After Tim's (toc) description of his DEC wobble, I thought I had better have a look at mine, and RA at the same time. Here is what I found, and photos are attached:
I removed the covers of both DEC and RA belt drives, and I will say I was surprised by the belt alignment as delivered. I had not removed the covers before. The mount I had before this CGX was a Rowan Belt Mod NEQ6 Pro, and it was a well engineered bit of gear, which I installed, so I know something about belt systems and alignment.
First I did a slew to M42 and watched the belts in action, they were doing their job and did not exhibit any bulging or wobbling action that Tim described. However, it did seem like they were out of sprocket alignment and on RA hanging over the end of the sprocket. The motor drive sprocket does not have an edge flange so it is free to move across the sprocket seeking to centre itself. The Worm sprocket does have an edge flange. What I noticed about the motor sprocket is that if was positioned fairly far out on the axle, like they were trying to get it far enough out but the motor shaft is too short, it is just on far enough to engage the lock screw on the shaft but it looks to me that it is too short for that arrangement. I can see how the motor sprocket could develop a wobble as the belt will move outward and put more force on the outside of the sprocket which may cause the sprocket to move against the flat of the shaft. The worm sprocket cannot be moved inboard any further due to the casting so the motor sprocket has to be out as far as possible to make it work, and the narrow sprocket cannot be made larger because of the confines of the plastic cover and the risk of rubbing on it. Both DEC and RA belt drive designs were similiar, and frankly a disappointment, this is bad design and they have tried to fix it by positioning of the sprockets.. I expected better, certainly as good as a Rowan Belt mod. Perhaps Rowan can come up with something to make this work better, a thinner belt and worm sprocket might allow a straight alignment. The constraint to a solution is going to be the plastic covers.
Now some will say "it doesn't matter what it does when its slewing, it's the tracking or guiding performance that counts", and this is true; but it bugs the heck out of me none the less.
Photos attached of DEC and RA.
I wonder what they have done with the CGX-L, there is potentially more room there as bigger rings are used, and the drive areas look larger in general.
Well I don't know enough about engineering to say Glend.. I'm more of a computer guy than a gear and cogs one
I can say it's good to be back online in the obs tonight and I'm getting great results again fully loaded with the 11" RASA, guide scope, finders scope, camera and filters.. about 19-20 kgs. CGX is rated to 25kgs.
2 x ASPA routine with the hand controller after 2 star align + 4 cal stars.
Guiding graph attached. Also how amazing are these 60 second Ha subs with the RASA (!!)
Thanks Dylan - out of interest, did Bintel replace or repair? How long did it take? I guess my issue is that the mount is not really malfunctioning, its just not performing as I would like/hope.
The head was sent back to Celestron to investigate and after 3 weeks I received a new head, at least I think it is.
They changed CEO and have had a lot of internal changes recently which I believe added to the delay but I think they're back on track now.
There was no need to ship the tripod / weights etc of course. Work with Bintel to determine if the mount is performing as it should or if it is indeed faulty. There are an embarrassing few posts on CN where a guy swore blind a returned Celestron mount wasn't performing but it turned out it was perfect so it must've been him.
But these things do happen for us early adopters so after you nail polar alignment and star alignment 101% you should be able to demonstrate whether its performing to spec or not. I've used the CGX for fairly long unguided subs too which might be a good start. That way your guiding setup is out of the equation.
Looks a bit skew-if doesn't it? Would that have an effect on backlash though? In theory its the gaps between the "teeth" that would increase backlash and having it skewed would mean a tooth edge is likely to close the gap quicker wouldn't it? I don't know anything about anything though.
I can say I loaded the CGX up to about 24kgs and it didn't handle it. It physically would not slew back against gravity, which is not ideal. They should probably pull back it's rated payload.
Definitely looks wobbly - the gear looks wobbly too
If the shaft was straight and the only issue is a bad cog it sure would have been a lot easier for them to send a new part, even if they sent it to Bintel.
It does not look that much different to mine, the main issue i see is the belt is hanging over the side of the motor sprocket and it appears to walk back and forth slightly, perhaps due to the bad fit on the shaft of the sprocket; or the belt is being guided by the worm sprocket edge flange. An over hanging belt puts alot of force on the outside of the sprocket. I believe both RA and DEC motor sprockets need to be replaced by a longer sprocket in these mounts. Walking belts, or belts hanging off of drive sprockets, is not good design. I am comparing it to my old NEQ6 with the Rowan Belt Mod, where both sprockets had edge flanges and lined up perfectly. I am wondering if Rowan Engineering could produce a replacement motor sprocket set for these mounts. Ideally Celestron should fix this on all CGXs.
Now the big question is, does visual wobble in high speed slew have any affect on very low speed guide or tracking movements? Perhaps not. However, it could be affecting worm gear end play at various places in the rotation of the worm as the belt tooth might catch on the worm sprocket edge flange when running at that angle.
I hope Celestron is on this, and not just focused on the CGX-L launch. Tim did you file a report with them?
Glad to hear you guys are shakin' the mounts down for the rest of us
Hope Celestron get on the case with this, the lack of belt stop on the top cog looks a bit of an obvious oversight, but I'm far from a mechanical expert!
I'd be keen to get one of these once the teething issues are over.
After Tim's (toc) description of his DEC wobble, I thought I had better have a look at mine, and RA at the same time. Here is what I found, and photos are attached:
I removed the covers of both DEC and RA belt drives, and I will say I was surprised by the belt alignment as delivered. I had not removed the covers before. The mount I had before this CGX was a Rowan Belt Mod NEQ6 Pro, and it was a well engineered bit of gear, which I installed, so I know something about belt systems and alignment.
First I did a slew to M42 and watched the belts in action, they were doing their job and did not exhibit any bulging or wobbling action that Tim described. However, it did seem like they were out of sprocket alignment and on RA hanging over the end of the sprocket. The motor drive sprocket does not have an edge flange so it is free to move across the sprocket seeking to centre itself. The Worm sprocket does have an edge flange. What I noticed about the motor sprocket is that if was positioned fairly far out on the axle, like they were trying to get it far enough out but the motor shaft is too short, it is just on far enough to engage the lock screw on the shaft but it looks to me that it is too short for that arrangement. I can see how the motor sprocket could develop a wobble as the belt will move outward and put more force on the outside of the sprocket which may cause the sprocket to move against the flat of the shaft. The worm sprocket cannot be moved inboard any further due to the casting so the motor sprocket has to be out as far as possible to make it work, and the narrow sprocket cannot be made larger because of the confines of the plastic cover and the risk of rubbing on it. Both DEC and RA belt drive designs were similiar, and frankly a disappointment, this is bad design and they have tried to fix it by positioning of the sprockets.. I expected better, certainly as good as a Rowan Belt mod. Perhaps Rowan can come up with something to make this work better, a thinner belt and worm sprocket might allow a straight alignment. The constraint to a solution is going to be the plastic covers.
Now some will say "it doesn't matter what it does when its slewing, it's the tracking or guiding performance that counts", and this is true; but it bugs the heck out of me none the less.
Photos attached of DEC and RA.
I wonder what they have done with the CGX-L, there is potentially more room there as bigger rings are used, and the drive areas look larger in general.
agree yours and Tim's look significantly out of whack - that is definitely not how a precision belt drive should look. Hopefully these are teething problems and they will be fixed by the company - the basic design concept looks exciting.
Thanks Ray i had a look at that video. The L is using the same belt as the CGX, a T5x165 P07, designated with the TLnis branding. They cut away the video of the belt drive before the mount rotated around enough to see if the belt was skewed on the sprockets, but the inside gap is the same so i think it was probably hanging over the drive sprocket as well. Surely they have to be aware of the problem. The sprockets look to be the same as well. The belt is only about 10mm wide. The motor shaft dia is 5mm. I am trying to track down a wider motor drive sprocket with the same teeth and diametre, but would surely void the warranty if i pulled that assembly apart to change the sprocket. I doubt Bintel would want to tackle disassembling the drives, so what is left, a return to Celestron?
Dylan can you have a look at your 'new' mount and see if they fixed the misaligned belt?
Thanks Ray i had a look at that video. The L is using the same belt as the CGX, a T5x165 P07, designated with the TLnis branding. They cut away the video of the belt drive before the mount rotated around enough to see if the belt was skewed on the sprockets, but the inside gap is the same so i think it was probably hanging over the drive sprocket as well. Surely they have to be aware of the problem. The sprockets look to be the same as well. The belt is only about 10mm wide. The motor shaft dia is 5mm. I am trying to track down a wider motor drive sprocket with the same teeth and diametre, but would surely void the warranty if i pulled that assembly apart to change the sprocket. I doubt Bintel would want to tackle disassembling the drives, so what is left, a return to Celestron?
Dylan can you have a look at your 'new' mount and see if they fixed the misaligned belt?
Perhaps we could engage the services of a talented iceinspacer with a CNC machine or lathe, and get some new ones made up
I have opened a ticket with Celestron Tech Support re the belt alignment issue. Tim i assume you have a ticket open re the DEC wobble? I have also inserted my photos in the CN thread i started on CGX Belt Misalignment. To quote Mal, from Serenity, "I aim to misbehave". I want some answers.
Tim i assume you have a ticket open re the DEC wobble? I have also inserted my photos in the CN thread i started on CGX Belt Misalignment. To quote Mal, from Serenity, "I aim to misbehave". I want some answers.
I do have a ticket open - they asked for a link to my video of the wobble, and I sent that off a few days ago. No response so far.
I also have a thread in the Team Celestron site, but that is more around the backlash issue.
I imagine we will see an silent update to the design in a year or so
Thanks Ray i had a look at that video. The L is using the same belt as the CGX, a T5x165 P07, designated with the TLnis branding. They cut away the video of the belt drive before the mount rotated around enough to see if the belt was skewed on the sprockets, but the inside gap is the same so i think it was probably hanging over the drive sprocket as well. Surely they have to be aware of the problem. The sprockets look to be the same as well. The belt is only about 10mm wide. The motor shaft dia is 5mm. I am trying to track down a wider motor drive sprocket with the same teeth and diametre, but would surely void the warranty if i pulled that assembly apart to change the sprocket. I doubt Bintel would want to tackle disassembling the drives, so what is left, a return to Celestron?
Dylan can you have a look at your 'new' mount and see if they fixed the misaligned belt?
just a thought Glen. The belt will run off if the two shafts are not parallel. The worm is held both ends, so it is OK, but it looks like the only things holding the motor in place are the two adjusting screws pulling the motor end plate up against the worm block. If this is the case, any tension on the belt may be sufficient to flex the motor plate and pull the whole motor assembly out of alignment. Maybe back off the belt tension and see if you can tighten the belt by putting finger pressure on the far end of the motor. If you can, it could possibly be worth shimming the front of the motor where it attaches to the end plate or using a zip tie? to stop the far end of the motor from lifting when the belt is tensioned - obviously keeping away from anything that looks delicate.
It sort of looks like maybe they used their spring loaded worm block design ideas from the CGEpro and put it in this mount - however, the original was designed for gears, not a belt under tension???
Ray, I need to have a close up look at the assembly to see how the motor and worm shafts are held in alignment. They are both attached to a cast frame, so hopefully there are some adjustment screws and it maybe a simple process to line them up - and then maybe the existing spocket arrangment would be ok ( although i still think it needs and end flange).
This sprocket, with the right pitch dia etc might be the ideal way to offset the belt slighly, and not have to mess with the alignment if the shafts.
I have just had a closer look at both belt drives and have some new photos to share (first 4 photos are RA and the last 3 are DEC). I have marked the DEC photos with red arrows to indicate the adjustment points. It seems the drive casting, while supporting the Worm at both ends by bearings as expected, does not support the motor at both ends. The motor is only supported at the sprocket (transmission?) end by a separate casting. It appears that when they tension the belt they push up the motor casting by set screws, (via the small wedge adjusted to push against the bottom casting) and when the lock screws are tightened motor is tilted (and thus the sprocket as well), in relation to the worm sprocket, and this leads to the belt walking over to the end of the motor sprocket. I tried an experiment, pushing the belt into alignment on the motor sprocket (on both DEC and RA) and then running a slew to a target. By the time the belt had rotated a few times it was hanging off the end of the sprocket again. A flanged sprocket might stop it but at what cost, as the motor assembly is still skewed.
Perhaps simply putting more turns on the wedge set screw (before locking it down) might actually level it up. Opinions?
Interesting arrangement of sprockets.
So the worm sprocket has flanges but the motor sprocket does not?
Have you measured the distance between the middle of the motor sprocket and end of the motor spindle where it attaches to the sprocket and compare that measurement against the distance on the worm spindle? (like the Rowan belt alignment routine)
If there is a bit of room for adjustment in the alignment, is it possible to shim a couple of thin but slightly oversized washers to the motor sprocket to act as a flange?
Just a thought.
Bo
