[Mike Everman]

Hi. You've got it correct. We generally scale up the motor to a maximum of a NEMA34-4 stack to get a peak force of 150 lbs linear. Keeping it direct driven preserves our near-zero backlash.

For use of smaller motors, a gear head is required, which as you say, adds some backlash, but not an appreciable amount of "springiness".

Our standard belting for ServoBelt is T5mm pitch, 32mm wide, so our recommended minimum pulley size is 15 teeth, for a linear travel of 75mm per motor rev.

Gearheads are readily available with 6 arc-min backlash, which at our pinion size translates into only .0008" on the linear. Not too shabby. You can get up to 10:1 in a single pass planetary to get that low backlash. Higher ratios/muliple passes (usually) makes for higher backlash. If you design your own, then smaller belt pitch is allowable, which on a 3mm HTD or T series can get you a smaller pinion.

Hope this helps!

[ger21]

I've been playing with a design using a 3:1 belt direction driving a 13 tooth T5 pinion with a 570oz Nema 23 stepper. Forget the results right now, but thought I'd get around 100lbs of force, even at over 500ipm speeds.

[Mike Everman]

Cool, Jerry. I'd love to see some pics!

[ger21]

http://www.cnczone.com/forums/showpo...&postcount=171

Still a few months away from building a mock up.

Edit. Added another image.

[Mike Everman]

Neeto. I LOVE skate bearings.

[posix]

Mike, what exactly is the purpose of two rows of tensioners? Wouldn't one row on each side suffice?

[Mike Everman]

what's not shown in the illustration is the static belt "rack" that this all pushes on. the two rows each side doubles the number of teeth being used in shear, to make the linear force.

[posix]

no, I was fully aware of the "rack" belt, just didn't know why two rows of "tensioners" instead of just one but that makes sense, more teeth, stiffer. maybe the question should've been why two different SIZES of tensioners rather than two rows. another question if I may, you mentioned earlier on that the top "belt" belt should be flapping about and the lower "rack" belt should be fixed (with double sided sticky tape) onto the surface. correct? would just sandwiching the two belts together and clamping them onto the surface have negative effects?

[Mike Everman]

The rollers have little to do with tension, rather they are there to keep the upper belt engaged with the lower, as close to the pinion as we can manage. They are different diameters because the nearest ones to the pinion force the belting to back bend, and there is a minimum diameter for long life, just as there is on the pinion itself.
As to clamping the belts, I'm not sure I understand which way you mean that...

[Hi-Vac]

Mike,
I love your idea / concept. I read thru this thread and I am VERY impressed with your design. I am assuming that you are not selling the belting / pulleys yet to us DIY’ers and I looked at some of the links posted to the other manufacturing companies but it appears that although it works, is it true it doesn’t work as good as your set-up? (backlash?)

I couldn’t help but to start thinking about a system that might give you more raw thrust (milling?) would be to machine the lower track out of aluminum, let’s say in 2’ increments, that would bolt down similar to a rack and pinion set-up. This track could then be machined to fit an industry standard (XH, L, XL, etc….) belt that is readily available.

If this (by any chance) is available (even on a trial basis!) for sale, please PM me!

I’d love to be your guinea pig!

John

[NEATman]

John-
If you want to make a version of this using a metal rack, you could use multiple sections of belt clamp - such as these extrusions:
http://www.brecoflex.com/index.php?C...ID=3&SMENID=14
However, I think that the stiffness is more defined by the section of belt that is not engaged with the rack than what the rack is made from.
Plus, I have seen this device in person, and it is very cool watching it zip at insane speeds back and forth.

Keith

[Hi-Vac]

I agree in regards to the section of the belt that is not engaged is the weakest link. I am also assuming (and I may be totally off base!), that the belt itself being made of polyurethane could be the weak link as well. I think this belt is selected because the belt “times” with itself when meshed and most industry standard belts do not. If the aluminum base was machined to a stronger belt pattern that for example would not mate to itself, but could mate properly to the machined aluminum (stationary) belt, you could get the same effect but with the benefit of the stronger belt and the stretch would be less.

Just a thought……….

[rowbare]

From previous posts by Mike there are at least a couple of factors that come into play for the belt choice. One as you mentioned is that the belt profile can mesh with itself, another is that the belt can bend backwards i.e. many belts can't wrap around a roller teeth out without self-destructing.

I am not sure that making the fixed rack out of aluminum would buy you all that much compared to the expense. There is only a very small part of the rack in play and handling load at any given time probably 3-5 teeth either side of the point where the belt comes off the rack to form the loop. Note also that the only part of the belt that is subject to stretch at any given time is that part that isn't engaged with the rack.

bob

[ger21]

If the aluminum base was machined to a stronger belt pattern that for example would not mate to itself, but could mate properly to the machined aluminum (stationary) belt, you could get the same effect but with the benefit of the stronger belt and the stretch would be less.

Just a thought……….

The stretch would be the same, because it can only stretch where it goes over the pinion. There's no tension on the majority of the belt.

[Mike Everman]

Hi Guys, good comments all! You're making it so I don't have to respond at all, ha ha.
Agreed that the stiffness of the static belt is not the primary issue. Deflections there are second order for sure.
An issue not covered before, and has come out in repeatability/velocity error testing recently is due to the loading/unloading of the downward legs of dynamic belting, even though they are preloaded by belt tensioning.
It's not just the stretching of that short bits of belt, but that the unloading leg exhibits a little curvature until the tensioned leg achieves full commanded acceleration (tension).
It's not much, but visible. I'll take and post up a high speed video to illustrate.

And, hi Keith!

[billblack]

VERY cool system.
VERY long time since you first mentioned DIY sales.
I've been reading this thread waiting for pricing information.
I've got a new prototype machine to build NOW.
So....is there any DIY/small quantity pricing?

[Mike Everman]

VERY cool system.
VERY long time since you first mentioned DIY sales.
I've been reading this thread waiting for pricing information.
I've got a new prototype machine to build NOW.
So....is there any DIY/small quantity pricing?

Sorry, still not set up for it.

[billblack]

OK, Mike.
Thanks for the honest answer.
I was hoping to give your system a try, but I guess we'll go a more conventional route on this new build.
Good luck to you, sir.

[Mike Everman]

My ServoBelt Router is coming along! Base frame should be done soon.

[asuratman]

Hello Mike,
Do you plan to sell linear servo belt kit for router? How good this kind rubber for plasma applications due to heat and a lot of debris? Thanks.