Check out these weird servos belt drives, they are probably too expensive for the PhlatPrinter, but the concept sure is cool.
http://www.bell-everman.com/serbvideo.html

 

"As for speed, the standard ServoBelt offers acceleration of 4 g and speeds up to 4 m/s. An optional bearings upgrade will increase the speed limit to 10 m/s, which is well into linear-motor territory."
At 4 m/s, the Z-axis would go from stop-to-stop in roughly 0.139 seconds! That's not nearly enough time to say "Oh sh**!"

 

From what I have seen on the web, the belt drives are common in diy cnc machines. I don't see why it couldn't be made, if someone really wanted one. Do you even need those, teeth grooves in the base under the belt? Obviously, it's not going to go that fast, unless it's precision made.

One time at work, I was asked to make a linear brushless motor. I was given the spec for the coils and magnet, from our motor design guy. I remember 3 magnets on the gantry and coils along the entire length, every 3 wired in series. It's not any harder than, what the guy that wind their own motors do. And mine did run that fast. The only problem is you needed to know position and a optical encoder will be expensive. I wish I knew more about stepper construction. A linear stepper would be the ticket.

Mark,
That's cool. I would have never consider that concept doable. I would really love to see a video of that stepper running. If you come across a video, please let us know.

-Kwok

 

I wonder if we could build something similar to these belt drives for our stock steppers? If we did would it be patent infringement if we didn't sell them? Kwok, it does seem like you wouldn't necessarily need the grooved belt to do this type of drive, friction may be enough, on the other hand maybe the grooved version has less friction and therefore is faster. What I like is that there is probably almost no side stress on the stepper pulley.

 

Mark these platen based two axis steppers you posted on are Whacked out! I can't even begin to think how those work!

 

meistertek,
Actually I meant the groves in the metal base. You can see it at the end of the video. I would still want the tooths in the belt, which mesh with the stepper pulley. I wouldn't rely on friction because if it slips; you've lost your position big time, not just a few steps like on the PP.

Belt drives are common. Just do a search. Here is one of the same type:
http://www.youtube.com/watch?v=9QgnD-bgd_4

The more common type is where the belt is clamped to the moving part to move it. And looky here, Mark, I finally found the CNC machine that I remembered with one of our model airplane brushless motors:
http://www.youtube.com/watch?v=-cOkD-4wF2I

Look how fast they are!

-Kwok

 

No, it would not be patent infringement. ;-)

 

If you do a search for "Sawyer" motors, you'll find a lot about these. You're right, it's a hybrid stepper, with magnets and coils in the forcer. The platen is etched or machined with a cross-hatched groove pattern to make the same type of teeth as the outside of a stepper motor's rotor. The grooves must be filled with plastic or epoxy, because the only way to fight the *huge* attraction force is with an air bearing. So the platen must be very, very flat, and heavy steel.
My buddy Mark Wilson designed Baldor's (they bought his company for that and other things).

Thanks for the kind words about ServoBelt. I can help with anyone wanting to DIY one.

 

Cool hopefully I can help. I have some ideas for you, too! Being from a different world (in that my company is not set up for this), I am faaaascinated by low budget CNC designs, and have spent a good deal of thought on it over the years, with no outlet through Bell-Everman.

 

Nice site and great product, BTW. I hope it's doing well for you!

 

Hi Mike,

It's nice of you to chime in and your kind offer of help is appreciated. I saw you servobelt thread over at cnczone. Also, when you responded in a thread about cable drives. Do you have an opinion on a cable drive for our machine? It certainly would be cheap. I figure 8 turns on a ~1" diameter with get us the 20+ " of travel we need.

-Kwok

 

Hi Kwok,
I do think the cable drive is interesting, and you can probably get away without the anti-creep method I showed, but it would require a spring tensioner (not a big deal), and a linear encoder or some other trick to make a good rotary to linear conversion at the motor encoder.
Speaking of linear encoders, there are some very inexpensive cable reel encoders for that kind of travel, but I know little of how the gecko's or mach 3 will deal with the inevitable synch errors between linear feedback and steps. Is this an open loop stepper?

All that being said, I've got a new one I think would be perfect for your machine. I just talked today to one of my vendors about some prototypes of the critical element, which will ultimately be dirt cheap, but do not exist. It's truly simple as it gets, and I can't wait to show it. I do believe I will go straight into public domain with the idea, but would like to prove the basic assumption as soon as I can.

 

Now that I think on it, I think it possible to let the cable creep over 8 turns be a simple linear scale correction in your software.

 

Mike, yea all our motors are all open loop steppers. Our machine is just not that sophisticated. Anyways, I was just curious. But like most here, will be drawn to a belt upgrade in the near future. I don't know if you are familiar with the history of the PP. But basically Mark and Trish, our host & designers, have introduced version 2 with belt drive. And it's proven to have huge speed increases. So a plane or whatever that would have taken an hour, can probably be cut in under 10 minutes.

-Kwok

 

I will gladly do that when I have something solid to present.