[MrBean]

What do you guys think about belt driven axis?
I've had my machine running for just over a week now. The X axis is driven by a 30mm wide, 5mm pitch, HTD belt. The Y is Ballscrew, hardened shafts and bearings. The Z is ballscrew, THK track & skates.

The machine's a bit of a Frankenstein, but that's the parts I got for free, so I used 'em. It's going great. The only thing holding it back, is me. (still learning)

If I had to pay for parts to build a second machine. I'd like to go with belt drive for X & Y, as the cost compared to ballscrews is about 1/10, but performance still seems good (at least it does to me). Also, if the machine is moving gantry with a wide span, it's far easier to put belt drives both sides, and link with a shaft. I suppose it's a bit like rack/pinion drives.

Is there any reasons you would no go with belt drive. E.g. accuracy, wear, backlash...etc.


Thanks for any input.

[sol]

Well, your question has been sitting dormant for a few days so this fool will jump in where the wise fear to tread.
From experience I know that racks can allow the pinion to jump and skip teeth thus losing position. A good belt system can be built to avoid this while still giving the advantages of a rack including no long whipping lead screw and more convenient placement. But you already knew this...
There have been positive comments about belt drive on other CNC forums...
I intend to replace my 5ft 1500mm lead screw with a belt when I get time....
Dean Couillard of Axxus made a number of tables that use belts for X and Y; I cannot find the pics on his site anymore but did happen to copy this one a while back which shows the gearing for a belt on Y.

[ger21]

Dean Couillard of Axxus made a number of tables that use belts for X and Y; I cannot find the pics on his site anymore but did happen to copy this one a while back which shows the gearing for a belt on Y.

http://www.axxustech.com/Axxus4x8/axxus_4x8.htm


I think the reason you're not getting any responses is that very few people are using belt drive systems.

How do you like the system you're using. I think the belt drive has a few advantages, such as no backlash, and not having to worry about keeping a greased or oiled screw clean.

[samualt]

The only downside I see in using a belt drive is that if you have motors with a fast rpm, say 2000 rpm, then your looking at a huge step-down to run the belt. All those extra pulleys, bearings, and belts cost money and time. Whereas a ball screw usually takes care of some of that by having 10 TPI (turns per inch). Ball screws and lead screws have a built-in step down.
If you have to have a large step down then your looking at about 2 pulleys for the step down and 2 more for the axis and one for the motor. 5 pulleys, 3 belts, and extra bearings. It's probably about the same price as a ball screw all said and done.

However, like ger21 said, there are advantages to belt drives also!

[MrBean]

For me, my belt driven X axis is performing best out of the three. My machine is mostly made out of crap destined for the bin. The stepper drives are also crap. No microstepping & no current limiting, so steppers are running at rated voltage. (slow). Belt drive suits this situation quite well. Low revving motors, with some gear reduction, and bugger all amps * volts. I am getting 90 IPM rapids on X axis with no lost steps. Resolution is not as good as the ballscrew axii though. Minimum movement of X is 0.007" which is ok for me at this time. Another bonus is, it sounds really cool compared to the ballscrews.

This is my first machine and is a learning experience as much as anything. CNC is only a hobby for me, so I cannot justify spending huge sums of money on equipment. I do plan on building some PIC microstepping drivers I saw in the forums somewhere. That's my next step. More power to the motors.

[balsaman]

what ratio are you running on the belt?

Eric

[MrBean]

The main belt runs on 26T pulleys either end, 5mm pitch.
The reduction is 3.6 : 1

So:

1/(26*5)=0.0076923

0.0076923*3.6*200=5.53844 steps/mm

1/5.53844=0.1805562mm/step or 0.00710"

That's with a 200 step/rev motor. Microstepping would increase the resolution.

I'm going to try a 6:1 ratio this week, time permitting.

[balsaman]

I want to cnc my lathe. I am considering a belt on the carriage. Mostly for no backlash. I am wondering about "strech" tho. Maybe it's not a good choice for metal working machines.

E

[MrBean]

Not sure about stretch specs of belts. I do know that mine will lose steps well before the belt would stretch or snap, which ever would come first?

If you have a CNC router that can do aluminium, you could try machining your own pulleys. Tho' they're not expensive to buy.

I still like my belt drive and plan to build a small, entirely belt driven machine as an experiment.

Regards Terry...

[Kookaburra]

Mr Bean,

From experience watching my customers come up with various ideas to drive their axes, there have been mixed opinions regarding belt drives. I have one customer in particular that is using a 25mm (1") wide belt driving an axis 3.5m (137.8") long and there was a noticable bounce in the belt as the machine axis started and stopped. It took a little longer to tune in the PID control of the servo drive but in the end it ran quite smooth. We have other customer that runs over a shorter distance without any problems at all. I personally think that the guy with the 3.5m travel should have used a much wider belt and used support idlers on the belt return side to eliminate some of the bounce. I guess what I'm saying is that belt drives can be good if they are engineered properly for CNC use in your particular application. Nice and cheap, I agree, and easy to install. Let us all know what you decide to do when the time comes I will be interested to hear the outcome.

[JavaDog]

Hmm...I didn't even think of belt drive when starting to design my CNC router. I might have to go back and look at that - lower cost is always a plus!

Besides, my set-up will be fairly small (26" x 30")...so it sounds like belt drive will work well.

[ger21]

Hmm...I didn't even think of belt drive when starting to design my CNC router. I might have to go back and look at that - lower cost is always a plus!

Besides, my set-up will be fairly small (26" x 30")...so it sounds like belt drive will work well.

Lower cost would depend on what you are comparing it too. Ballscrews probably. Acme with Delrin nuts, belt drive would probably be more.

[JavaDog]

Lower cost would depend on what you are comparing it too. Ballscrews probably. Acme with Delrin nuts, belt drive would probably be more.

Well, I am looking for durability and accuracy within my budget (aren't we all!) and I like the idea of belt drive and no backlash. My particular needs would probably be best described in another thread however...