Problem with Solsylva DIY Machine

[mlammert]

Hi all,

I finished building the 25 x 37 Dual Leadscrew Machine from Solsylva about 3 months ago (http://www.solsylva.com/cnc/25x37dual_leadscrew.shtml)...

It probably has about 45 hours of cut time on it...

All of a sudden the Y axis (longest axis of table) has suddenly started binding up and stalling the stepper... This is the axis that uses the belt system to turn the leadscrews...

What seems to be the problem (from what I can tell) is that the belt is all of a sudden starting to wonder on the tension pulley... And, when the belt wonders too far it either rubs the wood of the machine causing a stall... Or, it wonders so far that it stretches the belt causing too much tension and stalls...

I have been pondering this for days... I can't understand why it all of a sudden has started doing this and for the life of me cannot seem to fix it...

I have been focusing on the belt system... But, it now occurs to me that maybe the problem lies elsewhere and it is just showing up in the belt system...

Has anyone else experienced this problem and/or maybe have any suggestions for a starting point to begin troubleshooting...

I was able to carve some beautiful work with it while it was running flawlessly... And, now all of a sudden it is useless... Very depressing...

Any and all suggestions are greatly appreciated!!!

Mark

[solodex2151]

Have you checked to make sure that your gantry isn't binding? If one leadscrew slips and loses a couple of steps, your gantry is going to bind and stall the stepper. I would recalibrate each leadscrew by taking the belt off and turning them to move the gantry all the way forward. Put the belt back on and try running it.

Nate

[mlammert]

Thanks Nate!!!

I will put this on the list of things to start troubleshooting...

I wonder however, if this would/could cause the sudden issue of the belt wondering on the tension pulley???

Thanks again to everyone!!!

Mark

[Richster]

Hmmmm sounds like possibly one side is binding as others have said, and the imbalance in torque between the two sides is causing the belt to wander somehow. I know on mine when I started out I had to bash around (with a big hammer) the bolt which holds the pulley since it wasn't quite straight in the hole. That made the belt run centered more or less, and it has been fine since. I like the idea of a pulley with a ridge to guide the belt, if you can find someone to knock one up on a lathe.

I'd disconnect the belt, and revisit each leadscrew to check they turn freely over the entire range of movement, like when you did the initial assembly. I don't have anything like your hours yet, but it's sweet so far

BTW the plans call the long axis X - not that it matters whatsoever, but had me confused for a short while...

[mlammert]

We basically dismantled the entire machine over the weekend and reconstructed and realigned everything...

I am 99.9% sure that the side rails and gantry are square...

Before all this we tried making our own pulley with side walls as others have mentioned... This did not work... The belt still drifted and simply bound up when it reached the side walls...

The lathe thing sounds like a good idea...

Before all this we did try bending the pulley bolt to try and compensate for the drift... We can try more...

I am also considering getting an auto-tension pulley from this company:
http://www.econobelt.com/Q460/RFQ/de...nsioner/13.htm

The reason I am thinking about getting one is that for brief moment I thoughts I fixed the belt drift... But, then realized that I had forgotten to tighten up the stepper mount after connecting the belt... When the stepper mount was loose it drifted a little but like an 1/8 of an inch... When I tightened everything up, it started drifting to the edge again and binding...

Thanks for the advise... Off for more troubleshooting...

Mark

[BigMike4u]

Sounds to me like your motor pully is pulling out of alinement. Maybe flexing under load or some such thing.

Mike

[mlammert]

Mike,

Would you be so kind as to possibly go into more detail on this...

Could you possible shed some light as to what might cause this (besides the obvious flexing) and maybe some ideas on how to correct...

Thanks,
Mark

[CarveOne]

In the pulley alignment adjustments that you make you should take into account that the belt always drifts toward the tight end of the pulley. Each of the shafts must be parallel to all of the other shafts (lead screws, bolts, and stepper shaft). Even a minor adjustment of any one pulley can make the belt drift more, or less, or overshoot in the other direction.

If the belt tension is tight enough to pull the lead screws together (and the stepper shaft and idler are parallel to each other) the belt will drift toward the router frame end board. I can easily do this with my 1/2-10 acme lead screws. You need just enough belt tension to remove any free play but not enough to bend the lead screws together even a little.

CarveOne

[BigMike4u]

Carveone said just about all I had in mind. I don't have any hands on time yet. I'm building one, but real slow at getting it done.
I am going on past problems on rock crushers. Belts will run true until loaded and then everything goes haywire. This means to us that a motor mount or bearing mount is broke, out of line, bent, etc.
I would bet that if you run your machine and put pressure on the stepper motor that you can make the belt move on the idler.
Also as carverone said, to much tenshion on the belt can cause problems.
I hope this helps and I'm looking forward to the fix. I might have the same trouble in the future.

Mike

[CarveOne]

My belt drive experience comes from testing and aligning an old RCA shipboard radar antenna drive at the ITT Telecom factory in Raleigh, NC. The belts were 3" wide and were very hard to align, then tighten the bolts while keeping the belts running true. Each of three belts had to be aligned in sequence. Fortunately, we only had to build 12 of them. Any more than that and I would probably have gone suicidal....or quit the job.

Once it was running correctly, it could probably be used to crush rocks.

CarveOne