Made worse by the large cut-outs in the motor box sections. To allow the drive motor power sockets to clear the box during insertion.
"Cosmetic handicapping" is the fine art of making something look pretty, for its own sake, even if completely fails to function mechanically.
 

[Horwig]

If I started this project all over again, my polar wedge would look totally different.

The hinged plate model of Beacon Hill mounts might heve ben adequate for visual, but for high accuracy imaging it's pants. (My highest resolution imaging is 0.4 second per pixel on this scope).

The lower part of the wedge is a commercial gate hinge welded to two steel plates, the fulcrum created by the adjusting rods at the top of the plates pushes the hinge apart at the bottom, any imbalance in the load when moving between East and West of the Meridian rocks the mount, hence the bracing plates. But it still sucks.

If I were to start with a clean sheet, I reckon something like this as a polar axis would work:

http://www.astronomyblogger.com/homemade-stuff/homemade-equatorial-mount/

but naturally, in 20mm ali, not 18mm ply😀

 

Naturally, with so much invested in the mount already, and with so few clear nights this far West, it's a law of diminishing returns, so I'm fettling what I have, not stating anew.

I've spent a fortune in this last year on an encoder for the Polar axis, this is going to be my last big spend on this mount.

 

Huw

 

Edited December 8, 2020 by Horwig
typo

[Rusted]

Thanks Huw.

Looking at your mounting I see that you have the B.Hill channel section, worm housing backed by a thick plate.
My worm housing channel is arranged "sideways on." So likely much more prone to flexure.

I have dismantled and rebuilt the RA worm/motor housing several times this morning.
I keep trying various [scrap metal] profiles against the assembly to see if I can considerably increase the overall stiffness.

As the Irishman is supposed to have said: "I really wouldn't start from here!"

Tell me more about this RA shaft encoder? It would be great not to be limited to [slow] motor slews. [Minutes per 90°!]
I understand AWR offers encoders for its drives [at a price] but I haven't researched the issue further due to the cost.
 

[Horwig]

On 09/12/2020 at 13:24, Rusted said:

Tell me more about this RA shaft encoder? It would be great not to be limited to [slow] motor slews. [Minutes per 90°!]
I understand AWR offers encoders for its drives [at a price] but I haven't researched the issue further due to the cost.

This might be getting OT (😀)

My mount is controlled by a Sitech II system, using dc motors with encoders.

The beacon Hill RA worm has 'wonderful' periodic error, so I've mounted a second hi-res encoder directly onto the end of the RA shaft.

They are not cheap at about £800, but give very smooth tracking with no PE.

This is the raw unguided mount in blue, with guiding in yellow, and guided with encoder in pink:

 

Edited December 10, 2020 by Horwig
typo

[andrew s]

@Horwig what the y scale on your plot?

Regards Andrew 

[Horwig]

15 minutes ago, andrew s said:

@Horwig what the y scale on your plot?

Regards Andrew 

Arc seconds per pixel

[Rusted]

6 minutes ago, andrew s said:

@Horwig what the y scale on your plot?

Regards Andrew 

Hi Andrew,

Try double clicking brings up an enlargement.
It looks as if the raw Beacon Hill drive error is +/-15" with a 5 second period. 

Huw,

Is the £800 you quoted the full cost to be up and running with a commercially available set-up?
Or just the cost of the encoder?

[Horwig]

1 hour ago, Rusted said:

Is the £800 you quoted the full cost to be up and running with a commercially available set-up?
Or just the cost of the encoder?

the £800 is the cost of the encoder only (from memory), Sitech II is 900us, motors new about 150 us each, I got my motors second hand off ebay.

 

Huw

[Rusted]

14 hours ago, Horwig said:

the £800 is the cost of the encoder only (from memory), Sitech II is 900us, motors new about 150 us each, I got my motors second hand off ebay.

Huw

Thanks. Much the same as AWR charge for encoders. [From fading memory.]

I'm on the points of making a new worm [trough-shaped] housing.
If I could get the worm off the [stumpy] Beacon Hill shaft I could turn one to a sensible length between centres.
I suppose I could turn a new worm, between centres, but haven't the patience darkest midwinter.

By the time I've paid 25% VAT on top of everything and all the additional import and customs clearance charges, the E Byers wormwheel sets would be 50% higher.
Gorgeous, they may be, but I'm not sure I could cope with that much bling in the observatory.
During solar imaging sessions it would be like one of those disco mirror balls during a Meridian Flip.
Ironically the same website you linked to sells wormwheel sets. No idea as to their quality though.

[Horwig]

Just a thought Rusted, is any of the movement you were showing in the video due to movement in the bearings on the axis?

My mount, as was, had pillow blocks without any preload applied, and I was getting appreciable sideways movement of the shaft.

I've replaced them on both axes with pre-loaded taper roller bearings and precision cut shafts, that, and the work on the worm block hinges solved the wobble for me.

 

H

[Rusted]

1 hour ago, Horwig said:

Just a thought Rusted, is any of the movement you were showing in the video due to movement in the bearings on the axis?

My mount, as was, had pillow blocks without any preload applied, and I was getting appreciable sideways movement of the shaft.

I've replaced them on both axes with pre-loaded taper roller bearings and precision cut shafts, that, and the work on the worm block hinges solved the wobble for me.

 

H

Thanks Huw.

I checked and without any deliberate end loading the flange bearings had no shake when I fitted a 3' pipe to the bottom of the PA and gave it a good workout.

My worm housing will soon be fully enclosed in an aluminium "trough." I'm just waiting for some M6 SS CSK screws to be delivered to hold it all together.
The motor mounting plate and rear reinforcement will close off the ends of the Beacon Hill channel and provide some serious resistance to linear bearing movement.
The motor will also be shimmed tightly in its alloy, box profile to help to stiffen the whole motor/worm assembly. I'll brace across the end of the open motor slot too.

Some pictures of the mock up so far:

 

[Horwig]

Nice one, battleships were produced with lighter stiffening😀

[Rusted]

11 hours ago, Horwig said:

Nice one, battleships were produced with lighter stiffening😀

Better to squash any chance of flexure in one go. Rather than by lengthy trial and iteration.

[Horwig]

True, I've been trying to attack this itereratively, and it;'s sort of working.

My latest challenge is a problem thrown up by the high res encoder on the polar shaft.

The dc gearmotor is mounted to the worm block plate by its face plate, trouble is, under drive, the 3.5mm ali plate flexes, just a little, but it's enough, the encoder on the motor, and the one on the shaft fight, and an error of an arc minute or so throws the servo controler into a tiz, it tries to correct, and results in a 4 arc second bump, every few minutes, and yes, at 0.4 arc seconds per pixel, that's a lot.

It's too cold and damp here to start diss assembling the mount, it can wait till warmer weather returns.

I'm interested in your idler wheel tensioning the belt, is it adjustable, and if so how?

 

Huw

[Rusted]

It sounds if you are experiencing closed loop,  positive feedback on your drive system. Best avoided!

My idler wheel is a simple, turned brass roller running on a polished, stainless steel, bolt shank.
The roller was sized to reduce belt slop and no more. I tried a larger roller but it was much harder to insert it.
It seems I don't need a tension idler now the motor has moved relative to the worm. Or vice versa.

36-37F here, constantly overcast and downright unpleasant to spend hours "fiddling" in an unheated building.
I have retreated to the comfort and safety of the computer chair.

[Michele Scotti]

Hi Rusted, sorry I missed some of the last messages.

Did you identify the source of the substantial wheel movement as per your vid here: https://youtu.be/bWKQBCOkrNw?t=14 ?

Also, I recall you ordered angular bearings - were you able to instal them? Did you notice any improvement?

[Rusted]

18 minutes ago, Michele Scotti said:

Hi Rusted, sorry I missed some of the last messages.

Did you identify the source of the substantial wheel movement as per your vid here: https://youtu.be/bWKQBCOkrNw?t=14 ?

Also, I recall you ordered angular bearings - were you able to instal them? Did you notice any improvement?

I have the angular contact bearings and have fitted them.
The images above show my partial mock-up of a heavily reinforced RA worm housing.
Between waiting for new screws in the post and bouts of severe dizziness I have made little further progress.
I have visited the doctor and had a negative virus test and negative on routine blood tests.
A low sun and weeks of overcast sky have taken the pressure off the burning desire to finish the mounting improvements.
So I am preparing the ground for a much bigger dome while I am still upright.

[Rusted]

A positive update:

It took an online dealer a month to deliver new and slightly longer drive belts.
I had already fitted the angular contact bearings recommended here.

Today, I had my first solar imaging session in ages. Of only a few minutes before it clouded over.
The newly reinforced drive system is a great improvement. It stops and starts instantly. 

The images show how the simple addition of bolt on profiles to the motor housing have helped.
I may still add another profile to rest against the end of the motor to retain the rear worm bearing.

For scale: My home GEM mounting has an 11" RA wormwheel and50mm solid stainless steel shafts.

[Merlin66]

Well done mate!

[Horwig]

Very nice work, they look just the job.

 

Huw

[Michele Scotti]

Well done - much of satisfaction when things turn for the best.

Would you be able to make another video just like those before to check what it's fixed and what are the residual contributions.

Btw is now the motor/worm assembly fixed solidly and not sprung anymore, right?

[Rusted]

Thanks for all the interest.

It was never truly sprung. For most of the time.

I tried various springs to pull the worm into the wheel and then [usually] locked the pivot screws.
That didn't work well. Nor did it work with the springs and loosened pivots.
It would have needed a very firm pivot with no play at all. I just used a fixing bolt shank as the pivot. Sloppy!

The latest mods are a great improvement but I still want to make a new, rear worm bearing steady.
We are enjoying extensive frosts so working with metal is painful or clumsy due to wearing gloves.
My workshop is unheated and often well below zero even after it has warmed up a few degrees outside.

Thankfully I am no longer suffering from repeated stalls nor backlash.
My massive mounting used to feel springy or sloppy when the telescopes were moved manually.
This was obviously due to the drive sloppiness, backlash or worm housing flexure. Or all of them!

The wormwheel/worm set-up could not restrain the heavy loads being applied by my long and heavy refractors.
So I looked at commercial mountings with [unbelievably] high [claimed] load capacities.
Despite their being relatively tiny and lightweight compared to my huge and massive GEM construction.

These mountings [usually] seemed to restrain their smaller worms and wheels rigidly within box castings.
So I tried to copy that particular feature using simple, bolt-on profiles. Still incomplete.

I should have abandoned the flimsy Mount Beacon worm housings and line bored far more solid sections.
Now I have a far simpler alternative. I like to keep my constructions simple.
So anybody foolish enough to copy them won't need machine tools.

I'll make a new video once I have completed the rear bearing support.

The image shows an early set-up with three refractors. 180/12, 150/8 and 90/11.