Small but mighty tuning for the EQ6-R. Change the pulley! Sky-Watcher EQ6-R · Pinguru · ... · 44 · 4026 · 35

Added this pulley to my EQ6RS project need list.

Hi Rostokko,
You could get yourself a simple puller to help with this job.

Example:
MANNIFEN 2 Jaw Puller, 2 Claw Gears Puller for Removal of Pulley Flywheel Bearing,Small Gear Puller (4inches) https://a.co/d/hyvQhfX

I can't check the correct dimensions for the puller you need, but you should easily find a matching one after taking off the small belt cover. 
Regards, Volker

Check for "thin jaw puller" 
You don't need a lot of force. As mentioned in my opening post, add a drop of thin oil to help loosening the pulley. Something like WD40.
And keep it square while working it off the shaft.
Regards, Volker

Just as a general comment: I am surprised that belts and belt tension in EQ6-r pro's are not more prominently discussed when talking about backlash.

My mount worked perfectly well for about 9 months. I did use it extensively in that timeframe, and it did spend time outside (protected from the weather) in cold and hot days. Then I started seeing a major degradation in DEC guiding - and it was sometimes hard to tell if it was due to backlash, stiction, damaged worm or something else.
I started researching the topic, and the vast majority (almost all) of information I found was about how to tune the worm position, or its end float.
So, I went through that path - in a few iterations, and I did fine tune those two aspects as well as I could; but the problem persisted.

Finally, I took a closer look at the belts - which I had previously inspected, but not as careful; this thread actually pushed me to take a closer look.
Slewing at low rate (2), it was *very* easy to see where the problem was - similarly to what @Pinguru posted here originally: after slewing in one direction, it would take 3-4 taps to start movement of the bigger pulley in the opposite direction. While the belt "felt" OK tensioned, clearly it wasn't. In fact, comparing it to the RA belt, the difference in tension I could feel with my fingers was fairly small, but the RA's belt "induced backlash" was about just 1-2 taps.
Once I figured that out, the fix was easy: I increased the belt tensions (I also tuned the RA's one a bit), and now the belt induced backlash is down to < 1 tap. I can't really go lower than that without applying what I would consider excessive tension. But now I am back at guiding consistently at <.50" the whole night, and typically <.25" in DEC.

So, a few things are now clear to me:

1. Next time I have suspicions of increased backlash, the first thing I will check are the belts. Being flexible components, they will inevitable stretch and deteriorate faster than other components; and checking the backlash they introduce is quite easy and deterministic
2. The choices made in the original design, as already mentioned by @Pinguru , are, to say the least, sub-optimal... The amount of "intrinsic backlash" that the original combination of pulleys and belts introduces in the system is appalling to me - especially when I compare it to other discussions I read about rebuilds and fine tuning tolerances...
3. Next time I need to go through a belt tuning or replacement, I will most likely pull the trigger on replacing the pulleys and belt design as suggested by @Pinguru ; it just makes sense to limit as much as possible the intrinsic belt drive backlash

Hopefully this thread will help many other EQ6-r Pro users!

I've arrived at this thread twice now. Once googling for belt backlash on this mount, and then separately and for unrelated reasons: trying to diagnose the very clear spike at 10s on my guided FFT plots using the PHD2 log viewer:

Looks like that 10s periodic error is a pretty repeatable flaw with the stock belts! I have lots of experience with GT2 belts from self-built 3D printers, and love the idea of replacing this with a GT2. Unfortunately the 15t drive pulley is a less common size and I couldn't find anywhere in North America to purchase some with a 5mm bore, so I'm waiting on shipping from China. I'm really looking forward to making the swap and will report back here after I do! Which based on current shipping estimates probably won't be until about mid-February.

Some side notes for those looking to source their own pulleys:

I made the following baseline center-center measurements:
My stock DEC center-center distance was about 48.8 mm
My stock RA center-center distance was about 49.0 mm
Based on the above, I definitely agree that if going for a 15:60 reduction, 89 teeth (178mm) is the best size for the belt.

Don't get fooled by the widely available 16 teeth pulley - I haven't found a matching 64 teeth "big pulley" (and it would be an even tighter fit in the chassis), so the very commonly available 16T GT2 pulley will not work.

Also, a 12:48 setup would be feasible (both sizes are manufactured), but I also wasn't able to find 12 tooth pulleys available from local warehouses. So you lose out on the additional tooth meshing as mentioned in the OP, and you still are dealing with international shipping. So I also agree 15:60 is the way to go.

Tony Gondola:
I'm getting 0.3-0.2 pixel RMS totally stock

At what pixel scale (arcsecs / pixel) for the guide camera?

Great news! The GT2 pulley and belts showed a marked improvement over the stock T2.5 setup, just like in the OP.

As a reminder, this was my original guided frequency domain plot:


Just like @Pinguru , I had a sharp peak at 10 seconds. I bit the bullet and ordered correctly sized GT2 parts from varying overseas sellers. They came in a couple days ago, and I finally had some time tonight so I installed the RA parts in place in my backyard (perhaps not the smartest decision). And the results speak for themselves:


That 10s peak is completely gone. And all of the high frequency errors are within a pretty consistent magnitude. Note: overall my guiding is slightly worse, but I believe I had the PPEC profile I programmed into the mount active for the first data acquisition (the 10s periodic error was on every single one of my guide logs, PPEC or not. It is definitely a hardware problem)

Some installation thoughts:

• I too had glue on the OEM set screws, and even worse, on the pulleys themselves. Getting the OEM pulleys removed was a major pain. I wish they hadn't done that, because everything else is surprisingly serviceable and accessible (I like the little access door for the big pulley grub screws, for example)
• I also used M4 screws to pry off the worm gear pulley. What I didn't notice was that the outer edge of the chassis - the part that is closer to the pulley - is actually just barely not wide enough to pry against. I accidentally screwed into the edge of that surface, not realizing that I was screwing into something rather than prying.
• If anyone else attempts this, take a close look at the pictures in the original post to plan out where you will pry. And get the longest M4 screws you can, because the deeper part with a good prying surface is quite deep. My M4x20 screws were not really long enough.


• The shoulder on my 60t pulley was quite wide, and to get the belt centered(ish), I had to place the small pulley quite far out on the stepper shaft. I'm not a huge fan of that - there's a decent bending stress on the stepper shaft now. But, steppers are cheap and easy to acquire, and I've seen them take a lot of axial torque abuse.
• That said, if anyone else is interested in doing this, I would recommend trying to get 60t pulleys with the narrowest shoulder that you can find.


• I noticed there was quite a lot of range on the belt tension adjustment. It might be possible to get away with an 88 tooth belt, which is easier to find. I definitely recommend the 89 tooth 178mm belt if you can find it, but a 176mm belt might just be within the adjustment tolerance. Definitely proceed at your own risk though - there's a chance it might be too difficult to get on even with the most slack.

Here are some installation pics, although probably not too helpful. The pulleys and belt are after the GT2 parts installation and I believe after I tensioned the belt. And the chassis picture with the electronics plate loose isn't really for any particular reason, but there aren't a whole lot of pics of the inside of this thing floating around, so I figured I'd include it too while I'm here.


Thank you so much for figuring this out @Pinguru and for documenting your methods so well!

Finally some decent seeing for properly testing the tuning; not bad!

Matt Link:
@Martin W@Rostokko Would you guys be able to provide where you found the 60t GT2 pulleys? (that will ship to US).

I got a “ReliaBot” pulley for the 60t one: https://a.co/d/4BO9Y9C The style on the Amazon listing is “8mm bore, for 10mm belt”. Like I said in my post, the hub is definitely pretty wide on it. It fits inside the case and it works, but it makes the smaller pulley have to be pretty far out on the motor shaft. So you might not like it, but I can at least confirm that mine fit inside the case.

Hi all,
I really love to see that my mod is used all over the world :-)

Just to give you an update on the longevity of the GT2 belts:
Yesterday I have taken my mount apart to check the belt tension and the play of the worm drive. Guiding performance was still excellent.
I have taken the two pictures without any cleaning, this is how the belts look after 1.5 years of extensive use. Both RA and DEC look similar. They are still in excellent shape with no signs of wear.
I have seen other standard belts from EQ6Rs that looked very beaten up.