Hi, I am new machining and especially CNC. I bought a Grizzly mini mill to learn milling with. I then found out about the fascinating world of CNC and now have a burning desire to convert my mill to CNC. I have purchased most of the raw material to build the Z axis from Hoss. I do not have the ball screws or rods yet. I have not yet pulled the trigger for the X & Y material. One aspect I am struggling with is direct drive verses belt drive. I don't like the look of motors that are directly coupled to the axis ball screws.

Is there an advantage doing direct coupled rather than driving the axis with a belt. It seems that most of the conversions use direct coupling but that sure hangs the motor way out in the breeze so to speak. There are a few conversions that use a belt to drive the axis.

I would like to know the pros and cons of each method if ya'll could please discuss this for me.

My plan is to use Gecko drives and probably motors/power supply from Keling. It looks like 270 oz for X&Y and something in the 380 oz for the Z. Any suggestions on the actual motor. From what I can tell a 48v power supply is the right power as well. Am I correct on that.

I will be using Mach for the controller software.

Looking forward to discussing and learning more.

RWW

Direct drive or belt drive is strictly user preference.
2 pulleys and a belt will cost a little more than a flexible coupling per axis.
The motors should be best matched to the driver and power supply,
the G540 with 48V PS is suited for a 318 oz/in for the X and Y and a 387 oz/in for the Z.
A 1:1 pulley ratio will give the best speed.
Gearing down on the X and Y could let you use a smaller motor but you only save like $10
and the motor wouldn't be as good a match as the 318's which will work great direct drive or 1:1.
Hoss

Thanks for the reply Hoss.

Will the belt drive introduce backlash into the system? And what size Pulley would you recommend? Size and number of teeth?



RWW

from my very limited experience, it seems that the belt drive rigs usually are used with servo motor set ups.

not sure why this is, but suspect that small servos can run fast enough to deliver sufficient table speeds speeds while geared down for power.

that's my guess

then again, if you've got money to pump into a servo setup, you're likely not going to be concerned about the add'l cost of the belts, pullies and more complex motor mounts.

Using a belt system, I think (you'd have to ask someone with more engineering than I have) you can effectively add more accuracy to a cut because of the larger number of microsteps due to the gearing down of the system. If you run your screw at 1:2 with your motor, you get 2X as many steps in a full 360 degree rotation (I think-and please correct me if I'm wrong on this) resulting in a smoother, truer part.

One of the effects of adding a pulley to the setup, though is you introduce another set of moving parts to the system. Every part you add will introduce more errors to the overall equation, theoretically, and thus results in more errors. As this is a hobby/learning machine though, I doubt this makes a difference unless you have a contract you want to run on the thing for tolerances as high as this might impact.

-Farasien

You're both right on.
A servo is opposite of a stepper, it's weak going slow and strong going fast.
Gearing down gives it more power for milling speeds with plenty of rpm left for rapiding.
Gearing down a stepper will give more accuracy but in most system, direct drive is more than accurate enough with microstepping from the driver.
The g540 10 microsteps with 5tpi ballscrews will give .0001 resolution.
Doubt most benchtops would need any more.

20 tooth XL (.200) pitch would be good, be about a 1.25 OD pulley.
The 3/8 belts from SDP/SI (https://sdp-si.com/eStore/) I've used have kevlar inside to I wouldn't worry about backlash from stretch with such short lengths needed.
Hoss

Using a belt system, I think (you'd have to ask someone with more engineering than I have) you can effectively add more accuracy to a cut because of the larger number of microsteps due to the gearing down of the system.

Gearing down a stepper will give more accuracy but in most system, direct drive is more than accurate enough with microstepping from the driver.


This is a common misunderstanding. It is due to the confusion between RESOLUTION and ACCURACY.

Adding a belt and gearing down the lead screw will more likely REDUCE ACCURACY. However it will INCREASE RESOLUTION.

Accuracy is the tolerance to which a specific positional location can be selected.

Resolution is essentially the pitch of the discrete positions available (microsteps for example) and doesn't actually indicate the positional accuracy of these positions.

My intention is to build a CNC machine with the stepper motors not hanging straight out from the end of the X&Y axis. I think that I will do a 1:1 ratio. Hawkjet, I think that doing a 1:1 I will not need to be concerned with the resolution/accuracy issue that you mention. Correct me if I am wrong.

RWW

Hoss, if I do a belt drive conversion will the ball screws that you offer on your website be long enough? My thinking is that they will because I am basically replacing a motor coupling with a pulley.

Your thoughts please.


RWW

Hawkjet, I think that doing a 1:1 I will not need to be concerned with the resolution/accuracy issue that you mention. Correct me if I am wrong.
My apologies if was not clear. My intention was only to provide a vocabulary lesson. I don't know if (in your situation) there is an issue or not. There are pros and cons to both methods and as the designer, you would weigh them against your needs.

Also, accuracy/resolution questions are not the only factors in determining motor/leadscrew drive ratios. Matching the motor speed range to the leadscrew speed is a major consideration.

I might be missing something but why would you use a belt drive with a 1:1 ratio. The motors will hang to the side and then out front anyway. In addition you will need to take care of radial as well as axial loads with the bearing arrangment. Seems like a lot of work for zero gain to me.

Phil:)

My intention is to build a CNC machine with the stepper motors not hanging straight out from the end of the X&Y axis. I think that I will do a 1:1 ratio. Hawkjet, I think that doing a 1:1 I will not need to be concerned with the resolution/accuracy issue that you mention. Correct me if I am wrong.

RWW

Seems like a lot of work for zero gain to me.
Not necessarily. First, I don't believe it is a lot of work. Mounts are still needed and a coupling is still needed. When Bridgeport made their Boss 3 CNC with steppers (I have one), they had a 1:1 ratio and the motors were turned so they did not stick out - they were under the table.

Also, there are potential challenges with a direct drive and coupling arrangement. Care needs to be taken to get shafts to align and couplings can also be problematic (although with hobby machines there is far less demand on them). Plus there is the added cantilevered length to handle. The bearing consideration you mention is minor if it even exists.

Most commercial CNCs use belts. IMO, I'd rather deal with a belt than a coupling because there is more design flexibility.

If I go with a 1:1 belt design for X&Y axis resolution remains the same as it would with a direct drive. Correct?

Accuracy could suffer with a 1:1 belt drive if backlash is introduced by the belt. Correct???

I am concerned about coupling in direct mode. Currently I do not have a lathe so getting the couplers machined properly.

Maybe I should be about concerned about a belt drive system as well but right now I think that I will be able to do what
I need to do to make a belt drive system work with just my mill. I do need to research motor shaft size and pulley bore size.
It is possible that I will not be able to do anything without first purchasing a small lathe.

RWW

You are correct sir.
I can machine the ballscrews to whatever length is needed and make the tip diameter
fit the bore of the flexible coupler if you want,
no need to machine them on your part.
You can order them SDP/SI and McMasterCarr with different bores as well as the timing pulleys whichever route you take.
Would imagine you want to use a belt system to be able to tuck the motors in under the table
to give the mill a more compact footprint.
Hoss

How would you "turn" the motor on a mini mill x-axis without having interference with the base or having it hang a long way out to the side.

Phil:)

When Bridgeport made their Boss 3 CNC with steppers (I have one), they had a 1:1 ratio and the motors were turned so they did not stick out - they were under the table.

Also, there are potential challenges with a direct drive and coupling arrangement. Care needs to be taken to get shafts to align and couplings can also be problematic (although with hobby machines there is far less demand on them). Plus there is the added cantilevered length to handle. The bearing consideration you mention is minor if it even exists.

Most commercial CNCs use belts. IMO, I'd rather deal with a belt than a coupling because there is more design flexibility.

How would you "turn" the motor on a mini mill x-axis without having interference with the base or having it hang a long way out to the side.
I dunno... it would depend on the size of the motor, the length of the belt, the mill geometry, the clearance of the table, etc., etc., etc.

It's an opportunity for the designer to be creative, right?

My intention is to build a CNC machine with the stepper motors not hanging straight out from the end of the X&Y axis. I think that I will do a 1:1 ratio. Hawkjet, I think that doing a 1:1 I will not need to be concerned with the resolution/accuracy issue that you mention. Correct me if I am wrong.

Please clarify what problem you're intending to avoid. There is no problem with a stepper "hanging straight out" from the end of the axis. A belt-drive servo would hang off the edge the same way too, unless you're talking about replacing the whole leadscrew with a very long belt (which is possible).

They are typically not exactly direct-coupled. We use a flex coupler between them to isolate the motor from misalignment and leadscrew vibration forces.

Here is a link to a YouTube post that shows a belt drive on the Y axis. At a minimum this is what I want to do. If I can figure out a way to tuck a motor in between the frame and table for the X axis I will do that as well.

I do not know who the YouTube poster is. Just want to show what I am trying to accomplish.

RWW

YouTube - Test run of my CNC machine


p.s.

Hoss, I will be in touch with you to discuss machining of the ball screws after I get the motors and make the mounting plates. I am going on a roadtrip next week to Myrtle Beach. Be gone for a week.