Author Topic: Servo Driven 4th Axis for CNC Mill (Read 82502 times)

awemawson · « **Reply #75 on:** January 23, 2014, 07:23:14 AM »

So I've finished the CAD for the cover, drilled an access hole for the encoder pulley Allen key, all I need now is to work out how to hold the cover when machining - the perennial problem

awemawson · « **Reply #76 on:** January 23, 2014, 11:57:36 AM »

So I cracked on this afternoon and got the belt drive cover made.

First I cut the inner shape so it sits nicely with a lip

awemawson · « **Reply #77 on:** January 23, 2014, 11:59:41 AM »

Then I reversed it and sliced 10 mm off it's thickness using the CNC mill 'sort of manually'

Then I screwed it down to a scrap chunk of aluminium for profiling:

awemawson · « **Reply #78 on:** January 23, 2014, 12:01:23 PM »

And when that was done clamped it to the main body, and transferred the holes though with a 4 mm transfer punch and drilled and tapped them

awemawson · « **Reply #79 on:** January 23, 2014, 12:04:08 PM »

Then just for devilment I thought that I'd reverse the cover to see how symmetrical it is. It was drawn symmetrical, but I've not allowed any tolerance on the screw holes. 4 mm screws in 4 mm holes and I measured the screws as 3.93 mm.

I was quite amazed that it reversed with no problems at all

awemawson · « **Reply #80 on:** January 23, 2014, 12:05:40 PM »

OK enough mucking about, get on with it

So hopefully this is the last 'trial assembly'

awemawson · « **Reply #81 on:** January 23, 2014, 12:12:46 PM »

In all I'm quite pleased with it. If I were doing it again I would probably have not stuck so rigorously to the diameter of the servo motor, and reduced the curve of that end to the size of the stock to avoid those two flats. I'll probably blend them on the belt sander before the final 'put together'

Also I must try and get my hands on some corner rounding cutters - the cover would look far better with a radius on it's outer edge.

Making the four elements of the mounting using CNC have made me much more confident in going from a CAD drawing to something tangible, and I've also gained confidence in the use of the Heidenhain probe, allowing me to break down work, and re-set it with some confidence that it's accurately placed. The 'reversal exercise' is also useful in that there are occasions when I'll want to reverse a symmetrical part to machine it, and the 'reference point' will thus move.

Andrew

Pete. · « **Reply #82 on:** January 23, 2014, 01:50:39 PM »

Nice looking job Andrew!

You must have a whole barn full of 'projects to be got round to - you seem to own 'one of everything but it needs fixing'

awemawson · « **Reply #83 on:** January 23, 2014, 02:11:45 PM »

Quote from: Pete. on January 23, 2014, 01:50:39 PM

Nice looking job Andrew!

You must have a whole barn full of 'projects to be got round to - you seem to own 'one of everything but it needs fixing'

Thanks Pete for the kind words.

Projects : yes well I've been saving them up

When we moved here not much got done engineering wise as we had rather a lot of sorting of the domestic buildings, and actually building the workshop, but now those things are on an even keel I'm going through them ticking off the list. However it's a long list

Need to sort the foundry, need to cast some aluminium billets

Andrew

awemawson · « **Reply #84 on:** January 24, 2014, 08:51:12 AM »

Today officially is MY LUCKY DAY

I recently won a reasonable quality 200 mm chuck on eBay that was nice and local.

First task today was to draw up a converter plate to mount the three jaw chuck on the 4th Axis table. Table has four main slots and four auxiliary smaller ones. As the face mounting chuck has three holes at 120 degrees and the slots are 90 degrees nothing matches - hence an interposer plate.

Then I looked at the extra four tapped holes someone has added to the table - hey just maybe two of those line up with the chuck and the third goes in a Tee nut in a slot. Tried the first pair and they were close but not quite right by about 3 mm. Tried the second pair - spot on, exact, bullseye

I'll need to round the end of a Tee nut to slide further into the short slot, and no need for an adaptor plate - yipee !

awemawson · « **Reply #85 on:** January 24, 2014, 08:59:53 AM »

Then the second bit of good luck today:

I wanted to sort the air solenoid that drives the 4th axis table brake - it needs to work off 24 v DC and the original is 100v AC. I fired it up via the variac, and although it works it hums like an angry wasp. Pulling the coil off it, it became obvious that it wasn't viable to re-wind the coil as I had intended as it was a sealed unit

I considered using it via it's own transformer arrangement, but I really don't want 100 v AC in the environment that it will have to live in with coolant and swarf flying everywhere.

So what to do. Well it's obvious really - chuck it out and start again. Now the original sits on the end of the operating cylinder, and the valve ports directly link to the cylinder so unless I was VERY lucky and found a similar one, I need to make a simple adaptor to clamp on the cylinder end to take normal push fit pipe unions, and use a standard air valve.

ebay provided me with a brand new one with 24v DC operation for the princely sum of £9.99 including postage

awemawson · « **Reply #86 on:** January 25, 2014, 11:32:14 AM »

Today was really a day for plotting and scheming.

Firstly to decide on how to get cables in and out of the unit. The encoder has an umbilical cord which I don't really want to re-make, and the servo motor, inbuilt tachometer, in built thermal alarm, coil for the brake solenoid, and magnetic sensor that says the piston has moved, along with the reference micro switch takes 12 cables, however due to some being common we need a ten pole connector. So I will make up a second umbilical cord and have a box mounted on the machine with chassis mount sockets. Decided on the Amphenol '97' series connectors as I already had the plug, just need the chassis mount socket. Order placed with RS components promised delivery Monday.

Secondly, how to adapt the end of the pneumatic brake cylinder to give me a standard 1/8" BSP port where previously the solenoid valve mounted.

A bit of poking around convinced my the easiest way was just to clamp a suitable adaptor on the end using the existing mounting screws and gasket.

So adaptor turned up out of a convenient aluminium bar end:

awemawson · « **Reply #87 on:** January 25, 2014, 11:37:58 AM »

Then I needed to decide how things like the new pneumatic valve, cable clamps and covers would be mounted. First off I set too cleaning up the original mounting bracket so I could see what was what, and how the original cover was retained.

awemawson · « **Reply #88 on:** January 25, 2014, 11:43:13 AM »

I will probably retain the original bracket, cutting a few bits off where it fouls the encoder cable, and extending it to give more support for a far larger cover and the cable clamps.
The original cover is far too small. I want it to extend over the rear to afford protection to the encoder and servo motor, and be stout enough not to suffer when the 4th axis is lugged about. It is a VERY heavy beast, and it's hard to be delicate moving it.

I'll probably use the existing cover as the start of a pattern to develop the new one. Knowing how, when the machine is in full flight, chips go absolutely everywhere and coolant gets splashed over everything, it needs to be all embracing, but with an open bottom for anything that enters to fall out

Can't do much more to this until the new pneumatic valve arrives, as the bracket and cover need to be designed to fit round it.

Pete W. · « **Reply #89 on:** January 25, 2014, 11:47:21 AM »

Hi there, Andrew,

That project looks to be coming along fine.

Have you decided what sort of a test-piece you're going to set it all to to try it all out?

awemawson · « **Reply #90 on:** January 25, 2014, 12:16:34 PM »

Well Pete that's a good question and perhaps one for suggestions from the forum

I'll be very pleased to see it under servo control and being able to shift it back and forth using the TNC355 controller - I expect it'll be simple hex or the usual coarse square screw thread, though I was watching some cylindrical engraving on Youtube that set me thinking . . . . need to read up on that

awemawson · « **Reply #91 on:** January 30, 2014, 03:34:13 PM »

A little bit of progress over the past few days:

I've been researching the power side of the servo system. In the Bridgeport Interact where the Bosch Servo driver card and SEM servo motor originate, there is a whacking great 3 kW auto transformer with enough tapping to confuse a Bletchley Park decrypter:

awemawson · « **Reply #92 on:** January 30, 2014, 03:43:46 PM »

Now I don't like autotransformers at the best of times, as they offer no isolation, and that number of tappings is silly

Going through the Bosch manual for the drive, what it needs is two 180 volt windings aranged 180-0-180, and a further 200 volt winding added to one of the 180's to give 380 volts. The main power is taken by the 180 volt windings, and the 380 volts uses a measly 10 VA, and feeds an internal transformer generating 15-0-15 volts DC and also it uses this winding to get the synchronisation of the thyristor drives.

Searching around there were no suitable transformers 'off the shelf' or available second hand, so I'm having one wound. But what power? Well in a Bridgeport Interact it would feed three axis's and potentially a 4th axis rotary drive, and was 3 kW. Pro-rata that means that my 4th axis only needs 750 VA, and I reckon that's probably over stating it as the main axis drives will take more than the 4th axis.

I've simplified the secondaries to just the three needed ones, but have added a 415 v primary for isolation, with a 240 volt tapping so I can drive it off mains on the bench for testing. I upped the 200 volt winding to 50 VA so that they don't use stupidly thin wire.

As a 1 kVA transformer was not much more than a 750 VA that's what I plumped for:

awemawson · « **Reply #93 on:** January 30, 2014, 03:46:59 PM »

Meanwhile I've got on with the physical mounting of the Bosch servo board and it's twin wound choke. Nothing too complicated, just a bit of a pain as the space is a bit tight. Back panel of the enclosure is 2 mm steel, so everything is mounted using 6 mm hank bushes as 2 mm although it could be tapped and would be ok for first mounting, would strip after a few 'ins and outs'

awemawson · « **Reply #94 on:** January 30, 2014, 03:48:47 PM »

The X, Y, and Z axis are driven by Indramat AC servos - it would have been nice to use one for the 4th axis, but they fetch telephone number prices even second hand

awemawson · « **Reply #95 on:** January 30, 2014, 03:50:37 PM »

So now I'm waiting for the transformer to be delivered - (expected middle of next week) and also the air solenoid for the brake operator, which should be here already but isn't

. . . getting impatient in my old age

Pete. · « **Reply #96 on:** January 30, 2014, 06:12:32 PM »

I have a couple of 12v air solenoids left over form a project. Got some random servo/stepper drives too I'll be selling soon. You could have had one of the soleniods for your project Andrew.

awemawson · « **Reply #97 on:** January 31, 2014, 04:09:05 AM »

Pete that is very kind. The requirement is pretty specific - 5 port with 1/8" bsp tapping and 24 v DC coil, also physically small. Hopefully the post man will bring it today !

awemawson · « **Reply #98 on:** February 09, 2014, 06:50:32 AM »

Well time goes by - I've made quite a bit of progress, but some has been up blind alley ways

Firstly, in the picture above where I've mounted the Bosch servo card with it's associated choke just below it - I've had to re-locate the choke, as I found a tiny little paragraph tucked away in the Bosch manual saying that the servo card must be at least 300 mm away from transformers and chokes - it is presumably susceptible to magnetic fields. I'm sure that they are closer than that in Bridgeport Interacts !

I wanted to have the 4th Axis with it's cables easily de-mountable, so have placed a termination box on the machine right hand side where the Encoder, and two Adaptaflex covered cable forms will plug it. This also has a rotary enable / disable switch

awemawson · « **Reply #99 on:** February 09, 2014, 06:57:14 AM »

Now getting wiring to that point was a bit of a b****r to say the least. There is a semi-protected way in which the power for the machine light runs in armoured cable, but although protected by shields, it is in the 'wet' part of the machine and chips do fly under the shield. I am running two Adaptaflex IP54 rated 20 mm flexible conduits in this channel, but pulling them in took much ingenuity and messing about. The chassis sockets on the termination box have to be soldered in-situ, so to avoid soldering upside down in a confined space have minimum length tails to reach a table on which I could do the soldering, they then will be coiled up and put inside the box.

Fortunately I'd made a dummy test plate before I drilled the box, so used this to keep things at the right spacing.

News:

Author Topic: Servo Driven 4th Axis for CNC Mill (Read 82502 times)