Author Topic: Servo Driven 4th Axis for CNC Mill (Read 98445 times)

awemawson · « **on:** January 08, 2014, 01:37:08 PM »

Its been an ambition to have a 4th axis on my CNC mill for several years, and to this end I've been collecting suitable bits. Frequently people adapt manual rotary tables, in fact I did one myself many years ago using a stepper motor and a 4" table, but to be properly integrated into my Beaver Partsmaster it needs to be servo driven.

The other issue is that standard rotary tables end up on the left side of the mill table so that the motor sticks out forwards - as I have an automatic 16 station tool carousel on that side my 4th axis needs to be a mirror image

awemawson · « **Reply #1 on:** January 08, 2014, 01:46:56 PM »

I thought I'd cracked the problem when I 'caught' a Nikken 4th axis positioner on ebay a few years back - but when it arrived it proved to be stepper driven not servo although it is 'handed' correctly

Never mind, prompted by the success of Jim McCracken on the CNCZone forum getting his working I've decided to put a servo motor on the table of the Nikken

So this is what was there originally:

awemawson · « **Reply #2 on:** January 08, 2014, 01:52:29 PM »

And here are the bits that together hopefully will drive the table in future:

a/ SEM HT 30H4 -44 with ROD450B 125 count sine wave encoder and integral tachometer (more about the encoder later!)

b/ Bosch Servo drive card rescued from a Bridgeport Interact

c/ Bosch special choke for servo card from same source

awemawson · « **Reply #3 on:** January 08, 2014, 01:57:28 PM »

So far I have removed the stepper motor and started dismantling the rotary table. The table has a pneumatically operated locking brake, and it had jammed in the on state. Fighting my way into the mechanism proved to be anything but simple, and has required making some special tooling, but I'll leave you to ponder on this shot of the bit that had me stumped for a couple of days, while I go and have my supper

awemawson · « **Reply #4 on:** January 08, 2014, 02:45:19 PM »

Ok so what's the big deal?

Well firstly from the top there is little indication what you are dealing with, and there seems to be two concentric threads and a retainer with two forms of locking. The retainer has been slit so that a hex headed bolt pinches the slit gripping the thread, and also there is a grub screw bearing on an almost detached 'crescent' of thread to press it into a locked state. They really didn't want that to come adrift!

The grub screw was in a most inaccessible place and didn't seem to have a standard hex socket - I ended up grinding the end off a 3 mm one so it was short enough, then gently reducing its a/f size a few thou with a diamond file.

The retainer is bored for a pin wrench - so first of al I applied my Facom adjustable pin wrench. The retainer won and I sheared the pin on the wrench - lucky they screw in !

awemawson · « **Reply #5 on:** January 08, 2014, 02:47:31 PM »

As the retainer is recessed the Facom was not really getting where it was needed. So then I realised I'd have to make something to fit - I made a pin wrench out of a bar end and counter bored it 3/8 for a tommy bar I had lying arround

awemawson · « **Reply #6 on:** January 08, 2014, 02:52:04 PM »

You guessed it - the tommy bar ended up like a banana

Time to get serious - if only that pin wrench I'd made could go on my air impact wrench. Well it can if you cut a 1/2" square in the end. Lucky I fixed that EDM machine

So let's make an electrode:

Brass_Machine · « **Reply #7 on:** January 08, 2014, 02:56:19 PM »

Quote from: awemawson on January 08, 2014, 02:52:04 PM

You guessed it - the tommy bar ended up like a banana

...

Ouch!

awemawson · « **Reply #8 on:** January 08, 2014, 02:57:16 PM »

You'll notice that the lump of copper I used had had an odd multi-pointed star shape in the other end, and a little broke through - never mind it won't affect anything

So here is the result:

awemawson · « **Reply #9 on:** January 08, 2014, 03:01:09 PM »

Now I don't like using masses of force to dismantle things - things tend to break

So I set my air impact wrench to it's lowest setting and had at it. Nothing, no movement. Well I went through the four power levels, and at last on the highest setting off it screwed - relief

Now I could see what was inside, and here it is:

awemawson · « **Reply #10 on:** January 08, 2014, 03:05:35 PM »

As you can see, there is a cast 'shoe' on the shaft that is pulled up against the wormwheel to lock it. The shaft has an acme thread onto which a gear threads, which engages with a rack which is pulled up by an air cylinder triggered by a solenoid valve - simple really

awemawson · « **Reply #11 on:** January 08, 2014, 03:08:15 PM »

OK now I'm in so far at least hopefully the rest of the mechanical side should be reasonably straight forward. Currently I can only rock the worm and wheel about two degrees - I suspect that a bit of the broken brake pad has got between them - so they'll have to come out for cleaning.

Watch this space . . . . . . . .

Andrew

Pete W. · « **Reply #12 on:** January 08, 2014, 04:40:28 PM »

Hi there, Andrew,

There you are, you see. That wasn't so difficult, was it?!?! (To share it with us, I mean.)

Seriously, congratulations on your tenacity and ingenuity - I'm pleased you got a result and I hope the way forward is complex enough to be interesting but not so awkward it's frustrating.

awemawson · « **Reply #13 on:** January 09, 2014, 06:24:44 AM »

Thanks Pete,

So quite a bit of progress this morning. Objective was to free the worm 'capsule' so that it could be swung out of engagement with the worm wheel.

First I attacked the drive end, removing the gear that engages with the servo motor, then a cunning micro backlash adjuster, and loosened the various locking devices:

awemawson · « **Reply #14 on:** January 09, 2014, 06:30:14 AM »

Then I attacked the other end of the worm shaft. Turns out that the cover has an oil sight glass I'd not spotted before. It was all pretty gungy in there - obviously water based coolant had got in. When I managed to locate all the various locking devices, (and realised that some of the grub screws were two to a hole !) the worm capsule rotated nicely away from the worm wheel.

At this stage the worm would turn freely but the table was still only turning a few degrees, so obviously not foreign matter between the worm and wheel.

awemawson · « **Reply #15 on:** January 09, 2014, 06:34:26 AM »

This allowed me to gently tap the table with worm attached out of the casting to reveal the very nice end thrust race and top thrust race.

The construction reminds me of a swivelling tank turret

The side thrust race deposited all its rollers on the bench - all eighty of them - but all were located and put in a tote bin for cleaning.

awemawson · « **Reply #16 on:** January 09, 2014, 06:41:58 AM »

Those rollers look pretty grotty, but an initial inspection shows that they measure spot on 4 mm, so it is more discolouration than wear.

I'd replace them if easily available - anyone know a source of roller bearing rollers in various diameters?

So far I'm pretty pleased with progress - I need to get it re-assembled then work out just how I'm going to mount the servo motor, as its fixings are bolts into its end bearing plate, whereas the original stepper had a mounting flange with externally accessible bolts. It needs proper planning as the mesh of the gears is set by moving the motor, and if the motor has to be fixed before the cover goes on, there is no way of adjusting it

I'd hoped that I could make a dummy flange, but the sizes don't allow that

. . something will occur to me hopefully

. . .watch this space,

Andrew

Meldonmech · « **Reply #17 on:** January 09, 2014, 09:14:14 AM »

Hi Andrew

Its looking good, a good clean and new oil. together with your new actuator, should last for ever.

Interesting project
Cheers David

lordedmond · « **Reply #18 on:** January 09, 2014, 09:35:30 AM »

Andrew

Is this what you are looking for
http://simplybearings.co.uk/shop/index.php?cPath=c23_5043

Stuart

awemawson · « **Reply #19 on:** January 09, 2014, 10:08:33 AM »

Stuart, yes many thanks, Google got me there as well, but their 4 mm rollers are either too short at 4mm or too long at 11.8 so I decided to use the originals.

So onwards and upwards! I removed the vertical and horizontal thrust races. The Horizontal one has 80 solid 4 mm rollers all of which were ok measurement wise so they went back as is.

The vertical one on which the weight of the table and work is taken comprises 36 hollow rollers of 8 mm diameter, one of which was broken

However I now think that this was why the table was jamming. I was surprised to find they not only were hollow, but had really quite thin walls. I've come across hollow rollers in bearings before in for instance the Gamet headstock bearings of a Colchester Student, but they were quite thick walled. So what to do?

Options - a/ Leave a roller out b/ Try and source one c/ Make one. I opted for c/ on the grounds that although it probably would be perfectly ok leaving one out, it's not very professional, and I was unlikely to be able to find just the right roller. I turned a piece of silver steel to just oversize in the hard state using carbide tooling, then lapped to size. I decided to leave it solid as no doubt the flexibility of the steel would be totally different to the original.

So here we have the broken roller with it's replacement, and the bearing back in situ:

awemawson · « **Reply #20 on:** January 09, 2014, 10:12:17 AM »

Look carefully at the picture and you can see the new roller at 'ten o'clock'.

So on to the side thrust bearing. This has 80 little rollers all of which need to stay in their places as the table / worm wheel assembly is re-fitted. The usual trick is to smother them in grease, and this I did, but needless to say it was 'fun' as they kept sticking to my fingers and coming off again

However we got there :

awemawson · « **Reply #21 on:** January 09, 2014, 10:15:29 AM »

And not only that, but I got the table on and it's hold down keeper fixed.

AND The table is now free to turn

Hoo- flipping - ray !!!

So I am now sure that the jamming was that broken roller.

. . more to come later . . . watch this space

Andrew

RussellT · « **Reply #22 on:** January 09, 2014, 11:14:36 AM »

Hi Andrew

Well done. It's coming along nicely.

The hollow rollers are unusual and I wonder whether there's a reason. I wonder if it's to prevent shock loads from denting the bearing surfaces by making them a bit springy. It seems unlikely as I'd expect them to be too hard to spring much.

Russell

dsquire · « **Reply #23 on:** January 09, 2014, 01:54:31 PM »

Andrew

Well done. I love to follow along as you take it apart so you can figure out why it doesn't work as it should. As you have found out some times it can be a very small piece that has caused all the problems. You keep fixing them and I'll keep watching and cheering for you.

Cheers

Don

awemawson · « **Reply #24 on:** January 09, 2014, 02:48:05 PM »

Thanks Russell and Don,

So a bit more progress: The table is back on and turning so I've started re-assembling the end bearing assemblies for the worm shaft. First the drive end:

News:

Author Topic: Servo Driven 4th Axis for CNC Mill (Read 98445 times)

lordedmond