MechMate CNC Router Forum

MechMate CNC Router Forum (
-   Driving Mechanisms: Rack/pinion, gears, screws, belts & chains (
-   -   Reduction Drive by belt - a collection of various designs and how to build your own. (

Mike Richards Thu 24 August 2006 10:53

Reduction Drive by belt - a collection of various designs and how to build your own.
This is Version 4 of my 3:1 belt-driven gearbox. It takes me forever to design something that I like enough to keep.

This version is cut from 1/2-inch thick polycarbonate. The spacers are 1-inch aluminum. (The extra-long shaft and rusty gear are left-overs from other projects.)

Starting at the shaft, there is a 1/2-inch clamp collar, #9 rubber O-ring, 1/2-inch shielded ball bearing, motor plate, 60-tooth XL gear, main plate, 1/2-inch shielded ball bearing, #9 rubber O-ring, 1/2-inch clamp collar, 20-tooth, 20-pitch, 20-degree spur gear. The aluminum spacers are 1-1/2-inch high, threaded for 1/4-20 machine screws. There is a 20-tooth XL gear bored out to 14mm to fit the Alpha stepper motor. The belt is a 150XL037. The original 3/8-inch mounting bolt was replaced with a longer 3/8-inch bolt, due to the extra thickness of the polycarbonate. The motor mounting bolts are 1/4-20 X 1-inch socket head bolts (so that the belt can easily be adjusted from the rear of the motor).

This version took about 1.5 hours to cut (Y-axis, 2 X-axis, no Z-axis) using a 1/4-inch O-flute cutter at 1.25-ips. Machining the spacers took forever. (Whatever made me think that using a metal mini-lathe would be fun? Can you say one full day machining and threading spacers?)

The 3:1 gear reduction on the X and Y axes reduced the normal Alpha 'chatter' problem enough that I can accept the occasional 'chatter' mark. Torque is 3X greater. Ted, at Shopbot, suggested that I try setting the Alpha drivers to 500 steps per revolution, rather than the default 1,000 steps per revolution so that the factory software/controller could supply pulses fast enough for a 3:1 reduction. The resolution on either 500 or 1,000 steps per revolution were nearly identical. In fact, the 500 steps per revolution setting seemed to produce smoother edges in the majority of cuts. It seems that the reverse should be true, but the proof is in the cut. I'm beginning to think that the increased torque is the main factor in getting better cuts. It might just be possible that the 3X increase in torque holds the 'inactive' axes in position so that they're not shoved around by the active axis.

I would have used higher reduction, but that would require a multi-stage unit - which would have been fairly bulky. The 20-tooth XL gear is the smallest that can be bored to fit the 14mm shaft on the Alpha stepper motor (according to W. W. Grainger's catalog). The 60-tooth XL gear fits better than the maximum size 72-tooth XL gear.

In any case, the 3:1 unit really helps with cut quality.

Evan Curtis Thu 24 August 2006 14:34


Nicely done! How is the unit tensioned up to the rack? The same as the Stepper was, with spring and turnbuckle?


Sheldon Dingwall Thu 24 August 2006 17:59

Very nice Mike. I wonder if the belt has any smoothing effect on the step marks?

Mike Richards Thu 24 August 2006 22:34

The motor can be adjusted vertically about 1/2-inch, which is more than enough to adjust the tension on the belt. The spring and turnbuckle are still used, except that I used 4-turns rather than the 3-1/2-turns recommended by Shopbot.

The belt is pretty tight. Actually, it is tighter than I would normally run a belt - only about 1/8-inch deflection when tightened. However, it could be that the natural 'softening' of the belt helps cut down the 'chatter'.

Gerald_D Thu 24 August 2006 22:49

Nice job Mike!

Any comment on how it stands up to sawdust?

Mike Richards Thu 24 August 2006 23:01

In my shop, I've mounted a sign stating, "Only 'anti-cling' sawdust is allowed". Because of that sign, I don't have to worry about sawdust collecting on the belt and gears.

Seriously, sawdust hasn't been a problem. After every sheet, I customarily sweep the table and blow off the gears, rails, and motors. So far, there hasn't been any problem of any kind with sawdust.

Gerald_D Thu 24 August 2006 23:43

"I'm beginning to think that the increased torque is the main factor in getting better cuts. It might just be possible that the 3X increase in torque holds the 'inactive' axes in position so that they're not shoved around by the active axis." I totally support this theory. I believe that a stepper motor cannot "hold" a micro-step. It will hold a full step, and even so, with some "spongy-ness". I plan to clamp a stepper in a vice, put a 2foot long lever arm on the shaft, and then feel by hand how a stepper actually steps, micro-steps and holds.

Good news on the dust. An open casing is probably better than a closed box.

Mike John Thu 24 August 2006 23:55

Can you send me the file to make that sign for my workshop,please?

..............Mike, the other one!(or is he the other one?)

Gerald_D Fri 25 August 2006 00:16

Your sawdust doesn't understand English

Mike John Fri 25 August 2006 02:36

Drat! foiled again!
Hang on! Does anyone from Utah speak real English?


Gerald_D Fri 25 August 2006 04:28

Yes, the sawdust does.

Mike John Fri 25 August 2006 05:34

Now that is true, as the trees were probably planted back in colonial days!

Mike Richards Fri 25 August 2006 06:53

Mike J,
English? Never heard of it! Here in Utah, we speak American - when we're speaking to outsiders. Among ourselves, we speak the pure Adamic language.

One of my concerns a few months ago was reduced current holding torque on stepper motors. It seemed to me then that as soon as an inactive motor went into standby mode (automatic reduced current mode to reduce motor heating), that axis basically lost almost all of its holding power, meaning that that axis could be pushed around by the active axis. My thought back then was to have Ted modify the Shopbot code so that whenever any axis became active, the other axes would move one step CW and then one step CCW continuously to keep all axes active (full holding torque).

I think that holding torque while motionless is the biggest argument in favor of servo motors. If I understand the concept of servo vs stepper, the servo motor/driver consumes almost no power when stopped, but instantly goes to as much power as necessary to resist movement - meaning that a servo motor can't be shoved around like a stepper motor. Of course there are significant factors favoring steppers over servos, but, if we have to use gear boxes to make a stepper work properly, why not go the full distance, install servos w/gearboxes and encoders (at 1/3 the cost of the Oriental Alpha stepper/driver)?

Gerald_D Fri 25 August 2006 06:57

And Adam spake to Eve: "...with what router shall we smite the apple tree?"

David Rosenbleeth Fri 25 August 2006 17:09

"You're in luck if you gotta a McCullogh chain saw"

Moral of the story: If you wish to smite the tree-use a saw; to carve it use a spindle and Geckos.

Been cutting parts with the Agek rig for a while now and my guys still have nothin but smiles.


Joe Crumley Fri 25 August 2006 19:15


Could we see some of these parts? I allways enjoy seeing products made with a router.

Mike Richards Thu 07 September 2006 15:08

Here's the 3:1 ratio belt-driven-transmission cut out of aluminum on my Alpha.

Cutting speeds in the 1/2-inch thick 6061 aluminum were 0.60-ips XY-axes, 0.05-ips Z-axis, 12,000 RPM, 0.075-inch step down per pass, using a HSS 3/8-inch cutter designed especially for cutting aluminum. Those speeds required flooding the cutter with tapping fluid made especially for aluminum. If you look carefully, you'll see some edge marking in various places. That was caused by letting the cutter get too dry (thus getting too hot).

After using the 3:1 ratio for several weeks, it seems to work perfectly for the kind of work that I do. There hasn't been any problem with 'chatter'. The open frame has not been a problem - no dust on the belt or gears. In short, with the 3:1 transmission, my Alpha is cutting like a real CNC router.

Dirk Hazeleger Sat 21 October 2006 09:23

Hey Gerald ran across this belt reducer produced in your neck of the woods. Ever heard of them? look at the bottom of the page

Gerald_D Sat 21 October 2006 09:44

Microdyne have always been out of my price range but I'll call them on Monday to see if they have respectable prices for these belt drives.

Gerald_D Tue 24 October 2006 04:32

$600 a piece. Ouch!

Michael Cunningham Mon 19 March 2007 21:37

Does anyone have a set of cad drawings to build a belt/gear reduction box.. maybe a list of parts? I need to build two belt reduction units for my nema 34 steppers.. say 3 or 4 to 1 reduction. I however an a complete noob at this with no lathe or mill. Something very similar to the ones shown here. Suggestions? Ideas? Plans? Does the metal kit have provisions for making belt gearboxes?
Does anyone make these on the side for people?


Greg Tue 20 March 2007 05:20


These are a couple of pics just to give you more ideas.

I don't have any drawings. And this is obviously not a mechmate.

Same principals apply though. Keep the relationship between the pivot and the pinion shaft no more than 45 degrees from horizontal. I would make it less if I was starting again.

I see the weak point of other designs to be the pivot mechanism. Any slop here goes straight to backlash.

Whenever I drill a hole and put a bolt or shaft through it, it usually ends up with about 3 to 5 thou slop. Then the rapid back and forth of the axis will increase this with wear.

That's why I put bearings on the pivot and made them a firm fit in their housing and on the shaft.

Also be careful in your belt selection. Mine are T5 and the "cog" on the belt is smaller than the "groove" on the pulley.

So far this has not showed up in accuracy but this is not a good place to start.

Michael Cunningham Tue 20 March 2007 14:46


Thanks for the info.. couple more questions..

On you pivot point I see you have a press fit bearing on each side of the shaft. What keeps
the shaft in place? Is there some sort of clips I dont see? Same question on the other shaft as well. Where did you get your gears from? How do you keep the large gear in line with the smaller one. Again another clip? Interested in making another set of these for some extra cash?


Alan_c Mon 04 June 2007 15:51

Reduction Drive by belt - Designs
This post originally made in the motors thread - now moved to here...

A request for all the great minds out there...

something we have not discussed much in detail is a design of a self built reduction drive using say a toothed belt and pulleys, although it has been mentioned on occasion and references are made to them at cnczone.

If one is limited to using only an ungeared motor (eg PK299-01AA or PK299-F4.5A) what are the best options / methods for achieving a reduction drive in the mechmate spirit - i.e. buildable with the resources one would have available for building a mechmate.

How do you decide on the belt and pulleys, how do you tension the belt, what sort of bearings should one use for precision etc?

If you are going to use a reduction drive, are these motors mentioned best suited or should you be looking at alternatives?

I suppose another interisting question would be, is it really necessary if so many guys are using direct drive, or am I just chasing the elusive resolution genie:rolleyes:

edit - sorry its late, found the answer to some of the above in the drive mechanism thread

Gerald D Tue 05 June 2007 01:29

Alan, all very valid and important questions.

Let's start with the belts. Most people are using XL (eXtra-Light) belts, but if I look at motor torques and speeds in the book, the XL seems to be too light. The proof of the pudding is whether the folk with XL belts have needed to replace them?

Then you can look for the range of pulleys available for the selected belt. The theory says that the small pulley must not have too few teeth for a precision/smooth/high-speed application. It seems that folk are getting away with "coarse" (few teeth) small pulleys?

On the big pulleys, I have found that the off-the shelf stuff doesn't go all that big, and then I worry about them behaving as flywheels. Idealy we want big alu pulleys and they are a bit scarce around here.

Converting the direct-drive motors with reduction doesn't seem to be an issue. You could actually go to smaller, lower-torque motors, but there is no apparent harm in recycling the big direct-drive motors.

Bearings, adjust and hinge points don't worry me - it is that darn book on the belts that is my problem.

(Years ago a lecturer asked our class to design a gearbox and gave us the torques/speeds/etc. The smallest box any of us could come up with was about the size of a chair. Then he showed us what specs he gave us - Ford escort! Where the reliable box is about the size of a shoebox. This is what happens if you stick to the book.)

Richards Tue 05 June 2007 06:52

So far there has not been a problem with the XL belts. (Remember that the specs on some stepper motors specify < 8 lbs. of radial force on the shaft. The belt design keeps radial force to a minimum.) My plan is to replace the belts every year when I replace the pinion gears. That way I'll probably never now how long a belt will actually last. I'm coming up on about 8-months of use with absolutely no adjustment of any kind.

If you look carefully at the JPG in the post at the top of this thread, you'll see that there is about 3/8-inch of up/down adjustment for the motor. That is more than enough for belt tensioning.

The toothed pulleys that I use are 20-teeth an 60-teeth. The belt is the 150-XL size. Center to center distance on the pulleys is 3" minimum and 3.375" maximum.

I've made several versions, all very similar to the polycarbonate version shown in the first JPG. My latest version uses Delrin instead of polycarbonate. The Delrin works perfectly, but I'm willing to guess that 1/2-inch thick Baltic Birch plywood would work just as well.

Gerald D Tue 05 June 2007 10:00

Oriental Motor allow 75lbs radial load on the shaft of the PK29_ series motors, 0.4" in from the tip of the shaft (57lbs right at the tip), so that is not an issue for belt loading (nor spring-loaded direct pinions for that matter).

McMaster-Carr offer the following pulley teeth numbers for XL series belts:
10|11|12|14|15|16|18|20|21|22|24|26|28|30|32|36|40 |42|44|48|60|72

If a 10 tooth were used with a 72, the ratio would be nice, but only about 3 teeth of the small pulley would be engaged with the belt. To get more "wrap" on the small pulley, the center distance would have to be increased drastically, making a big drive system. Mike's 20/60 choice is a good conservative selection, giving a compact package.

Richards Tue 05 June 2007 21:57

I couldn't find the specs for the radial load on Oriental Motors, so I'm happy that you found that information. In fact, I may have been remembering data from a tech seminar that Oriental held years and years ago. (With my short term memory beginning to be just a memory, I find it hard to remember anything that happened less than 15 years ago; however things that happened in my ancient past are still sharp and clear.)

The reason that I selected the 20 tooth pulley and the 60 tooth pulley for my units is that the 20 tooth pulley is the smallest pulley that I could use with the 14mm shaft of the Alpha stepper motors. Pulleys smaller than that had a recommended maximum bore less than 14mm. The 60 tooth pulley was mostly selected because I had six of them on hand and only two 72 tooth pulleys. Also, the 60 tooth allows a more compact unit than the 72 tooth pulley. The difference in ratios 1:3 vs 1:3.6 does not seem to be an issue.

You're correct about the danger of using a 10 tooth pulley that has only 3 teeth engaged with the belt. I used a 10 tooth and a 60 tooth pulley extensively with a PK268 motor in a Kodak S-printer electronics package. Testing showed that I needed to use a much longer belt to get more teeth engaged. That particular design required almost no torque and minimal speed, which is much different from the requirements in a CNC router.

Gerald D Wed 06 June 2007 00:21

Mike, the radial loads spec is from the sheet that comes in the box with the motors. I hadn't seen that same sheet online before, but a new search this morning revealed this excellent document. See page B-37

I have a few L-series pulleys lying around, but they are a lot bigger and downright ugly. The book on belt drives has to assume continuous operation at full speed & torque. On the stepper driven CNC router we don't get to full speed at full torque simultaneously, nevermind doing it consistently 24/7. If the XL belt is working in this application for everyone, then the book can be revised.

Gerald D Wed 06 June 2007 03:23

Being an ex-automotive person, I have a fondness for using car parts.....

A waterpump bearing could be useful to the belt drive designers. That is if the motor is swung around compared to the above layouts......

All times are GMT -6. The time now is 14:11.

Powered by vBulletin® Version 3.8.3
Copyright ©2000 - 2020, Jelsoft Enterprises Ltd.