开云体育

Just testing with RA and DEC control turned off in the guider. Running with multiple stars. And to be sure that OnStep has no secret cross talk, disconnected the DEC cable.? After reading about the 76 second cycle, guess what we have here.? It's 75 seconds according to my estimate but I may be off.



Let me check what happens if I tighten the blocks thereby disabling the DIY spring loaded worms.? Regardless, I need to check the block alignment.

When I tighten the outer block (so both blocks are tightened) I have the standard non-spring-loaded Losmandy G11S mount.? It gives me an extra harmonic of 225 seconds (=3*75).? Hmm.?



I can't get the size of the images right, maybe there's a limitation.? Here are just the error plots:

This one has the DIY springs enabled.? The outer block is floating and is pressed by a strong tangential spring and a weak radial spring.


This one has both blocks tightened.? This is the standard non-spring-loaded G11S - except that I had my hands on all screws I could find.


Clearly the DIY spring loaded worms got rid of one harmonic.? But what would cause it, thoughts anyone?

Henk,

You have probably seen the discussion. Other than replacing the bearings, or honing out the right bearing block fit and compressing with a Belville washer, you might try squeezing the blocks toward each other along the worm axis. This may allow the balls to ride smoothly along the inside track, and not ride on the outside race where there could be some rough spots. That seems to be the best theory by consensus.

Henk,

Just saw your additional post made while I was typing. Seems like you may have two separate occurrences with overlapping periods. So a glitch with both bearings? Wondering if some additional block to block tension may be needed. Or perhaps to much pressure on the worm towards the ring gear is overcoming the axial tension, and you are again riding too much on the outer races. Perhaps starting a bearing purchase may be in order. I bought a pack of 10 the last time round so I would have extras at home to experiment with, or to see if one is better than another.

To make sure what the frequencies are I imported Ekos' .csv file of the 2nd plot into Scilab and ran an optimization over the sum of a linear and 2 sines (8 parameters altogether).? The outcome is? amazingly clear:? One cycle is 75.63 seconds, the other 240.62 seconds.? The amplitudes are 1.66 and 3.61 arc seconds, respectively.? The fit is like this:


So, by relinquishing the spring loaded worms I get stuck with a significant worm cycle of 1 turn per 4 minutes, Earth's rotation.

Wow!? I thought I was provided with a "high precision worm"?

Ok well my DIY spring loading apparently takes care of it.? So my next quest is to fix the 76 second cycle due to the ball bearings in the worm blocks.? Ugh.

Hi John,? thanks for the advice and the pictures of your bearing fitting in Alan's thread.? I believe the explanation by John Goetz about the 76 second cycle. Just spring loading will not fix it because the bearings are still misaligned.? Both the worm and the crown wheel are uneven so the spring loaded worm must be moving constantly to compensate.? Therefore the bearings must follow so Belleville washers are the solution.

If I widen the holes in the blocks I should be able to reuse the bearings that I have, should I not?? If yes, why did you buy new bearings?

Also, my outer block bearings should realign themselves because mine float and are pressed by a tangential spring.? So I think the inner block needs the Belleville washer not the outer.? Maybe it's best to do both if there is enough space.? Any thoughts?

Hello Henk, the high precision worm should give a periodic error of about +/-5” so it all looks ok to me (and the period would be 239 seconds). I found misalignment of motor and worm could give extra harmonics (and this would include any bearing misalignment), so I would also check if everything is aligned well. To get lower values you would need to use PEC. Don’t know what focal length you are imaging at but autoguiding should work well with that level of periodic error.

Thanks Ian, that's good to know.? I think the inner block was misaligned because I had to make the flex couplers of my stepper fit, and those have some stiffness while the motor and worm were not perfectly aligned also.? Let me see if I can do a better job at that.??

Maybe an autoguider can handle it but I started looking into it because I get over 1" RMS and I expect this mount to do better.?

I added the spring loading because the mount had more backlash than I liked.? Tightening the worm would simply lead to the mount binding up, I could feel it manually.? The ring gear is just not round enough to have zero backlash.? Looking at the first graph (spring loaded configuration), it doesn't look much better than the second but at least it keeps the mount from binding up when slewing.

So I can pick my poison.? I can align the blocks as best as I can, tighten them, abandon the spring loading, allow backlash (hardly a problem for RA when tracking) and put up with the 240 second cycle and the backlash for slewing and plate solving.? Or I can keep the spring loading, improve it so the 75 second cycle is more regular than what we see in graph 1, and put up with the 75 second cycle.? Or I can drill out the holes in the blocks, apply Belleville washers and spring loading.? Let me try the non destructive options first.?

There is actually another option, namely to stack every 10 seconds and use no PEC or autoguiding at all.? That way the errors will never exceed the errors in a 10 second interval.? The technology is available, I know someone who does this with his ATIK.? Perhaps the live stacking of Sharpcap can do this for general cameras.? If not now, some time in the near future.? The computing power should be there already, it is just a matter of time until people start wanting this feature.

Henk,

The bearings are pretty inexpensive, about $10 (US). Since the 76 second error is related to the bearings, new will be best. Old bearings could have wear spots or may have been slightly warped during original installation or removal. If the block seats were too tight, or changed with temperature (as per Michael's explanation), the outer races could be slightly out of round, or bent. This may cause the ball hang ups, as per the white paper. So the same bearings may retain their deformations, even after you change the seating or springs tensions.

As per Michael's advice, it seems the right bearing is the one to float, to allow Belville washer tension, but also to permit some slight flex along the worm axis. The left one should remain as the solid base. I suppose that seating could be loosened as well to prevent some crushing, but I have not heard much discussion about that. But installation of the left is still critical, being careful to evenly tap into place without pinching, so tap or push slightly around the edges gently to wiggle into place.?

I have followed Michael's instructions, and read the whitepaper and discussions, so I am following the instructions of those experts. Figuring this out on one's own would take a quite a bit of study and research. Good luck.

Regards,

John

Also, my outer block bearings should realign themselves because mine float and are pressed by a tangential spring.? So I think the inner block needs the Belleville washer not the outer.? Maybe it's best to do both if there is enough space.? Any thoughts?

Yes - purposely set the blocks so that there is too much axial play in the worm.
Bias east heavy to test the right block, west heavy to test the left block

Good idea Alan, thank you.? I will hold off on the Belleville washers for now since I just changed my DIY spring loading such that the blocks and bearings should be aligned properly (see my other thread that I just started).? But if I still have problems the Belleville washers and smaller bearings will be next.

On Sat, Apr 3, 2021 at 11:28 AM, Michael Herman wrote:

Stuart Hutchins, a very experienced and knowledgeable group member, has been explaining the details of bearing seals, concerns about polishing dist entry, and precision belt drives to me. He just suggested that you could mount the R4 bearing between a pair of R4 Belleville spring washers to compress and hold the bearing on a 1/4 inch precise mandrill, then use like a drill to precisely spin the bearing...and some fine sandpaper to buff down the bearing OD more consistently.? I have not tried this nice idea yet.? He was concerned that the polished bearing OD needed to be precisely round, but since the OD of the bearing merely needs to slide, not rotate, I don't feel that level of roundness perfection is necessary.? The inner races must be precisely round as they must roll without "rumble" which is why I prefer the highest ABEC quality...but even so I get extra bearings and feel them for smoothness.

Hi Michael,

After reading through your info, I was considering how to build a rig to get a circular polish on a bearing. I came up with the following using a 1/4" bolt. It consists of a nut, washer (non-Bellville), rubber washer, Boca bearing, rubber washer, washer, and a second nut. By tightening the nuts toward each other, you can seal the bearing between the two rubber washers to prevent any sanding debris from entering. Then when done polishing you can wash with some solvent without having any residue get in between the races. I tried to spin it in a drill tonight and it seems decently centered with little wobble. Now I could hit the outer race with a Jeweler's file or a small strip of sandpaper while rotating.

I measured a new bearing and a new bearing block and their OD and IDs, respectively were both 0.625". Now I am wondering how much tolerance to apply for a proper sliding motion of bearing inside block. You mentioned the blocks will contract with cold, so thinking what the final polished bearing OD should be. Take off 0.005"? What do you think?

John

Hello Michael,

Thanks for your thoughts. I guess I will just have to try a few polished bearings and see how they slide with decreasing ODs. Once I get one that easily moveable, I should put it the freezer and see if it still moves with the same freedom.?

I am not suffering from the 76 second error problem, but hope I am giving some useful info to those that are. Perhaps with a standardized bearing procedure., we can offer something to these folk who need that.

John

My opinion is that you need just enough clearance to allow the bearing to tilt inside the block by a tiny bit.? Maybe about 0.001 - 0.002" clearance

Alan/Michael/all:

Just tried the bearing spinning rig today with success. I placed the assembly in a drill press at 100 RPM and started trying small metal files and types of sandpaper to hone down the bearing. Surprisingly the outer race is fairly hard steel and doesn't shed diameter easily. Finally I used the corner of a sharpening stone, like that used for scissors or straight razors, and that was the trick. Subsequently I polished with 200 then 600 mesh sandpapers. It looks like reducing OD about 0.002" is the point where the bearing will begin to slide freely inside the block (see pics). The outside race also seemed to be virtually round and smooth after this process. Before pulling apart the rig, I sprayed away all the sanding dust with some WD-40 and then some compressed air. Upon finger turning the finished bearing I did not sense any hang up or grittiness inside.

Perhaps this is a new way for reducing the bearing OD to allow the Belville washer to do its thing. I found this a bit easier than turning down by hand. Perhaps a bit more OD reduction might be needed for those who wish some off axis tilt, but too much might start the bearing turning inside the block. But I think polishing down to 0.623" is the place to start. But just a note a caution for those who are not that familiar with moving machinery: the spinning drill assembly could easily cut the skin, or send particles into the unprotected eye. Use precautions.

Hope this info is of use.

John

On Tue, Apr 6, 2021 at 12:00 AM, Michael Herman wrote:

Very nice work.

You have nicely demonstrated a faster and more consistent way to polish down the R4 bearing OD.??

?

As you say, you must always be very careful around any power tools.? We are in a hobby for fun, not for danger.? So care if the hands and eye protection, and a face mask to avoid inhaling dust are important precautions.?

?

That said, I have used my bargain priced Wen brand drill press to polish ODs down (on brass rods) before, so this is an accessible way of polishing small round items at rather low cost.? The drill press has no explicit cutting tool in it...just a round bearing.??

?

The only thing doing the cutting is an emery cloth or fine sandpaper.? (I like the 3M "wet or dry" paper because the grit is well embedded in a kind of string waterproof adhesive.? You can wet the paper to keep polishing dust down.? ). The sandpaper can get hot, but it generally won't cut you.??

?

The notion about extra benefits of allowing the bearing (at least the far bearing away from the gearbox) to self adjust it's angle in the worm block...these are new ideas.? We don't really yet know a "spec" on the best amount of OD slack to use.? Ideally you don't need any slack at all if all the parts are perfectly made.? Most people never see the 76 sec oscillation so never need this extra workmanship.??

?

A theory for the riddle:

? ? "why is a single Belleville washer under the (OD loose) far end bearing sufficient?? Why don't you need that spring under both bearings? "

?

An idea:?

?

Many people preload their RA axis with an intentional East bias. Say a weight on a cord, wrapped so the weight hangs to the East side of the RA axis.? We have discovered that the East bias eliminates a vibration ("chatter") that occurs if you bias to the West instead.??

?

When you preload the RA axis (ring gear) with an East bias, this forces the worm to move West.? That presses the worm against the West "far" end RA worm bearing center, and leaves the near (near the gearbox) bearing center looser.? So the West far worm bearing becomes the more critical surface for the RA drive: the worm is mostly rotating on that ball bearing race surface, not in the bearing closest to the gearbox.? ?That's why eliminating rumble of that far bearing becomes more critical for low noise sidereal tracking.??

?

This idea might explain why we see a reduction or elimination of the 76 sec oscillation when we reduce the OD (allowing tilt) and put a Belleville spring behind just the far worm bearing... the spring allows the bearing when tilted to still ride on a "self-leveling" seat.? ?

?

Otherwise, reduction of the 76 sec oscillation should only be effective some of the time...but it seems to be working all the time.? The East side bias effect could explain this riddle.?

?

By this reasoning, using a West side bias might create a 76 sec oscillation even with a Belleville spring under the far West bearing? ...there's another experiment to try!? ?

?

All the best,

Michael

?

?

?

?

?

On Mon, Apr 5, 2021, 10:01 PM John Kmetz <jjkmetz54@...> wrote:

Alan/Michael/all:

Just tried the bearing spinning rig today with success. I placed the assembly in a drill press at 100 RPM and started trying small metal files and types of sandpaper to hone down the bearing. Surprisingly the outer race is fairly hard steel and doesn't shed diameter easily. Finally I used the corner of a sharpening stone, like that used for scissors or straight razors, and that was the trick. Subsequently I polished with 200 then 600 mesh sandpapers. It looks like reducing OD about 0.002" is the point where the bearing will begin to slide freely inside the block (see pics). The outside race also seemed to be virtually round and smooth after this process. Before pulling apart the rig, I sprayed away all the sanding dust with some WD-40 and then some compressed air. Upon finger turning the finished bearing I did not sense any hang up or grittiness inside.

Perhaps this is a new way for reducing the bearing OD to allow the Belville washer to do its thing. I found this a bit easier than turning down by hand. Perhaps a bit more OD reduction might be needed for those who wish some off axis tilt, but too much might start the bearing turning inside the block. But I think polishing down to 0.623" is the place to start. But just a note a caution for those who are not that familiar with moving machinery: the spinning drill assembly could easily cut the skin, or send particles into the unprotected eye. Use precautions.

Hope this info is of use.

John

Michael,

By loading the outer bearing's outer shell you are loading all three pieces, putter bearing, worm and inner bearing. So the side of the mount should not make any difference and the 76 second error will not be present on either side of the pier if done correctly and no other issues are present in the mount or drivetrain.?
?
--

Chip Louie Chief Daydreamer Imagination Hardware?

? ?Astropheric Weather Forecast - South Pasadena, CA?