[SCzEngrgGroup]

I watched with amusement the other thread discussing toolchangers, and the pros and cons of Geneva mechanism vs stepper drive. It seems some people consider the Geneva to be mechanically complex and difficult to fabricate, and to have all kinds of shortcomings. No doubt some of the doubter have never built one, and perhaps never took the time to really understand how it works.

As I am in the process of building a carousel-style ATC for my knee mill, I chose to go with a Geneva, and today I spent about 90 minutes fabricating a prototype out of scrap aluminum. I'm happy to report it works absolutely perfectly! Even running bone dry, it is silky smooth and quiet in operation, positions perfectly at each tool station, and locks in place with only just barely perceptible angular backlash - certainly nowhere near enough to cause any kind of problem whatsoever.

I expect it would have taken just as long, and cost a lot more, to do the same with a stepper and belt drive. To drive it, I will be using an off-the-shelf $34 gearmotor from McMaster-Carr, which will give rotate one tool position in barely a second, making the worst-case rotation time for a 12-tool changer only six seconds. Overall, I expect worst-case tool change time should be about 15 seconds start to finish.

Regards,
Ray L.

[hoss2006]

I'm glad you were amused but it still won't be as fast as a stepper system period, which was the point.
Who said a geneva has all sorts of short comings? Dave D. uses one which shows it's a perfectly viable mechanism for this task as do some real VMC tool changers.
Hoss

[691175002]

It seems some people consider the Geneva to be mechanically complex and difficult to fabricate, and to have all kinds of shortcomings. No doubt some of the doubter have never built one, and perhaps never took the time to really understand how it works.

It was never a question of whether it would work, there was video and diagrams all over the internet. The question was what advantages does this mechanism provide?

Cost isn't really a valid criticism since torque is not a major requirement of a carousel. Any 10-20$ stepper will do and you can get a driver for 12$:
Pololu - A4988 Stepper Motor Driver Carrier
NEMA 17 Stepper Motor
I'm not convinced a stepper would have taken any extra effort either. Sliding a timing pulley onto a shaft is not any harder than cutting a geneva mechanism.


I just wouldn't build a mechanism that is forced to make an unnecessary stop at each tool location.

This is a fairly trivial argument given that both solutions clearly work and are of similar cost and difficulty, but I felt like it was worth explaining my position.

[Geof]

....torque is not a major requirement of a carousel....

It could be if you want fast starting and stopping on a large disc with a lot of mass near the periphery.

[rwskinner]

It could be if you want fast starting and stopping on a large disc with a lot of mass near the periphery.

So then, a stepper driving a Geneva for the best of both worlds. Speed and locking.

Believe me when I say the locking and accuracy is not an issue.

Richard

[SCzEngrgGroup]

So then, a stepper driving a Geneva for the best of both worlds. Speed and locking.

Believe me when I say the locking and accuracy is not an issue.

Richard

Nor is speed, except for bragging rights.....

[hoss2006]

You darn bragger Richard, how dare you.
Hoss

[wizard]

It was never a question of whether it would work, there was video and diagrams all over the internet. The question was what advantages does this mechanism provide?

1. Mechanical locking has already been suggested as a big feature as the unit should "lock" in place without power. This can also be a disadvantage if you want to move the carousel around un powered.
2. Electrical simplicity is another big advantage as you can get by with one signal to trigger the index and one "home" input. You are generally better off with a set of switches implemented to encode each position though.
3. In my estimation a mechanical indexer is more rugged and can better handle unexpected loads. Through a really heavy tool on a stepper driven dial and you may find yourself retuning the drive. Now obviously excess load can be a problem for a mechanical indexer too, but I wold suggest that it is easier to get such a mechanism to work reliable with a variety of changing loads.

Cost isn't really a valid criticism since torque is not a major requirement of a carousel. Any 10-20$ stepper will do and you can get a driver for 12$:
Pololu - A4988 Stepper Motor Driver Carrier

torque is a huge factor. The big problem with tool carousels is the changing nature of the load. Going with too small of a stepper could be come a reliability issues.

NEMA 17 Stepper Motor
I'm not convinced a stepper would have taken any extra effort either. Sliding a timing pulley onto a shaft is not any harder than cutting a geneva mechanism.

A stepper system is certainly workable. The development effort is bigger though. I should say the electrical development effort. It is really a case of where you talents lay.

I just wouldn't build a mechanism that is forced to make an unnecessary stop at each tool location.

? I don't follow. You don't have to stop the motor, given a bit of creativity and knowledge of where you want to go you can just drive the indexer for X number of moves. Depending upon the stepper implementation you may be doin the same thing. At work we have machines with Camco type indexers and stepper indexing. The PLC control logic Is very similar with the stepper "indexer/driver" basically acting as an electrical substitute for a mechanical solution.

This is a fairly trivial argument given that both solutions clearly work and are of similar cost and difficulty, but I felt like it was worth explaining my position.

It is certainly good to see multiple opinions here. My experience is that mechanical indexing is the way to go for high volume & reliable operation. Stepper indexing can be a good solution here as long as the implementation isn't under engineered.

[CNC-Joe]

Ray - I hope that you post your design of your tool changer here - so that the rest of us can benefit from your build.

Your Geneva mechanism looks good.

[SCzEngrgGroup]

Ray - I hope that you post your design of your tool changer here - so that the rest of us can benefit from your build.

Your Geneva mechanism looks good.

Yes, I will. It will be really simple to build, and should be trivial to adapt to almost any machine by changing nothing more than the mounting arm. On a machine without a quill, it could even be simplified a bit.

Regards,
Ray L.

[sudip]

To under stand briefly the automatic tool changer and Geneva mechanism you may try this
Automatic Tool Changers of VMC 500 machine...How It Works ?

[fizzissist]

What I find continually amusing is the old... "What's better: A sports car or a truck?"

This discussion has those elements, and the end result is a sports-truck.

It always boils down to the specific application, and always the constraints of that specific application.... build time, money, power, speed, quality, size, etc....

The geneva is a wonderful, reliable, rugged mechanical mechanism. The stepper is a wonderful electronic mechanism.... They can be a wonderful solution together. In the end, they're both simply tools available for a solution, and one isn't "better" than the other. I know first hand, I've had both (though not integrated together).
Besides the occasional grease, I have NEVER EVER had to service or replace a geneva index. (with the exception of a bearing on an index pawl.. big deal)

I've had lots of steppers crash for a variety of reasons. I've never had a geneva crash on its own. ... And I still like 'em both.

.... Btw..... Ever notice that even the big boy's CNCs have mechanical locking (braking) on the steppers for things like the C-axis on a lathe, or the A and B-axis on a mill??

[johnedward]

I think the OP and others are misinterpreting a decision by others to use stepper driven carousels and not use a geneva mechanism as either a disdain towards the mechanism or an inability to understand how it functions. There is no evidence of either from any of the parties. These fellows seem to have built tool changers already and have the benefit of experience on their side. Implying they couldn't understand a concept is pretty egotistical. Maybe trying to find out actually what someone else's reasons for using a different approach than yours might teach you something. Or do you know it all already.

[uli12us]

An important part is, if you have a machine that needs more axes as you can drive, than you need a geneva mechanism, because you don't need any driver for that, just start or stop a whatever motor and count the steps back and forth of a Proximity switch. Another point is, the position of the chain or disk is always the same, what is important for the toolchanger. If a steppermotor runs against too much resistance, it loses steps and the position isn't correct. Much fun with toolchanging.

I had an Issue with the magazinedisk of my machine, sometimes it have a false position and I need months, to determine that the screw who connects the disk with the gear were not tight enough secured. If I had a geneva mechanism in it, the problem can't be anywhere in the motor/encoder.

[johnedward]

I think you need to bear in mind that the OP and these gentlemen aren't using real tooling like CAT40 we may be used to dealing with. They are using little Tormach tool holders for hobby machines so there isn't the need for as stout a system. I believe the pc controls they use can handle a minimum of 6 axis.

[uli12us]

OT: The 6axes are much often the maximum, more likely are 3 or 4 axes. And these are completly absorbed by XYZ and spindle. I have never seen more than 6axes by one of the simple controls. If I need more I must use a more or less professional control, which is capable to drive 20 or more axes.

[johnsattuk]

OT: The 6axes are much often the maximum, more likely are 3 or 4 axes. And these are completly absorbed by XYZ and spindle. I have never seen more than 6axes by one of the simple controls. If I need more I must use a more or less professional control, which is capable to drive 20 or more axes.

CNCUSB board will give you 9 axis

[uli12us]

Ok, but this board have only a very limited range of in and outputs and runs with USB, what I have read should not be very reliable. And the badest of all features is, theres no way to drive a +/- 10V Servo amplifier with it. I haven't seen, that have inputs for digital scales, so it will be good for a DIY project, but not suitable for a big machine with all these features to refurbish the control.

Okay the +/-10V can eventually substituted with a simple step/dir DC-Servodriver who makes the tension indirectly.

[johnedward]

uli12us, you don't seem to understand that the machines these fellows are putting tool changers on are very small, they are not VMC's. I looked at one of the programs that seems popular, mach 3 and it allows 6 axis plus spindle and several inputs/outputs so for their needs it's plenty.

[derek]

I might add that it's a discussion about Geneva mechanisms not controllers.