[acannell]

I'd like to make a 4th axis for my DM2900 mill.

They (dyna mechtronics) made one for it back in the day, and it turns out they used a HV-6 rotary table and just slapped a timing belt, encoder, and stepper motor onto it.

If it was built to the same quality as the mill, then its probably plenty good enough for me.

So if I get my hands on a super-common HV-6 rotab, maybe I can convert it myself and actually end up with a nice 4th axis.

I have a bunch of encoders and nicely sized stepper motors too. So it could be a great project.

So its seeming pretty good.

'cept that its hard to imagine that a pretty cheapo HV-6 plus a timing belt ended up not having a ton of backlash..what am I missing?

So I read a little about harmonic drives..and then I realized I had three brand new ones laying around!!!

Or so I thought..turns out they are just very high quality planetary reducing gearheads.

But reading the specs, they might be quite good for a rotab to 4th axis conversion like this (a 6" rotary table).

And, I only need 1 of them for this. I'll sell the other ones so other peeps can enjoy..

But I'm not sure if they are up to the task yet..here are the specs:

I have TWO of these:
mfg: Parker
p/n: PV40FB-050
ratio: 50:1
nominal input rpm: 4500
maximum input rpm: 8000
nominal output torque: 59 in/lbs
standard backlash: 18 arc-min

and I have ONE of these:
mfg: Parker
p/n: PV40FB-010
ratio: 10:1
nominal input rpm: 4500
maximum input rpm: 8000
nominal output torque: 31 in/lbs
standard backlash: 15 arc-min

How is that backlash looking? Is that really good or really not so great?

If these are good for a direct drive 4th axis, then what are people doing these days as far as that? I.e. just making a spindle headstock essentially from scratch?

If we assume a direct drive..i.e. the stepper connects to this and then this outputs directly to a headstock shaft..no belts or gears..then how much torque we talking bout here..

Lets say a NEMA 23 stepper motor running at 1500 rpm, putting out 3.75 in/lbs of torque

So with the divide by 50, that becomes 30rpm for the 4th axis speed, and 188 in/lbs of torque. Seems kinda slow..not sure if thats enough torque or not. Lets see if the rotary table is 6" then 30lbs at the periphery of the table would overcome that. Im pretty sure cutting forces could exceed that pretty easily right?

Are there any general guidelines for 4th axis torque requirements?

Using a 500 line encoder would give 0.0036 degrees resolution.

Is that good enough?

[acannell]

Assuming the backlash for these of 15 arc minutes is specified at the output shaft..meaning you can bounce the output shaft back and forth 15 arc minutes..that equates to about 10 thou at the edge of a 6" circle..not so great.

So what about using one of these on the output of whatever motor to increase the torque, and then doing a typical timing belt drive from there to the actual headstock? If the pulley ratio is say 4:1 then the backlash would get divided by 4 and the torque would get further increased...now its a backlash of 3 thou...not great but better

[louieatienza]

These are actually planetary gears, not harmonic drives. Harmonic drives are capable of less than 1 arc-min, and higher gear ratios. These however should be just fine for your needs, unless you're making aerospace parts...

I should also point out that these have pretty robust output bearings, and you probably could just mount an adapter plate for your chuck directly to the output shaft without much worries.