Alternative to gears?
Posted 24 February 2014 - 08:09 AM
not sure if this is the right section… but has anybody tried to use rubber wheels as an alternative to gears or timing belts for transmitting the motion from the RA motor to the mount's worm gear? I am not planning to get rid of the worm and wheel, I just would like to replace the gears between the worm and the motor with two rubber wheels to reduce some annoying high frequency oscillations I can observe during tracking (see this thread ). I tried to search around but could not find any info: in theory rubber wheels should transmit the motion much smoother than gears… so, where's the catch?! Not enough friction? Slippage? Not accurate enough?
Posted 24 February 2014 - 08:18 AM
Check this link for more info on different drive types:
Posted 24 February 2014 - 09:06 AM
Better to use a rubber timing belt and toothed pulleys.
Posted 24 February 2014 - 11:04 AM
Orlyandico , thanks for the comments: I was thinking of using a metal support and just a thin rubber band to provide the necessary friction. In this way the introduced error would be lower… I am tempted to give it a try!
Posted 24 February 2014 - 11:27 AM
Anything with rubber would deform if left stationary. So you'd get a flat spot on the roller that would cause a once per rotation bump.
Many mounts including the Paramount ME/MX and skywatcher Az Eq6 use pulleys with timing belts, it's a well known technology and cheap. See also the various belt drive mods for the Eq6.
Posted 24 February 2014 - 11:58 AM
Posted 24 February 2014 - 02:05 PM
I put a toothed belt/wheel drive on my old LXD-55 & Vixen GP from Warp.corp. They made an amazing improvement in tracking ability. My daughter is still running the LXD. My son is running the GP but upgraded to a Starbook-S.
Clear Dark Skies
Posted 25 February 2014 - 10:58 AM
A belt drive would also be an option: I will try and search for some options here in Europe (Germany). Any ideas about the number of teeth/diameter needed? I guess I would have also to "switch hemisphere", at least on my mount's controller, since with a belt the RA axis would now rotate in the opposite direction...
Posted 25 February 2014 - 01:26 PM
Here are some suggestions..
First, do the math. Based on the period and the number of worm wheel teeth, and the number of spur gear teeth, you sould be able to deduce where the error is coming from.
And notice the "Reversal" in the "Peaks" of the graph. For the first 3.5 minutes, the peaks are trending positive, and for the next 3.5 minutes, they are running negative.
My bet is that this corresonds to one complete turn of the worm wheel, and this would indicate that the worm gear itself may have an issue? I don't know. But it it takes one revolution in 6 or 7 minutes, there could be some small issue, but again, this is not that big a deal.
Anyway, what this would suggest is that if you have about the same number of peaks in the cycle as the number of teeth in the spur gear, then the spur gears could be your problem.
If the spur gears look like they are the problem then the next step would be to simply check the gear lash.
Depending on the shape of the gear teeth, you can get a deceleraton, a pause, and acceleration for each tooth engagement.
If the lash is to tight, the gears my bind somehere in the revolution, but if they are too loose, you can get this signature.
If the timing above matche the number of teeth in the spur gears then, look carefully at the mesh. This is adjustable on many mounts, but unfortunaly it is not "easy" to always get right.
But here is the proceedure I use.
Check the mesh. Using the worm input gear, try simply rotating the gear against the motor output gear to see if there is any play. If there is more than a tiny tiny amount, then you may want to proceed further.
If you have any meaningful play, then you may want to adjust it out.
First, find the adjustment that determines the mesh. I don't have a GP anymore, so I can't look at mine, but I think it was determined by moving the motor.
Before you do anything, find these adjustments.
Now, mark the gear positions on the worm shaft and use a hex key to loosen it.
Declutch the Motor Shaft.
If there was play obsevered before then the goal hear is to eliminate as much as possible but still keep from experiencing any binding of the gears.
Rotate a few complete revoltions and feel that pressure of a complete rotation does not vary.
Now, move the motor closer in a tiny increment.
In a perfect world though, as you eliminate more and more lash, you should feel the resistance growing evenly.
The goal is to feel the point where almost all of the play has been removed, but the gears still turn evenly.
Now, re-align your marks and re-position and tighen the worm gear on its shaft.
Now, turning the gear on the motor shaft (still unclutched) feel that the pressure is even.
If you feel it grow heavier, then ligher, you have a bent worm gear shaft.
This is either going to have to be straightend a bit, or you are going to have to go back and add a little mesh as necessary so that when you make one revolution of the motor shaft, you an even resistanace all the way around.
Of course now the problem will be that you are putting in some lash to correct for the bent worm gear shaft.
It might be possible to use the slow motion control rod on the worm shaft itself to try to straighten the worm shaft it you think it might be bent a tiny bit.
Sometimes you can even see this simply by putting the slowmo on the worm shaft and turning it and watching the end to see if it moves in a circle, but this is not the best way to do it.
Anway, the goal here is to get as much lash out as possible but to ensure that the complete revolution of the spur gears has the same amount of resistance.
Man it has been a decade since I have done this, so I am working off of memory, and I have owned a half dozen GP type mounts in that time, and all are a little different so I may have gotten some of the technical specifics incorrect, but now that you know what you are looking for and the general proceedure, you should be able to check it.
But the best place to start is to use the data you have to try to find if the number of teeth in the spur gear match up to the higher frequence peaks on the chart.
Make sense or did I do a horrible job of explaining it??
Posted 25 February 2014 - 06:39 PM
Your explanation is brilliant: the gears on my Vixen GP have 40 teeth, and one worm period should be 600 seconds, so if the spikes are given by the theet engaging I should expect one spike each 600/40=15 s. This is more or less what I see in my PHD graph. I will try to follow your instructions and check the gear lash as soon as I have some free time!
Have you tried any imaging with a Vixen GP mount? My fear is that even if I optimize the gear lash, I would be always limited by the mount mechanical limitations. I would like to push my mount to its limits, but I am not sure where they are!
Posted 25 February 2014 - 08:50 PM
Posted 26 February 2014 - 05:25 AM
Posted 26 February 2014 - 06:55 AM
Suggest replace the transfer gears with a 1:1 toothed belt pulley.
I looked at your other thread. The issue with the Vixen GP transfer gears is that they are too rough (have too few teeth) so there's a lot of lash in the tooth engagement.
If you adjust the motor so that the gears are tightly engaged, you'll get a huge periodic error because the gear holes are a little too large for the Vixen GP shaft (!) at least on mine.
A much finer-toothed gear would have much less inter-tooth spacing and thus less lash, and so would induce less periodic error. But the cheapest solution is to use timing belts.
The tooth count doesn't really matter; by its nature the timing belt engages the pulley over half of the pulley circumference (unlike a spur gear which is only engaging 1-2 teeth at a time). So there will be minimal backlash in the pulleys.
Also the timing belt itself is rubber with a steel core; due to the damping effect of the rubber, this would lessen lash even more.
Like I said the Paramount ME uses timing pulleys and a belt for the final transfer, and if it's good enough for a Paramount....
You can buy the pulleys and belts here - http://cnc-deutschla...anguage=en&a...
I believe the GP uses 6mm shafts on the worm and 5mm on the motor (you need to check) but the spur gears are 1/4" (6.35mm) which also causes the spur gears to wobble once every revolution.
Posted 26 February 2014 - 07:11 AM
The offset spur gear on the motor shaft will produce a large periodic error, but that error is at the worm shaft revolution period, which is 10 minutes.
What you're seeing (spikes every 15 seconds) are definitely due to the spur gears.
Posted 26 February 2014 - 09:44 AM
My fear is that even if I optimize the gear lash, I would be always limited by the mount mechanical limitations. I would like to push my mount to its limits, but I am not sure where they are!
Well, dude it is clear that you have waaaay to much lash here.
When the lash on a spur gear is to great, as the top land (the tip of a gear tooth is called the top land) move to the bottom land of the driving gear, the tooth actuallly stops transferring power for a brief inverval of time as the face of the driving gear transits the gap beteeh the driven gear faces. For that period of time, the drive actually stops transmitting power to the worm. When the top land transits the gap and the faces come into contact again, it goes from no motion and accelerates to the drive speed.
And you see this clearly in your chart, yes?
Now, remember, if you find that the gears are tight in part of a full rotation, this indicates that one of the shafts is bent, and my bet is that it is the worm shaft.
If when you loosen the motor gear and turn it by hand against the worm shaft gear, this will tell you the worm shaft is bent or not.
If it is bent, you have to decide whther you want to try to straighten it. This can usually be done by using the slow motion knob on attached to the worm shaft and very slightly applying pressure to slightly bend the worm.
The goal is to get a full rotation to have as even a pressure as possible.
I image. A lot of people don't know this because well I don't talk about it because I am not very good at it. Mostly, I don't know how to do all of the processing and just done care to learn.
But I have not used the Vixen to image.
Here is what I know though. Most of these systems have the potential to do much better than they do out of the box.
The raw potential though may take considerable patience to fully exploit.
But hey, I see numerous accounts of people using far more expensive mounts than yours that still wind up fussing with PE errors.
Mostly what is required to get a dramtic improvement is patience and a lot of small, sometimes tedious adjustments.
Lash is a big factor, and you can see that your lash is far to great.
But then you have the worm itself, the worm gear mesh to the worm wheel, and the end float of the worm gear.
The real problem is not that the mounts don't have the potential, but that the adjustment mechanisms are crude.
These tend to use corse threads and pinch mounts, which make it difficult to get the adjustment made with any precision. Just tightening the worm gear assembly can cause it to shift, ruining the mesh.
This is just my opinion of course, but the real potential is hard to fully exploit simply because the design itself does not allow for easy precision adjustment. You set something, but when you tighten things down, things move.
But with patience and a keen eye to detail this class of mount should be capable of excellent performance.
Now, some things might require more effort. Lapping the worm gears could improve things, but now you are talking about a much greater level of commitment.
Better to see if you can inspect the worm with a loupe and look for small flats or spurs on the worm gear teeth and use a jewelers file.
You will see this in your PE as either a single spike in one revolution of the worm, or in more than one spike that happens with the same offset from the beginning of the cycle.
For example, if you have one 2 mintes into the cycle, and another one 3.5 mintues into the cycle, always occuring at those intervals, I would look for a small spur on the top land of the worm gear, and if I saw it, I would dress it with great care using a precision file.
I am not saying that all of this is worth the effort, but often, mass produced mechanical things like this are assembled with very sloppy tolerances and many small imperfections in machinging are allowed to slip by.
My guess though is that you are well below the full design potential of the mount, and only you can decide if it is worth digging into.
But once again, you can spend a lot more money on a bigger "Better" mount only to discover the exact same thing.. These things are not assembled and inspected by NASA engineers.
Posted 26 February 2014 - 11:30 AM
The Warp belt drive replaces the brass gears with a pair of Nylon toothed gears & a toothed belt. The adjustment between them is still the same mechanism. I added a tension spring made from a wooden spring closing clothespin & a brass bearing made from a brass ink pen refill. I screwed this so that the brass bearing put slight pressure on the outside of the belt to maintain a consistent pressure in the drive to remove an slop in the belt section of the drive. To remove the gear slop, I made shims out of simple aluminum foil epoxied between the drive shaft & the drive hole. This removed all of the slop/slack/flex from the entire drive. No more backlash.
When this was done I still had a PEC of 10-15". I found that the worm had problems. I took the entire worm mechanism to a machinist friend & he found that the worm itself was not concentric & the threads weren't consistent in width or depth. He remade another brass set for me. A case of Bass Ale & weekend of deer hunting went a very long way. The new worms brought the PEC down to 2-3". We looked into that & found that the Wheel itself was also a source of error. This was actually easier to correct. The wheel wasn't centered in the RA axis. I marked the position of the wheel, moved it 90* marked another 90*. This let me know exactly how much to "Nudge" the wheel to center it. PEC came down to 1-2" as measured with a Imaging Source v.03 camera. I did an enormous amount of work on my GP & LXD-55 to make them both perform well enough for AP because the problems they had were really poor quality production. Unfortunately neither one & even the G11 that replaced them, had the capacity I wanted & got with an HGM-200. They weren't abandoned, junked, or sold. I wouldn't gotten enough for them.
My son is running the GP with a Starbook-S an AT6RC, & Orion Starshoot, doing solar system AP.
My daughter is using the LXD, Autostar & a Warp kit using a Meade 100mm f9-10(?) & is happy just looking through it.
I'm now running the Losmandy HGM-200 (Titans older brother) Gemini-2, Vixen 8"f4 R200ss, Celestron C6R with a 90mmf3 piggybacked guider & a Coronado PST all side-by-side on a tripod in a Skyshed POD. Quite a setup.
The short story is that I found that the Warp kit was well worth waiting to get. He only makes them in batches 1-2 times a year when the clamor gets loud enough. I think they're now ~$100 USD but I can't swear at that. Dead simple to install, no more greasing the gears, checking gear mesh, etc. HTH
Clear Dark Skies
Posted 26 February 2014 - 05:19 PM
Posted 26 February 2014 - 05:48 PM
The worm seems ok: by de-clutching the motor and manually turning the shaft, I could notice only a minor change in the resistance. By a closer inspection, it turned out that the gear which is pressed on the worm shaft when the clutch is tightened (red arrow in the picture, image taken from Company7 web page) was responsible for the uneven resistance.
I carefully adjusted the position of both the RA and the Dec. motors: now the lash in RA is absent, while a tiny bit remains in Dec. I can't observe any gear binding, but if there should be a bit of play, then maybe the RA gear is too much pressed against the worm driving gear. I have to wait for some clear sky to see wether the tracking improved!
I enjoy tweaking this mount: even if, as you said, the regulations are pretty coarse and it might be tedious to find the right settings, the satisfaction I get if things will prove to work out is a great motivation!
Thanks again for your advices: hopefully I'll post some positive updates soon!
Posted 26 February 2014 - 05:55 PM