By continuing to use this site, you agree to our use of cookies. Find out more
Forum sponsored by:
Forum sponsored by Allendale Jan 24th

Power feed for a VM32L mill

With full CNC later, maybe?

All Topics | Latest Posts

Search for:  in Thread Title in  
Iain Downs18/04/2020 17:58:23
687 forum posts
618 photos

Hi, there.

I recently took delivery of what for me is a BIG mill. VM32L which is pretty much the same size as WM18. Whilst, I love the size of the table and being able to have a couple of areas on it, it's quickly become obvious that a Y axis power feed would make life much easier. So I'm going to make one. But I need some help!

As a background, I've already built a CNC system for my (much much) smaller CMD10 mill. Even to the point of writing my own CNC driver and controller (I'm a software keep by profession even if demoted to just talking to people in my old age).

What I have in mind is to build the electronics (arduino) with 3 axis in mind, but only one motor for now. I have some TB6600 drivers a 24V 20A supply and a bunch or arduinos. I expect I will control it from an old android tablet with a Xamarin program. So arduino (Iain's Code) to drive the driver and a simple app to pick speed and stuff on the tablet.

So what am I missing? I thought this would be a decent motor. I think it's overkill but it's not much more expensive than a smaller one and maybe the Z would need it.

I'm thinking that I either want to have a clutch which can disconnect the motor from the leadscrew when I want to be manual, or to have a relay or similar which can disconnect power to the motor. In this case I would arrange it so that when the power was turned off, the relay would disconnect the connections to the motor (after a second or so).

I kind of prefer the mechanical approach, but have no idea how to go about it. I'm thinking some kind of clutch, I'm thinking something with splines. I'm thinking it sounds very complex and hard to make. I'm thinking I could do with some advice!

The second think that would be useful is a drawing or model of the endcap (if that's the right word) that the handle is connected to. I can take it off and measure it up, of course, but I doubt my accuracy and it would be irritating to make a frame and find that it doesn't fit.

Any advice or pointers would be much appreciated.

Iain

Martin Connelly18/04/2020 18:18:46
avatar
1463 forum posts
168 photos

If you are going to create a controller program on the Android tablet why not emulate one of the pendants used on a CNC machine. These give fast motion or jog. Jog is selectable for steps such as 1mm, 0.1mm, 0.01mm and 0.001mm. With a manual pulse generator (mpg) there is also an option for a slower motion with speed controlled by the rate the mpg is turned. It may be easier than designing a clutch if the programming is possible. There is a software version in Mach3 that you could look at for ideas as well.

Martin C

Iain Downs19/04/2020 10:52:39
687 forum posts
618 photos

I probably will put a jog option on it, but I feel I need to be able to move by hand for somethings where I need to 'feel' my way into a piece and back off if things look like they are going west.

One thing which confuses me is that the the motor referenced above claims 4.2A and 2.96V. Why, therefore, am I considering a 24V power supply?

Iain

Iain Downs20/04/2020 20:37:26
687 forum posts
618 photos

This is the sort of thing I had in mind..

clutch concept.jpg

The idea is that the motor moves forward on the bars and engages the pins. You'd need a lock of some kind and I don't know if pins would be better than a wedge or bar.

Youu can't see it, but the two timing pulleys aren't quite in line (imagine the belt, that's beyond my 3d skills!), so it still needs a bit of fettling.

Iain

SillyOldDuffer20/04/2020 22:20:22
Moderator
6328 forum posts
1388 photos

Hi Iain,

Gremlins got at the motor link in your first post and doubled the URL. Should be.

2.96V is the DC voltage needed to maintain the current (4.2A) in order to produce the motor's rated holding torque, ie ability to resist turning when stopped. The watts expended on holding won't overheat the motor within these limits.

Normally though stepper motors are turned by pulsing the windings rather than held stopped with a constant current.

Best step performance is achieved by getting the motor up up to rated current as quickly as possible. As the edge of a pulse rises faster with increased voltage it's usual to run the controller distinctly higher than the motors holding voltage and, because this could cause overheating, to keep the motor within ratings by managing the current. So the T6600 controller uses DIP switches to clamp the maximum current it will deliver irrespective of the input voltage.

With stepper motors amps and pulse rates matter more than volts and they work better when the controller has plenty of volts so the amps can be delivered quickly.

Rather than slide the whole motor, how about fitting a dog clutch between the motor and mill? Could a pair of contrate gears where one is slid to engage the other by a lever, or just a pair of bolted discs that the operator unscrews manually.

Dave

John Pace21/04/2020 09:52:42
203 forum posts
188 photos

I have two similar disconnects on stepper motor drives to go from
power feeding to manual ,they both use the same type of connection.
The first one here disconnects the cross slide of my Warco lathe,the hub
is driven by the stepper motor and the three legged spider connects
to drive the gear when the three bolts are fitted.
The second one is much the same fitted on the Universal cylindrical grinder
to disconnect the table drive ,both of these have a ball race at the centre,the
leaves of the spider and the drive on the grinder connections have a small
amount of flex to clear the drive when not connected ,when the bolts are fitted
the connection is clamped up tight.
Pin and dog clutch drives may eventually become loose because of the
action of the stepper motor drive and rattle .
It would perhaps difficult to fit this on the stepper motor drive in your application
but you could fit something like this on the other pulley which would work just as well.

John

stepper motor 1.jpg

stepper motor2.jpg

Alan Jackson21/04/2020 10:43:32
avatar
199 forum posts
82 photos

Hi Iain,

This approach might solve your problem;  the worm attached to the stepper motor shaft can be rotated on its eccentric mounting to engage or disengage with the worm wheel on the feed screw. You also get a good reduction drive (say 20 :1) to give a slow speed and high torque. The stepper motor speed seems to be just about right.

Alan

Cross slide stepper drive

Edited By Alan Jackson on 21/04/2020 10:44:03

Martin Connelly21/04/2020 11:29:13
avatar
1463 forum posts
168 photos

Iain, I think tentatively milling is a good way to get an uneven finish and rapidly worn, chipped and blunted cutters. One of the thing CNC milling encourages you to do is learn about feeding at the correct rate to get good cutting processes and avoid rubbing or undersized chips that mean you are not using cutters as they were designed to be used. Look at what Jason is doing with his build logs, no feeling his way in to the material there.

Martin C

Bazyle29/05/2020 13:27:23
avatar
5471 forum posts
206 photos

I know this thread is a little old but I came across it looking for something else.

Looked at the picture above by Alan.
It seems too clean to be real and too detailed to be a CAD drawing. Which is it?

JasonB29/05/2020 13:32:12
avatar
Moderator
18884 forum posts
2080 photos
1 articles

Have a look in Alan's album, there are a couple of his "Stepperhead" machine with swarf on it

Iain Downs29/05/2020 15:39:18
687 forum posts
618 photos

Hi, all.

Prompted by Bazyle's post, I thought I'd tell you where I'm up to with this.

This is what I plan to build

power feed from onshape.jpg

I'll replace the Y axis handle with a timing pulley, which will have a new handle bolted to it. I'm going to lose the graduations I think, but I've had news that my DRO kit may not be far off - and there's still the right hand handle.

The timing belt (not shown) connects to a timing pulley held on a bearing. On the outside (left hand side) there is a 3 fingered dog clutch (three clawed?). The motor has a 3mm keyway with the other half of the dog clutch on.To engage I slide it forward and tighten a lock screw. To disengage move it back.

I now have all the parts. I've built the electronics and the driver software on the arduino (well, I still have to include an acceleration profile and the code which stops when an end stop is reached). Even more I've tested it with a smaller stepper and it works! The big stepper arrived today.

I've also got the pulleys, timing belt and the bearing. And probably enough stock metal to make the bracket.

I'll post up a photo of the bits this weekend and of any progress I'm not too embarrassed about!

Who knows, it may even work!

I think the biggest challenge is going to be to get the motor lined up with the pulley so that the clutch engages correctly.

Iain

Iain Downs29/05/2020 15:42:59
687 forum posts
618 photos

Oh!

I forgot to ask if it is worth using shielded cable for the motor to reduce noise (which could affect the DRO when it arrives). If so is there any obvious source?

This on ebay looks like the right stuff, though I'm not keen on cable with 3 black wires and one green.

Iain

John Hinkley29/05/2020 16:09:16
avatar
950 forum posts
325 photos

Iain,

You say you are motorising the Y-axis but refer to the "right hand handle". Do I take it that it's actually the X-axis that's getting the motor upgrade - i.e. left to right? I'd be inclined to do that one first, anyway. Much more useful in my opinion and experience, limited though that is.

John

Iain Downs29/05/2020 17:32:20
687 forum posts
618 photos

Yes, John. Left to Right I got the naming wrong here.

Iain

duncan webster29/05/2020 23:33:16
avatar
2795 forum posts
41 photos

If you're driving it with a direct (1:1) stepper motor you don't need to declutch to manual wind. I have a stepper on the X travel on my Centec, 0.1" pitch leadscrew. The motor makes it 'notch' at 0.0005" increments, but I'll live with that.

ChrisB30/05/2020 06:55:07
548 forum posts
192 photos

I think a clutch would be a good idea Duncan. Although as you said, if the motor is a direct drive the resistance will be low, there still is a noticable resistance and worse than that, when you turn the handwheel you are feeding back some kind of voltage as I can see all the lights on the control box flickering.

I chose to make a simple friction disk clutch which serves a dual purpose, apart from decoupling the motor, it will slip if the load is too much or the end stop is reached.

Iain Downs02/06/2020 20:53:26
687 forum posts
618 photos

So I've got my design and I've started off.

The first step was the electronics, which looks like this

motor drive 02 electronics front view.jpg

The motor on the far left is a much smaller test motor. The bigger beast next to it is the real deal.

Two oddities are the switch which appears to light up when off. Hadn't expected that (or have any idea where it cam from) and the motor drive socket is too high, the switch interferes with the eject button. it still seems to work with a bit of a wiggle, though.

motor drive 03 electronics back view.jpg

And the working bits. On the left an arduino nano on some veroboard. I'm afraid I've never got to grips with making PCBs so it's old school. At the bottom a 24V 15A power supply and the black box claims to be a TB6600 drive. One thing I forgot when designing the front panel was indicators, so I'll be drilling more holes soon.

Incidentally, the way I made this was to do a design in OnShape, make a drawing, stick it to a piece of aluminium plate and drill, file and whatever against the template. Some of my filing's a bit rough, but it's hidden by the DB9.

All of these bits (apart from the motor I already had.

Next the hardware side

motor drive 01 hardware.jpg

The flat bits are for the mounts, the wide round steel bar for the dog clutch, the thin bar for the axle, the aluminium bit (probably) for the new handle and the pulleys, belt and bearing are obvious.

The plans are shown below

motor drive overview.jpg

Obviously the timing belt is missing between the two pulleys. THe first thing to make is the backplate

motor drive bracket.jpg

And the rest of the frame. The dog clutch is the only other difficult part (I hope)

motor drive dog clutch.jpg

Iain

Iain Downs02/06/2020 21:04:11
687 forum posts
618 photos

And I've already started

The big bit of 12mm hot rolled plate needs to be cut down, faced and then further machined.

FIrst cut on the Femi bandsaw (I love this machine!)

motor drive 04 cutting the back plate01.jpg

Which is all very well for cutting to length, but it's still too wide. So I dragged out my franken-grinder.

I build this before I had a chop saw (which melted) or finally invested in the femi, It wasn't a whole-hearted success on thicker items, but it was designed to be able to cut long pieces.

motor drive 05 cutting the backplate 02.jpg

This is an action photo. For the sake of those with weak stomachs let me say it was hot and I was wearing shorts. Sorry about the legs...

motor drive 06 cutting the backplate 03.jpg

I got better sparks at the top of the cut and it's only by chance that the wife came in and was persuaded to take this (not very flattering) photo.

Next to thin down to 10mm

motor drive 07 facing the backplate.jpg

Mark out

motor drive 08 marking out the backplate.jpg

And cut the slots and corner relief.

motor drive 09 slots and corner drilled and milled.jpg

Although I've marked out the hole for the bearing, I was really worried about how I could make sure it lined up with the motor spindle. I eventually thought that the best approach was to hold the backplate and motor plate together and drill through both to provide an accurate hole to work from. Lets see how well that works.

Finally to cut out the un-needed L part. This was on my recently made bandsaw table which worked rather better than I expected (in fact I tend to be quite pleased when things I make work at all! Perhaps I'm improving).

motor drive 10 cutting the corner out of the backplate.jpg

and the result (so far)

motor drive 11 backplate coarse cut.jpg

I've cut it a bit oversize so I can clean up on the mill and match this to the motor plate.

So far so good!

Iain

Iain Downs10/06/2020 20:11:12
687 forum posts
618 photos

Rightio. A bit intermittent over the last week, but progress has been made.

Here's the backplate (as above) plus the motor mount. I got the motor mount trimmed to match the backplate and then clamped them together to drill holes for bolts and, more importantly as pilots so to speak for the bearing and motor itself.

motor drive 12 backplate and motor mount.jpg

The motor plate motor recess was done on the lathe - spending (much) more time centering the 4 jaw than actually machining.

motor drive 13 boring out the motor mount.jpg

I swear that I get the direction to move the piece wrong at least one time in three!

The hole for the bearing in the backplate was done with a boring head, the new mill being a MUCH more pleasant instrument to use it on than the tiny CMD10.

One thing which was less good was that my (fairly) new co-axial indicator seems to be positioning the head a bit to the right of the actual centre. I can't even imagine a mechanism for this, but checking with a dial test indicator seems to suggest this. For now, I'm ignoring this issue, but I will need to review sometime soon sad.

I've also cut the axle to size, bored out the pulleys and started work on the dog chucks. The idea is to cut them on the mill in a hex ER32 collet, so I've cut a 40mm x 18mm cylinder as the support for a 40x10 cylinder. One done so far. I can't really take a cut bigger than 0.5mm on my lathe so it takes quite a while to reduce from 40 to 18, You can see the first one in the bottom right

Here are most of the bits.

motor drive 15 most of the parts.jpg

Tomorrow, with some luck, complete the preparations for the dug chuck. Not sure I'll get to cut the faces though.

Iain

Iain Downs14/06/2020 15:19:39
687 forum posts
618 photos

The more recent work for this has been mainly on the dog clutch. Sadly this turned out to be more of a dog's dinner as you will see.

The idea was to make a long stem for the dog clutch which could be held in an E32 collet for cutting the faces. You can just see one of the in the last photo in the previous post. Here is the face of the dog clutch being reduced (the clutch is 10mm high, 1 5mm plate and then 5 mm of doggy bits).

motor drive 16 dogclutch.jpg

Here are the 2 parts with the Hex ER32 collet holder in the background.

motor drive 17 dog clutch.jpg

And this is a complete mess.

motor drive 18 dogs dinner.jpg

The one on the right was my first attempt. It's not going to win any canine awards, I can tell you. The stupid thing I did was to rotate the holder by 6 degrees not 120, which doesn't work. I missed that in the otherwise rather good You Tube video which told me what to do.

The second mess up ( on the left) was just because I didn't tighten the chuck enough and it span.

However, the second shot worked rather better.

motor drive 19 dog clutch 2nd time.jpg

I had to do a bit of trimming with a file, as the lands were a littl bite too big, but better that than too small where it would rattle. The parts fit nicely together in all three orientations (after a little fettling) and there is only a barely perceptible slack.

Here's all the bits I've made so far and you can start to see how it fits together.

motor drive 20 more parts.jpg

The only major thing left to do is to cut keyways for the leadscrew and the motor clutch. I've decided to leave this for now as, although I've made the keyway cutters, I want to make a spindle lock for the mill before I cut the keyways. Then the terrifying part - put it all together and see if the bits line up without too much 'help'.

Iain

All Topics | Latest Posts

Please login to post a reply.

Magazine Locator

Want the latest issue of Model Engineer or Model Engineers' Workshop? Use our magazine locator links to find your nearest stockist!

Find Model Engineer & Model Engineers' Workshop

Latest Forum Posts
Support Our Partners
ChesterUK
EngineDIY
emcomachinetools
Warco
cowells
Eccentric July 5 2018
Eccentric Engineering
Subscription Offer

Latest "For Sale" Ads
Latest "Wanted" Ads
Get In Touch!

Do you want to contact the Model Engineer and Model Engineers' Workshop team?

You can contact us by phone, mail or email about the magazines including becoming a contributor, submitting reader's letters or making queries about articles. You can also get in touch about this website, advertising or other general issues.

Click THIS LINK for full contact details.

For subscription issues please see THIS LINK.

Digital Back Issues

Social Media online

'Like' us on Facebook
Follow us on Facebook

Follow us on Twitter
 Twitter Logo

Pin us on Pinterest