PDA

View Full Version : Belt drive on a router



John S
09-09-2012, 10:03 PM
Playing devils advocate there and like some opinions for and against with the two methods described below.

Both methods will rely on a tranverse mounted stepper motor at the rear, this will drive two pulleys, one either side via jack shafts.
All pulleys will be well supported on bearings.

Two lengths of open belting will pass from these pulleys to twin pulleys on the front of the machine and the open ends will be secured to each side of the two uprights.

Basically open belt drive to replace two ballscrews and this will be over a 1 metre length at max, probably more like 800mm.

Now the different design points.

Plan A will have one end of the belt secured on a fixed bracket and the other on a bracket that can be adjusted on a turn buckle.

Plan B will have both ends fastened to the same fixed bracket and adjustment will be handled by independently spring loading the front idler sprockets.

Thoughts on this please ?

JAZZCNC
09-09-2012, 11:33 PM
Plan A for me John.? Much simpler and just has affective.

Is this to move a gantry for a router.?

John S
09-09-2012, 11:57 PM
Yes moving gantry design

Robin Hewitt
10-09-2012, 12:10 AM
I'm just about to try out open belt. I went for Plan B, I pre-tension springs with circlips then run a bolt in to connect the tension roller to the spring.

The idea is that I can slacken the belts when not in use then set them by taking up the slack plus 2 turns.

I'm using 16mm T5 steel reinforced belting. I hope to find out how stretchy it is tomorrow, will 2 turns be enough?

m_c
10-09-2012, 12:25 AM
I'd say plan A, provided it's adjusted correctly.
However plan B would allow for a bit of idiot proofness if anybody happens to go tinkering with belt tensions.

JAZZCNC
10-09-2012, 12:33 AM
Yes moving gantry design

Well couple years ago built a belt machine for a friend in Ireland have a look here. New Machine Build Belt Driven extrusion framed wee beasty!! - CNCzone.com-The Largest Machinist Community on the net! (http://www.cnczone.com/forums/cnc_wood_router_project_log/108333-belt_driven_extrusion_framed_wee.html)

Built from scavenged bits with a few new. Check vids in post #5 and you'll see it wizzing about.!!
In the Vid with moving gantry it was running 3:1 ratio and that wasn't enough so ended up at 6:1. @ 3:1 the motors weren't happy at low speeds.
I was restricted with room to use larger pulleys and just happened to have a spare belt and pulleys so thats why it ended up using the setup it did. . . Not ideal but got the job done.!!

The point telling this was to make you aware of the ratio your likely to need for smooth machine at low feeds.

m_c
10-09-2012, 01:24 AM
The idea is that I can slacken the belts when not in use then set them by taking up the slack plus 2 turns.

Why?
Modern belts don't stretch.

Jonathan
10-09-2012, 10:29 AM
Why?
Modern belts don't stretch.

Everything stretches - it's a matter of how much, which in this case depends on things such as the width of the belt and type of tension members in the belt.

Page 7 of this document has a graph to show how they stretch:

http://www.gates.com/facts/documents/Gf000289.pdf

If you're using the belt on a machine with low or zero cutting forces (e.g. laser cutter), then the gantry can be light so the force should be small enough for this not to be an issue. However if it's for a router as the title suggests I would do the calculation to find out how much the belt will stretch, or at the very least use the widest belt you can fit.

Robin Hewitt
10-09-2012, 11:20 AM
Why?
Modern belts don't stretch.

Belts may be modern but I'm not :rugby:

If the belt doesn't change it's elongation over time when left in tension that could save me a very inconvenient switch. I was thinking to measure the length of the bed in stepper resolution at start up then scale the G code to match.

This is kind of crucial because I have designed for a 3/4" aluminium tube inside the gantry which carries the head control wiring to the flexi cable conduit that connects to the head.

If the belt needs constant calibration this tube goes full length to get a switch at the far end.

If the belt doesn't need constant calibration the tube only goes half way and the fitting is different.

It will be a pig to fix if I get it wrong.

This is the last part before I can fit the long belt, so what do I do? Full width or half way?

m_c
10-09-2012, 03:26 PM
You don't have to retension timing belts on cars, and the latest ones last 150,000miles or 10 years with no maintenance.
And given quite a few run with fixed tensioners, stretch isn't a concern.

deannos
10-09-2012, 04:01 PM
You don't have to retension timing belts on cars, and the latest ones last 150,000miles or 10 years with no maintenance.
And given quite a few run with fixed tensioners, stretch isn't a concern.


Every car i have had has always got some form of tensioner on the timing belt, in most cases its a idler pulley under spring tension (auto belt tensioner).

You could argue this just keeps the belt tight against the pulleys, but i have noticed a slight difference in length when i have changed them in the past. But whether you would get this much on a router to cause major problems, who knows.

JAZZCNC
10-09-2012, 07:12 PM
But whether you would get this much on a router to cause major problems, who knows.

I do and no it doesn't.!! Been using belts on my machine for years and they haven't stretched to the point I can see or measure. Had a hand in few belt drive machines and none of them suffer from belt stretch and they are spot on accurate and super reliable.
Must say doesn't work out much cheaper than ball-screws after you've bought all the belts,pulleys etc Unless like John S you can hob your own and have time.!

Was surprised to find the Jobs 5 axis machine I've been trying/about to buy connects to the ball-screws with 25mm timing belts and the moving head(X axis) on that thing must weigh 7-800Kilo. The bed(Y Axis) alone must weigh 200K and this thing was designed for machining titanium and precision moulds for Mclaren F1 to super fine tolerances.! So they wouldn't be fitting them on this quality machine if they stretched to any detrimental amount.! (That said they are probably something special and cost 500 each. . Lol)

Jonathan
10-09-2012, 07:59 PM
Belts offer numerous advantages for driving ballscrews, the stretch is almost invariably negligible there since the belts are very short. However if you're using belts to replace ballscrews (i.e. the subject of this thread), then the stretch will be a factor - otherwise belt driven machines would be good for machining more than just soft materials.

JAZZCNC
10-09-2012, 08:20 PM
Belts offer numerous advantages for driving ballscrews, the stretch is almost invariably negligible there since the belts are very short. However if you're using belts to replace ballscrews (i.e. the subject of this thread), then the stretch will be a factor - otherwise belt driven machines would be good for machining more than just soft materials.

Oh here we go.!!! . . . My machines is nearly 1500mm wide so thats over 3000mm long belt and stretch hasn't affected it yet.! This belt has been on machine probably coming up 2yrs, the first which I damaged had been on over 2yrs and hadn't stretched in that time.!

It's all relative to the machine and work load and I know, Again thru experience, that this size machine doing the kind of work John intends then it will be a very very long time before belt stretch even remotely comes into the frame.!! . . . Yes they clearly will stretch over time that's a given but ball-screws wear-out and I know a good quality belt will last just nearly has long has any Chinese ball-screw.!!

Jonathan
10-09-2012, 08:37 PM
We're talking about different things. I'm not talking about the belts stretching over time (since as you say that just doesn't affect us). I'm saying when you apply a force to a belt it will stretch then spring back when you remove the force. So I agreed that when driving a ballscrew with a belt there's no problem. However I thought John S was asking about using the belt instead of a ballscrew? I.e. long open belts on either side of the gantry to push it along...

C.AlveSilva
10-09-2012, 08:47 PM
Hello,

What do you think of using T5 Timing Pulley for my router?

I'm using HGW 20 Hiwin Block, 1605 ball screws, 4Nm step motor and I'm thinking of use Timing Pulley running at 1:1 ratio.

30 teeth pulley.

Thank you

Jonathan
10-09-2012, 08:48 PM
What do you think of using T5 Timing Pulley for my router?

You could do but the motion is 'smoother' with HTD pulleys and T5 may have more backlash, not sure.

m_c
10-09-2012, 08:51 PM
Every car i have had has always got some form of tensioner on the timing belt, in most cases its a idler pulley under spring tension (auto belt tensioner).

Not everything does though, and I normally deal with vehicles that don't.

You could argue this just keeps the belt tight against the pulleys, but i have noticed a slight difference in length when i have changed them in the past. But whether you would get this much on a router to cause major problems, who knows.
The normal difference you'll notice between new/old is flexibility. If you were to put them on a proper belt measurement jig, there would be neglible difference in length.

m_c
10-09-2012, 08:56 PM
Belts offer numerous advantages for driving ballscrews, the stretch is almost invariably negligible there since the belts are very short. However if you're using belts to replace ballscrews (i.e. the subject of this thread), then the stretch will be a factor - otherwise belt driven machines would be good for machining more than just soft materials.
Stretch/Elasticity during use will be a factor, but loosening the belts of when not being used is going to have a neglible effect.

Jonathan
10-09-2012, 09:01 PM
Stretch/Elasticity during use will be a factor, but loosening the belts of when not being used is going to have a neglible effect.

Agreed..
The manufactures do say you should check the tension after running for a little while as they stretch a bit when first installed then 'settle down'.

JAZZCNC
10-09-2012, 09:47 PM
We're talking about different things. I'm not talking about the belts stretching over time (since as you say that just doesn't affect us). I'm saying when you apply a force to a belt it will stretch then spring back when you remove the force. So I agreed that when driving a ballscrew with a belt there's no problem. However I thought John S was asking about using the belt instead of a ballscrew? I.e. long open belts on either side of the gantry to push it along...

Ermm Ok but still the same applies in John's application belts will be fine.

boldford
10-09-2012, 10:06 PM
I'm unsure where I saw it but one idea is to have the belt that wraps round the gantry mounted stepper motor meshed with another length of open belt firmly fixed to the machine bed. The only length of belt "unsupported" is that short length which is picked up to wrap round two idlers and the toothed pulley on the stepper motor IYSWIM.

You could also try the alternative of roller chain.

John S
10-09-2012, 10:30 PM
I.e. long open belts on either side of the gantry to push it along...

I was thinking more about pulling it, tried pushing a bit of string years ago and not a good experiment :panda:

boldford
10-09-2012, 10:39 PM
I was thinking more about pulling it, tried pushing a bit of string years ago and not a good experiment :panda:Bit like putting a sausage up an entry.
Have you looked at http://www.mycncuk.com/forums/linear-rotary-motion/1636-idea-linear-belt-drive.html