[Erick Davidson]

Greetings All,
I was inspired by this website this past May to build rather than buy my CNC plasma table. I went about gathering all the parts through surplus sites and the ubiquitous eBAy. I spent a good couple of months researching various design approaches and drew up my complete table in Solidworks to share on the site here as well as document the buildup. In the beginning I fell into the "more is always better" approach and over engineered a good deal of it based on the choices I had for components. I spent a bit more many times as I figured if 320 oz. motors are good 1300 oz. motors would be great. I wanted the performance of servos and the quiet backlash free advantages of belt drive. So I worked toward those ends.

Which brings me to my quandry. For the X and Y I have 3 1300 oz. servos rated at 3000 rpm at 90v. I have a complete robot module for the Z. I have geckos to drive them and intended to use a 75v power supply, which should give me 2500 rpm, correct? ( please correct me if I am wrong about any of this!?) Well, only after buying new brecoflex belt and pulleys (one of the costliest parts of the build) did I start to do the math for reduction to give me 200ipm to 400 ipm at 2500rpm with the 4 inch diameter pulleys driving the X and Y. With 12.567 inches travel per revolution, my intial idea that I only needed 3:1 reduction was a joke, a holdover from discussions of using rack and pinion. I only need 32 rpm to achieve 400 ipm! reducing 2500 rpm to 32 rpm requires 78:1. Hitting the internet for something in this range I only encountered a few 100:1 gearheads (I felt 314ipm a decent compromise) and a few 50:1, but almost nothing in my Nema 34 size. I finally settled on a couple of used 100:1 Bayside right angle gearheads I will have to adapt slightly. As well as a 50:1 Bayside that is freaking huge, I didn't pay much for it, so I probably wont use it as it would add 50 lbs to my gantry, which looks like it will weigh between 150 and 170 as it is, due mostly to the weight of the capistan robot module.

And now my question, I had purchased enough servos and geckos to drive both sides of the gantry, but with only two reducers in hand should I give up and use one for both the X and the Y and just run a jackshaft? I believe I have enough torque available, and that would leave me a spare gecko and servo for the future. I liked the idea of squaring the gantry with each home position, but I guess if both sides can be adjusted with the belt positioner / tensioner that wouldn't be an issue. I was also worried that the cumulative error that even the minimal backlash of the two gearheads might rack the gantry.

Any advice would be appreciated, as I am more mechanical in nature and this is all new to me. Thanks in advance!

Cheers, Erick Davidson
Zweihammer Armoury

[Torchhead]

Nothing like overkill (:-). If you can devise a 10:1 reduction using a jackshaft double set of toothed belts and look at smaller diameter pulleys. (4 inch????).

You have enough torque to pull Superman off the toilet! Nothing says you have to run the motors at 2500 RPM. In fact if you design your cutting speeds for half of that (1250) then it will be more realistic. If you end up with 800 IPM rapids then so be it. You can always tune the max velocity in the software.

Lots of machines (including commercial ones) use the jackshaft approach. Unless you can hold NASA tolerances on the table it will tolerate a small amount of racking.

I like dual drive for several reasons but either approach works.

[Erick Davidson]

Thank you, Torchhead, for your reply. You are most active on this board and if it weren't for you there would be far fewer answers. Hopefully when I gain experience I will have more to contribute.
Am I correct in assuming my 3000 rpm @ 90vdc will be 1500 at 45vdc, all else being equal?
Cheers, Erick Davidson

[Torchhead]

The speed of a DC motor (servo) is directly porportional to the applied voltage. Not absolutely linear in the below 10% range but close enough.

[lgalla]

How much were the belts?I was under the impression they are inexpencive.

[Erick Davidson]

Igalla, I am building a oversize 48 x 120 table and used the same 50mm wide belt on both the X and Y as well as 6 each of the same 4" aluminum pulleys and belt clamp plates, ended up around 1300 dollars. I like common parts and assemblies throughout if I can. Steel belted urethane belts, great stuff, never plan on replacing them. Like I said earlier, overkill. I could have saved about 800 dollars had I found this website, http://www.youngssurplus.com/tbelts.htm , before I ordered from brecoflex. They have the gates in 32 mm wide for about a quarter of the cost of new and can provide long lengths. I found their website about two weeks after ordering and after two months of searching.
Cheers, Erick Davidson

[lgalla]

Eric orTorchead,
I have seen other members using belt drive and had the same problems of requiring huge reductions.Is this an application better suited to steppers?
ie low RPM high torque.
Thanks Larry

[Torchhead]

Belt drive is similar to Rack and pinion in the ratios. The diameter of the pully engaging the belt determines the travel per revolution. While the reductions don't need to be huge running servos with high RPM ratings (eg 3000 RPM) may not provide practical reduction ratios. Steppers are more ideally suited for low RPM operation since they become torque limited with RPM.

Belts offer some advantages over rack and pinion. Because the belt is flexible, methods of engaging the teeth evenly and completely using idler pulleys and or partial wraps of the drive pulley allow for more teeth to be engaged. It can also mean the ability to use a finer pitch and smaller pulleys. The belt will run quieter and require no lubrication.

Just as in rack and pinion I don't think direct drive, even with steppers, is the best design. You just give up a lot of torque for the simplicity of the design.

I have long thought about my next bigger table and decided that a secured belt approach with the drive motor mounted on the gantry and the belt looped over the drive pulley (like a 35mm film projector feed mechanism) and tensioning idlers will be my design.

That's the fun part of this hobby. There are dozens of ways to approach a design and all of them will work. Some better than others (:-)