Anti-backlash rotary joint existance

[TomB]

I have been thinking about a design for a 2 rotary axis gimbaled table for use in a milling machine. So far it is still at the sketch stage, but it’s within the capability of my shop and skills. However, I’m worried about how cutting forces and drive backlash would interact as the cut crossed the axis of rotation. Currently I thinking about a worm and gear to drive the rotation so cutting forces really could not try to rotate the motor worm shaft, but worm shafts and gears typically have a lot of backlash.

Is there an alternative drive connection for rotary joints? For example, ball screws eliminate backlash for linear drives, is there a ball joint for rotary connections? Alternately, spring loaded dual nuts can eliminate/reduce linear backlash so is there an equivalent for rotary joints. So far all I can think of is having two worm drives on opposite sides of the gear, where both were driven at the same rpm but in opposite rotation and spring loaded in the same direction.

I’m quite serious about this project and just joined the group to gain insight into the CNC techniques required. While reading old posts I read the linear motion thread and it triggered the question. Couldn’t find an answer with a search and this seemed like the best thread for the question. As background on me, I practiced mechanical engineering 40 years ago and currently work as an applied mathematician. I have a metal/wood shop and spend lots of hobby time in it.

Tom

[RICHARD ZASTROW]

Tom, Is there room for direct drive motors such as the Danaher DDR cartridge motors? They can be attached directly to components like rotary tables etc. eliminating belts, gears, couplings etc.

You'll have to determine if there's enough speed & torque available and still fit in your "package".

Good luck, Dicx Z

[TomB]

I wrote the original post and have considered the suggestion to use a gear drive motor. I’m not sure I want that much mass on one side of the gimbals and I’m trying to keep the gimbal as open as possible. I know the parts I want to make but don’t want to specifically describe them. But milling out a rimmed oval soup bowel with drilled surface orthogonal holes requires most of the same cuts. Putting gear motors on the gimabal would put it in the way of some cuts.

I do need an anti-backlash mechanism and I think some sort of a ball screw would be more precise and lighter than dual worm gears. For the past month I’ve been sketching and it seems possible. Put a collar on a drive shaft that is hourglass shaped with a radius equal to the pitch diameter of a gear blank. Mill a helix-ball groove in the hourglass surface, and matching ball groove slots in a concave gear blank. Then put a housing to hold the gear in relation to the shaft, to contain balls on the non-gear sides of the hourglass collar and with a ball return hole. As the drive shaft rotates balls between the shaft and the gear would move toward or away from the drive motor and the gear would rotate.

However, what I described is nothing more that a worm and pinion that utilizes balls as the connection instead of gear teeth. To me it does not seem like a novel idea and therefore I would expect to find it commercially available. However, I have tried every variety of Google search that I’ve been able to imagine and so far have found nothing. Any ideas, pointers or criticism will be appreciated.

Thank you
Tom

[RICHARD ZASTROW]

Tom, FYI you might be interested in this concept: google "howimat" .

Dick Z

[TomB]

The link Dick Z offered is very interesting. I'm not sure I could build something as complicated as that integrated joint and motor. I was/am a mechanical engineer and may have learned a lot about digital chips, but I've always considered the field equations that define motor characteristics to be black magic. I was also surprised by their selection of rotary axis. Putting a rotary table on a gimbal is certainily easier than putting a gimbal on a gimbal, but can it accomplish the same things? Certainily any point on a convex surface can be rotated so that it tangent to the bottom of an endmill. Further the part can be moved and rotated so that the endmill will stay tangent to the suface with either approach. But are the required cutting paths equally easy to generate? I don't yet know.

Tom

[RICHARD ZASTROW]

True 5-axis SIMULTANEOUS contouring motion requires some very sophisticated and $$$$$ software. Something less like 3-axis contouring and other axis positioning is not as complex or expensive.

I went to the Detlev Hofmann "Howimat" website. I didn't find their normal page where they explain the ball driven rotary axis dividing heads. They function similar to a ballscrew in a rotating direction. Those things are amazing. They are fast, accurate, rigid and compact. They are not, however, inexpensive.

Fortunately, at the time the project I was working on was funded by a large German company. We used a German rotary table on a German milling machine with German CNC controls which made all the Germans involved very happy. Jawohl !!

PS, My family is from Germany

Dick Z

[wwendorf]

If you go to www.sdp-si.com, you will find some special anti-backlash worm gear and worm drive combos. I'm planning on purchasing a set, but they will set me back about $65.

Wade

[Clementvs]

Hi TomB,

I would recommand you a torque motor which can drive your axis directly, with permanent magnet you can limit the mass. The French company, names Alxion, makes nice product. To reduce weight, I would recommend the frameless version :

http://www.allytech.eu/index_fichiers/STKmotors.htm


Regarding the antibacklash point, I would suggest to google "splitted worm gear" which is a good alternative to dual lead system, may not be lighter (I don't know)

Good luck for your project

Clem

[NEATman]

TomB-
Take a read through this thread:
http://www.cnczone.com/forums/showthread.php?t=72261

Lots of info on anti-backlash rotary tables.

If you have a limited angle on one axis, you could try this approach:
http://www.centroidcnc.com/downloads...ng_Machine.pdf
It is like rotary table on a CNC Teeter-Totter with a linear axis on one end for porting engine heads.


Keith