Best timing belt and pulleys type for minimum backlash

I was at the only specialised store in my area and held in my hands the T5 and HTD 5M pulleys and belts. The T5 have a huge tooth clearance. I estimate the backlash to maybe 0.5mm. The HTD are much better fitted but there is still a visible clearance/ movement when pulled by hand. The pulleys are made by BEA, the HTD belt is Optibelt and the T5 belt I do not remember.

I made some research and found that the HTD were designed for high torque capability but hawe the greatest amount of backlash and the Trapezoidal are suited for positioning accuracy as they have minimum clearance/ backlash. Strangely this is exactly the opposite to what I has in my hands.

Here I found something interesting on page 19 about different tooth gap shapes for the same tooth profile (normal, reduced and zero backlash tooth gap) but I cannot figure out if the belt or the pulley teeth are different? From the drawing it seems that the pulleys are different.

Now there is the GT2 type, a modified HTD shape providing high torque as well as minimum backlash comparable to the Trapezoidal and the AT type which is an improved classical Trapezoidal profile for better performance and reduced or zero backlash. I even found a new type of "Non-Backlash Timing Pulleys" S8M.

As my research is mostly based on the manufacturers advertising which do not necessarily reflect reality, it will be interesting to see your practical experience with different types of belts and pulleys.


some more docs:
http://www.gates.com/facts/documents/Gf000282.pdf
http://www.sdp-si.com/D265/PDF/D265T003.pdf
http://www.sdp-si.com/D265/PDF/D265T042.pdf
http://www.gatesmectrol.com/mectrol/...older=brochure

I have a belt driven machine. Originally with XL belt, then i upgraded to AT. I wouldn't be too concerned with backslash, as there is no any when tightened properly.

Belt stretch under heavy loads is another thing. However an 0.05mm precision can be achieved doing the following. I made a Mach3 file for every material i cut by cutting and measuring square with a circle in it one pass full depth of the bit and using the Mach3 motor steps adjustment features. So i have a file for wood, plastic and aluminum.
The practical downside of this is if you have bigger machine, you have to make at least 2 files for various sizes similar to your normal work. The file for a 1m piece is different from the file for 10cm cut pieces.

Bear in mind my case is extreme as i use only 10mm belts. I imagine with wider belts it wouldnt be so much of a stretch. Another thing i learned is to use steel braided belts, not the kevlar belts.

AT belt is definitely the strongest from the normal belts, i mean apart from the exotic " special" ones. A bit stronger than the HTD and quite much compared to the XL.

Preloading the belt using a tuner and frequency calculator from their site is the proper way to eliminate belt backslash, i mean if you don't have a very handy device that measures the stretch.

Of course that is valid if the manufacturers tooth engagement and belt design rules are followed.

Quote:

---

Originally Posted by paulus.v

The T5 have a huge tooth clearance. I estimate the backlash to maybe 0.5mm.

---

Blimey he is right, I just tried a piece of T5 belt in a pulley and it was total crap.

Why would they make it like that?

There are some figures for belt stiffness here:
http://www.gates.com/common/download...talog_5_11.pdf

On the micro lathe I made with preloaded nuts (here), and more recently the CNC router with the same system (here), it was possible to obtain zero backlash (in steady state at least) by tensioning the belt properly. The tension on the belt was a lot higher than I expected would be required. I expect there's probably still extra hysteresis in the system caused by the belts.

You could make your own pulleys with lower clearance. I suspect the belts would wear out faster, but that's probably not an issue on a CNC router where they're only lightly loaded.

This is concerning. I'm working on my first CNC conversion of a milling machine and opted for belt drives in order to make the build more compact. And now, halfway through the build, I'm reading that using T5 belt drives will significantly increase the amount of backlash - something that I am trying my best to eliminate!

I'm looking into belts at the moment.
I have just got a sample AT5 pulley & belt from Beltingonline.com, Supplier of power transmission and conveying products

It seems ok, but a lot of people here seem to use HTD pulleys and belts so i might order a sample one of these to have a look at too,

Photo of AT5 belt on 44 tooth pulley
Attachment 9932

Quote:

---

Originally Posted by birchy

This is concerning. I'm working on my first CNC conversion of a milling machine and opted for belt drives in order to make the build more compact. And now, halfway through the build, I'm reading that using T5 belt drives will significantly increase the amount of backlash - something that I am trying my best to eliminate!

---

Sometimes reading the manufacturers general guide lines on the first page is better than reading 100 posts on forums. :toot: Its all there, the good and the bad in bold letters:
Attachment 9931

Are you too far to go back now? Belt and pulleys are so cheap. I would not be worried if i were you. As Jonathan confirmed, backslash is eliminated if tightened properly. However i would be worried if some of the pulleys you intend to use are the smallest for the belt size. Up to 18-20 tooth i mean. Implementing smaller than 18 tooth pulley is a call for problems with any belt.

Generally that's why people use so much HDT timing belt. Stronger than XL, better than the T for linear motion and requiring smaller pulleys than the stronger AT.

As far as i know Belting Online is the cheapest place, i buy from them so its worth a look comparing pulley sizes and so.

Damn, I wish I'd seen that chart before! I only chose T5 belts because someone on here recommended them when I started the build. I know you say pulleys are "cheap" but I've only recently bought two T5 belts from beltingonline at a cost of £9 each and 4 pulleys at just over £10 each with boring and grub screws. So that's almost £60 of "cheap" parts for the scrap bin. In hindsight, I can see that HTD5 pulleys and belts would of been a much better choice for cost, availability AND backlash reasons. Oh well, we live and learn.
:disturbed:

EDIT: Just a FYI, http://www.bearingstation.co.uk/ are generally a bit cheaper and have a "price match guarantee", however they don't offer a machining service.

Quote:

---

Originally Posted by birchy

Damn, I wish I'd seen that chart before! I only chose T5 belts because someone on here recommended them when I started the build. I know you say pulleys are "cheap" but I've only recently bought two T5 belts from beltingonline at a cost of £9 each and 4 pulleys at just over £10 each with boring and grub screws. So that's almost £60 of "cheap" parts for the scrap bin. In hindsight, I can see that HTD5 pulleys and belts would of been a much better choice for cost, availability AND backlash reasons. Oh well, we live and learn.
:disturbed:

EDIT: Just a FYI, Bearing Station - Adhesives, Bearings, Belts, Chains, O Rings, Oil Seals, Pulleys & Sprockets are generally a bit cheaper and have a "price match guarantee", however they don't offer a machining service.

---

Do not throw anything in the bin! And sorry for that "cheap" . I meant relatively compared to other linear movement.

If somebody told you is maybe because it works. In practice there is a belt backslash only if you have open belt clamped from both sides and the linear movement is transmitted by only one pulley

In a closed system where 2 pulleys are engaged, when you tension the belt enough / but not too much!/ the teeth on the opposing belt touch tightly the teeth of the opposing pulleys and preload the system, the way 2 ball screws are preloaded.

So whatever specification say, i believe you did nothing wrong and the belt will work without backslash.
And talking about backslash and various types of belts manufacturers refer more to the tooth strength and theoretical situations with one pulley only.

So cheers up!

OK, OK! I'll put the lump hammer down for now! :congratulatory:

I have got one 14T pulley to give me a 3:1 ratio (42T) on the Z axis, so might have to change that to an 18T or bigger. My thinking is that even if we had 1 or 2 degrees of rotational backlash in the pulleys, this would still not be enough to cause a problem if using ballscrews?

Quote:

---

Originally Posted by birchy

OK, OK! I'll put the lump hammer down for now! :congratulatory:

I have got one 14T pulley to give me a 3:1 ratio (42T) on the Z axis, so might have to change that to an 18T or bigger. My thinking is that even if we had 1 or 2 degrees of rotational backlash in the pulleys, this would still not be enough to cause a problem if using ballscrews?

---

Here is a simplified, why backslash in reality will not happen with these belts
Imagine 2 pulleys on a horizontal plane with a belt mounted. When tightened the belt, all the teeth from 6 to 12 o'clock on the left and from 12 to 6 on the right will preload.

Now imagine the belt start to move clockwise driven by the left pulley. From 9 to 12 o'clock is where the backslash will happen, but from 6 to 9 will be still preloaded. At the right pulley the same. From 12 to 3 backslash and from 3 to 6, tooth engaged.

So at any time at least half of the engaged teeth are without backslash.

Hence the tooth strength together with the tension plus the pulley size , if enough will not permit backslash.

18t pulley is what i use on my machine, changed from 14, he he. For other reasons though.

Yeah, for the Z it shouldn't be a problem.

When i changed belts even when the pulleys were tightly fitted they gave me problems, especially the small ones connected directly to the motor shafts. They were backslashing by slipping on the shafts even if i used 3 grub screws on each. And only under heavy loads so it took me a lot of time to figure the problem. A little "super" glue eliminated this.

Quote:

---

Originally Posted by birchy

EDIT: Just a FYI, Bearing Station - Adhesives, Bearings, Belts, Chains, O Rings, Oil Seals, Pulleys & Sprockets are generally a bit cheaper and have a "price match guarantee", however they don't offer a machining service.

---

I have used them before a few times and to be fair they have been ok. BUT this last time they let me down badly with regards to delivery times in that I ordered 8 20t pulleys on the 5th August and despite 10 -12 phone calls which they promise to return with an answer, they never do. I have now got 7 of the 8 and they are telling me that none of their supplies have got any more and its not their fault. I now have rung Belting-on-line and they have plenty.
I personally will not be using them again. Clive

I have servos with 19mm shaft and i want to retrofit a lathe (1.6m between centers) .
I want a pair of timing pulleys with 1:3 reduction for Z axis and a pair of timing pulleys with 2:3 reduction for X axis.
What kind timing pulleys/belts will be best for that application that they have NO backlash and good quality/performance ?

Also all car engines have inside timing pulleys ( What Is a Timing Belt ? - AutoZone Car Care - YouTube )
So you can extract for free timing pulleys from old broken car engines if you go at car repair shop and search at scrap engines.
I have extracted some timing pulleys from broken engines but i don't know the type of pulleys to order matched timing belts.
Where can i find a chart with specs with all types of pulleys to identify that i have extracted from car engines ?

Oof! And get ready for it... the deep dive on belts and pulleys, everything you every wanted to know and more: http://file.lasersaur.com/docs-third...ming_Belts.pdf

I'm very lucky in that my local bearing and seal supplier also do toothed pulleys and belts, it was there that I saw the steel wire reinforced belts, they have them in stock in stock too.