Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: wil on January 12, 2015, 08:40:09 AM
-
Hi guys
All 6 axes in mach will be utilised but I also need to precisely control a worm screw to advance the pin tree 1.85 mm each time into a weld stage.
After welding, a cut off disc detaches the 2mm dia pin onto the welded component and preps the next weld face ready to advance.
I tried a timer which turns off the motor supply after the set amount of millisecs, but it never repeated well enough.
I guess maybe an electromechanical brake attached to a dc motor with gearbox ? and possibly have a sensor underneath the welded part to activate the brake on contact ?
any suggestions please.
-
I was reading your post thinking this is all above my head, but then looked closer at your drawing and wondered if you could use a Geneva Wheel to advance the pin tree.
Plenty of videos on the net showing how they work if you don't know what they are.
The motors stopping position isn't critical with a Geneva Wheel, it just simply needs to complete one turn to advance the pin tree and you could simply use Mach to activate a relay.
-
cheers for that. I hadn`t seen them before.
I`d still need to control the motor position tho as it could stop mid cycle.
There could be as much as a minute or 2 between weld cycles while the rest of the machine does its thing. sorry I should have mentioned that.
-
You could just use a stepper (or servo) that has an "Indexing" type of drive.
Program a 1 revolution motion profile in the drive, then just call it when needed with a pulse on one of your available output pins.
The drive would then send an "In position" signal back to Mach when the move is completed and the program would then resume.
The Parker OEM 10 is a relatively cheap indexer that I have used with good success in similar applications.
Regards,
Russ
-
cheers for that. I hadn`t seen them before.
I`d still need to control the motor position tho as it could stop mid cycle.
There could be as much as a minute or 2 between weld cycles while the rest of the machine does its thing. sorry I should have mentioned that.
All you need to control a Geneva is a microswitch driven by a cam on the motor shaft that is open when the motor is in the "stopped" position. Use a relay to short the contacts on the switch briefly, long enough to get the motor started and move the cam off the microswitch so the microswitch is closed and providing power to the motor, then open the relay. When the motor has rotated about one turn, the cam should open the microswitch, stopping the motor. That's the nice thing about a Geneva - it requires absolutely no precision control whatsoever to give very consistent, repeatable position.
Regards,
Ray L.
-
Russ, thanks a lot. That sounds exactly what,s needed, but i have had trouble finding a Parker OEM 10, if you have a link that will be appreciated.
I actually do like the Geneva plan from Ray, only thing i`m not getting is how to short the contacts breifly ? is that with a timer ?
Would it be possible to do it with a mach3 brain ? can you change a pin from high to low from a brain ?
The other thing with the Geneva thats getting me going, is i should be able to mount 4 cut off discs, driven off a central motor, and rotate them one at a time to do the cut off opp.
4 times the wheel life. those discs are only 0.030" thick.
I was considering going with Kflop for the 8 axis, but you guys have again showed me how knowledgeable you are and will find a solution to any issue.
Thanks again.
-
I'm sure Ray will get back to ya on ozymax's geneva idea, the timed output could easily be done with a simple macro.
I think the OEM 010 is way beyond obsolete. They had an OEM 650 later on, not sure if still avail or not.
Gecko is due to release an indexing stepper drive soon that would work very well .... any day now. You might inquire.
The Geneva is cool, but I'd prefer less mechanics. Just the smart motor, a belt with 2 pulleys.
Cheers,
Russ
-
The Geneva is cool, but I'd prefer less mechanics. Just the smart motor, a belt with 2 pulleys.
Cheers,
Russ
...... or, just directly coupled. Shaft to shaft.
Russ
-
in this clip https://www.youtube.com/watch?v=7VDG7C5R5XQ he has attached an led via microswitch
I could do similar to activate the micro spot welder
-
yikes
I just realized to rotate a 4 lobed plate one revolution, the motor drive has to rotate 4 times !
it would now also need a 4 to 1 gear train
-
yikes
I just realized to rotate a 4 lobed plate one revolution, the motor drive has to rotate 4 times !
it would now also need a 4 to 1 gear train
The way I see it, you would direct couple the lobe plate to the worm screw that you have. This worm screw needs to rotate once to advance the pin tree 1.85mm each time.
You would have the motors contact switch acting on one of the 4 lobs of the Geneva wheel. That way when the switch is triggered the motor simply keeps running until the lobe wheel completes one revolution. The motor will need to complete 4 revs and thus no need for a gear reduction.
Maybe this triggering could all be done outside Mach. For example, another switch could be placed elsewhere on the machine so after the cutoff wheel has done it's part it retracts and trips the switch. I don't know if that's feasable or not without seeing the machine. Just trying to think outside the box.
-
Russ, thanks a lot. That sounds exactly what,s needed, but i have had trouble finding a Parker OEM 10, if you have a link that will be appreciated.
I actually do like the Geneva plan from Ray, only thing i`m not getting is how to short the contacts breifly ? is that with a timer ?
Would it be possible to do it with a mach3 brain ? can you change a pin from high to low from a brain ?
The other thing with the Geneva thats getting me going, is i should be able to mount 4 cut off discs, driven off a central motor, and rotate them one at a time to do the cut off opp.
4 times the wheel life. those discs are only 0.030" thick.
I was considering going with Kflop for the 8 axis, but you guys have again showed me how knowledgeable you are and will find a solution to any issue.
Thanks again.
A Geneva can be easily controlled by a brain or a macro. The output only needs to be active long enough to ensure the Geneva motor has turned far enough that the microswitch comes off the cam and closes, and it will power the motor from there, until it again hits the cam, which will stop the motor, leaving the wheel correctly positioned, and locked in that position, and ready to start the next move. For something that only needs a small number of fixed positions (up to about 16 or so), a Geneva is FAR simpler, and FAR cheaper. than driving it with a stepper or servo. All you need is a $30 DC gearmotor, and perhaps a relay. No motor drivers, no software, no complexity whatsoever, and it's dead reliable, and very rugged.
Regards,
Ray L.
-
Gecko is due to release an indexing stepper drive soon that would work very well .... any day now. You might inquire.
How about today?
http://www.geckodrive.com/gm215-step-motor-motion-controller.html
-
Sunofagun !
Just talked to Maris Marcus last week about it. He said "soon", but also said that a year ago.
Thanks Gerry !
I've got 2 on the way.
Russ
-
"For something that only needs a small number of fixed positions (up to about 16 or so), a Geneva is FAR simpler........"
...could be true, but this is only 1 position. 1 complete revolution.
With the Geneva, you'd need the aforementioned gearmotor including a drive for it, the Geneva mechanism with whatever indexes you might choose (say 4) then a timing belt (or chain) geared 4:1 , mounting and machine work to set it up ........
I'd still prefer the indexing drive and stepper motor directly coupled to the feed worm shaft.
Software is free, VERY small single incremental move to program (infinitely adjustable for acc, dec and speed) just a few wires to hook up.
Would be much simpler imo.
Russ
-
If it's one position, doing one rev each time, then all you need is a motor and a cam-operated micro-switch....
Regards,
Ray L.
-
That's more like it.
If that would be precise enough (repeatable stop pos.) ... good plan.
Would still need a gearmotor though ... but AC would be OK, no need for a DC controller.
Regards
-
Don't make it more complex than it needs to be.
The Geneva Wheel doesn't need any more than 4 positions because the important point is that the worm screw must rotate once and park in an accurate, repeatable position. That's why I suggested the wheel.
If you hook it up as I suggested in my reply #10 you don't need a reduction drive of any sort.
If you use a DC motor, relay and micro switches to drive the wheel then you don't need software, programing or driver boards etc, you simply use a DC supply feed that I assume can be got from the computers supply rail.
Simply using a cam operated switch mounted on a gear motor won't give you a repeatable stopping position as the motor will always over run to some extent once the switch is tripped. The Geneva Wheel addresses this situation.
I guess I'm looking at the original post, which talks about all 6 axis ports being used in Mach3, so the 7th axis would need to be controlled mechanically.
-
Hi ozy,
As I see your example, the Geneva would need to index through 3 full cycles, then stop on the fourth (without 4:1 gearing) to make 1 full revolution.
Quite time consuming just to make 1 rev at the feed screw.
But would certainly work as you describe.
Am I seeing your example right ?
Thanks,
Russ
-
Hi ozy,
As I see your example, the Geneva would need to index through 3 full cycles, then stop on the fourth (without 4:1 gearing) to make 1 full revolution.
Quite time consuming just to make 1 rev at the feed screw.
But would certainly work as you describe.
Am I seeing your example right ?
Thanks,
Russ
Yeah, that's how I see it.
The speed of the Geneva Wheel is soley governed by the choice of the actuating motor. They can run at a very high speed if made accuratly and with bearings etc.
To eliminate the time factor, the wheel could be advancing whilst a cut off wheel or welder was retracting for example.
I remember as a kid, my Dad gave me a people counter from an old showground that was being pulled down. The counter consisted of several Geneva Wheels in a row to change the numbers of people through the gates. I used to hook the counter up to an electric drill to see how fast I could make the numbers incresase. Believe me when I say that they can operate at quite high speeds and survive. :D
It was my knoledge of this counter that made me suggest it in the first place.
-
Another thought about the speed needed to advance the screw using a 4x lobe wheel.
If you have one pin mounted on the motor, then the motor must do 4x revs to advance the wheel 1x revolution, but if you have 2x pins on the motor then the motor only does 2x revs to advance the wheel 1x revolution.
I found this video that might help explain what I'm talking about.
https://www.youtube.com/watch?v=cunj4O_03e4 (https://www.youtube.com/watch?v=cunj4O_03e4)
I can't see why you couldn't have 4x pins acting on the wheel and thus saving time.
-
I tried to model it and I dont think 4 pins is possible due to the scallop cut away of the boss on the drive. This cut away boss keeps the wheel in it`s stopped position.
I do love mechanical devices and the Geneva is brilliant, i`m sure we could get this work as described with switches etc,
and a simple brain i wrote seemed to work ok (keyboard simuation mode)
Russ
The Gecko unit looks amazing, it would seem a hell of a waste to have most of it`s potential never used on my rig.
But have a look at this little beauty i found https://www.youtube.com/watch?v=DYG8SlrwGHM
at around £30 looks like just the job. Direct drive - one rotation.
-
That's a handy little bugger wil, nice find !
Thanks for posting,
Russ
-
My only concern with using a device like that is the slight movement you get on powering up/down a stepper.
It wont affect my welding stage use for it as i could put a spring behind the weld electrode to take up any errors,
There is another axis on my machine i`m tempted to try use the same device, all it does it rotate from zero to 180 degrees
and back to zero again on command, but it must be in the correct place on machine start up. easy enough with a homing sequence of a mach3 axis
but i`m not sure how to acheive that with the stepper movement driver device.
any thoughts ?