Another Banana Bender's Build !!!

[Rodm1954]

Chich is going to have to stop hanging around people who are a bad influence.
You like a challenge with servos and now toolchangers. Look forward to seeing how you get on and given your perserverance and Chich's skills I am sure you two will pull it off.

[epineh]

Hi Rod, thanks for the vote of confidence, I am not sure if it is going to work but it is certainly worth a try

I was going to have a moving rack that was made up of another axis (or two) but I think I will just have a stationary rack at one end that holds the tools, with about 500mm of travel on the Y axis I should be able to fit a few side by side, say 50mm per tool should make 10 tools by my calculation...been wrong before tho

By keeping the rack stationary it should simplify things considerably, at the expense of a lot of "cool factor"...lol

Another thing I want to toy with is to convert a small car alternator into a BLDC spindle motor, I have heard impressive things of these, and the extra torque MAY allow for some rigid tapping, again not high on the priority list but I do want to fit as many features onto this build as I can, what it is lacking in size I want to make up with goodies...

Cheers.

Russell.

[TOTALLYRC]

Hi Rod, thanks for the vote of confidence, I am not sure if it is going to work but it is certainly worth a try

I was going to have a moving rack that was made up of another axis (or two) but I think I will just have a stationary rack at one end that holds the tools, with about 500mm of travel on the Y axis I should be able to fit a few side by side, say 50mm per tool should make 10 tools by my calculation...been wrong before tho

By keeping the rack stationary it should simplify things considerably, at the expense of a lot of "cool factor"...lol

Another thing I want to toy with is to convert a small car alternator into a BLDC spindle motor, I have heard impressive things of these, and the extra torque MAY allow for some rigid tapping, again not high on the priority list but I do want to fit as many features onto this build as I can, what it is lacking in size I want to make up with goodies...

Cheers.

Russell.

First time I heard of converting an alternator to a motor. What do you need for a controller???

Mike

[epineh]

There are a couple of options:

Roll your own, there are plenty of BLDC motor control chips out there, I have a few Motorola chips that will do the job, the ones I have are really for just running up to speed, braking and reversing. I was going to build such a drive a while back but never got the time to do so...but now may be a good time. I have some IGBT power modules that may be usable but the motor currents are usually pretty high so a discrete power stage may be in order.

Some of the control chips allow for very good control, ie cnc servo control, with a quadrature encoder that also outputs hall sense info you can have a high torque BLDC servo, this is what I have in mind, also will help for future projects.

Other option is to buy a controller, the alternator is basically already a BLDC motor or even syncronous AC motor as the rotor is a made up of a field and number of poles that normally generate an AC waveform out of the alternator, but can be reversed so that a three phase waveform is applied and the rotor locks to the rotating three phase field... also BLDC type switching can also be used, same result - motor spins.

I have always wondered if a VFD could be used to run such a beast, most VFD's have sensorless vector control these days so this means no hall effect's need to be installed, and have selectable motor voltage, so maybe the voltage could be set to say 50Volts and the alternator may not catch fire... I honestly have no idea but have started to do some research on the matter.

I will report my findings here

Russell.

[chich2]

WooooooW Heavy!
I feel so Dumb!!!
Nice Z axis Russ
Chich

[epineh]

Hey Chich, thanks

Don't worry, you will make us all feel dumb if you get the toolchanger going, lol

Been watching all the toolchanger DIY jobs on the tube (youtube that is ) and the one thing they are missing that I can see, is speed, which is the one thing that I really hope to achieve with this build. Did some quick calcs tonight and with 10mm pitch ballscrews and at least 3000 rpm from the servo's, geared down 2:1, by my calcs this should give me about 15m/min, while on the Z axis being direct driven has a potential speed of 30m/min...a little scary considering it is only about 150mm of travel, should make for some interesting startup testing...of course top end speed is nothing without fast accelleration, something I won't know about until the end.

I have heard various values for the particular servo's I will be using on the X and Y axis, speed from 4000 to 6000 rpm and supply voltage from 30 to 60 volts... I intend to use a 50volt motor supply and be conservative with rpm and work on 3000, any more will be a bonus. I am using 30volts with one on my first machine and it has heaps of go, in fact the only thing limiting the speed is the pulse rate of the PC, the servo can double its speed easily.

Any way enough dreaming, I am off to start drawing up the toolchanger spindle, may take a while, my CAD skills won't exactly get me a new job anytime soon, lol

Russell.

[neilw20]

Hi Russell

This is how a commercial spindle I have works.

The shaft is the hollow shaft of the motor, VFD driven to 80,000 RPM, 1KW water cooled. Not shown on the tool storage is a light weight collar that the plastic part gets held in to pull the drill out. Only 20 or 30 grams required.

In the real machine, a second head picks up the drill and puts it back in the tool station which is on the deck of the machine.

To load no air on diaphragm and spindle pushes down against plastic collar forcing nylon covered friction slug up the hole via the solidly attached collet.
To release the diaphragm move the rod about 0.25mm to release the collet and drill bit ends up back in change station.

It took me a while to figure out how it worked, because at first appearance there are no moving parts. There is a grub screw up the center of the collet to adjust the position in the taper relative to the bottomed slug.

Not really hard to make one of these. All bits are 38mm end to tip.
Plastic collars are 7.5mm diameter and 4.5mm thick.
All shanks are 3.175mm (0.1250" = 1/8")

Collets are off the shelf replacement spares for the Italian/US machine.
I can get them, but no idea of the $s.

The little black bit in the collet is a hard rubber/nylon insert that springs the collet open.

Enjoy... My atc will be a monster for big tools, with some of these style shafts adapted as a normal tool in the carousel.

As a side note, while one hole is being drilled, you can manually load/remove drills by a simple hit with a punch, or just hit it against a jig to lock.
Two or three in the ATC would be sufficient for most small jobs.

Neil

[epineh]

Hi Neil, that is so simple, I wonder how it would go for 1/4inch shafts...may have to do some tests.

Thanks.

Russell.

[chich2]

Niel,
Interesting tool holder. I can see it would work well for drilling but feel the taper angle of the collet may release if subject to side loads during milling. I'd like to hear your thoughts and experiences with this.

Thanks for sharing,
Chich

[neilw20]

The pulling force on the collet is determined by how much force it takes to overcome the friction on the nylon slug. More friction means more force multiplied by the mechanic advantage of the taper, which is quite fine which by my measurement gives about 14:1. assuming it takes 15kg,and it may be a lot higher, that gives a clamping force of 200Kg minimum, on a 3.175mm (1/8") shank. There is no way the milling forces on these cutters is more than about 10% of the holding force. as they bust at maybe 10Kg or less.
These get used in high speed milling operations all the time whenever a PCB outline is routed. One of the nice things with this is that there are NO SPRINGS. The slug is moved against the friction to unlock and lock. The collet is firmly locked to the slug.

I am not advertising. Just observing...

[chich2]

Neil,
Thanks for the details. That's a very impressive clamping force.
Sounds like a great system.

Chich

[epineh]

Managed to source the electronics enclosure, I am going to use a standard electrical switchboard, the type without DIN rail circuit breaker mounting. This will allow me to completely wire up the transformers, caps, drives etc while it is all sitting on the bench, then move to the mounted switchboard later on once the framework is done, and simply "join the dots"...well something like that anyway.

I am going to try keep everything as simple as I can, my plan is to have everything in the switchboard, PC, MESA 5I20, BOB, servo drives and IO opto board, and only have a pendant as the user interface. This will mean a pretty busy pendant but I am sick of using a mouse and keyboard on my current machine as dust gets EVERYWHERE, and this machine is going to get some heavy usage with a bit of luck so it will need to be well protected. I need to be able to see the drive status without opening the cabinet so I will have a mimic panel with simple info, drive status, current limit, position error and also drive reset buttons on the door, tonight I am going to draw up the basic layout, and add to it as I think of extra things.

Nothing worth a photo yet but I am still going, soon I will start on building the servo drives, these will need output to the mimic panel as well as the rest of the electronics. I am also going to build an off board current trip so the enable has to be pluggable to the servo drive to allow this in future.

Initially I will use the same drives as my first machine, but bypass the "smarts" later when I work out the MESA card, but closed loop to the PC should be interesting...

Russell.

[neilw20]

Russell. DIN rails are the way to go. I built a complete 4 axis controller system, with VFD in 3 days last week. You build it all on a plate, then put the plate in the box. You can use AL plate to double as a big heatsink if you need one. The removable backplate was predrilled on by CNC for 10 plugs, all wired up with the rest of it, then put into the box with the finished assembly.

Look at the pix all you, but ask no questions. I have no answers! (They are not mine to answer)

[JLSD]

Hi,
Only just picked up on this thread, reminded me I've work to do in a direction that might be of mutual interest!
I'm thinking of using RC "outrunner" motors for home-brewed spindles, a) monster power/weight, b) big bang/buck!
I'm going to trial a drive using a VFD & 3 big toroids as an alternative to the RC controller, reasoning that the motor simply wants to see 3-phase at big amps, but low volts, and the stock controllers look like electronic grenades if anything goes amiss!
The only predictable downside is that injection braking simply won't work, & regen braking may be iffy, but neither issue is problematic (to me anyway!)
I need to do some more "research" (ask clever people questions!) before kick-off.
If anyboy else has already tried this, I'd be grateful for any info, especially if it helps me avoid silly mistakes!!( I was only supplied with the one life, and I've used most of it already!!)
Regards,
John.
"Ploughing your own furrow" gives you a ready-made ditch to drive into!

[neilw20]

A current mode controller system like used in VFDs can be made short circuit proof, but the development curve is treacherous and exciting.
Easy but inefficient is 3 big audio amplifiers, one for each phase with good feedback from a 3 phase sinewave generator. Easy with a micro with three D-A outputs. It will just weigh a lot and fill a big cabinet.

For less than $600 (i know it looks big) but you can then drive a 3 phase 2 pole 2850 RPM motor to about 6500 RPM with absolutely no drama, and a minimum speed of around 150 RPM.

If you shop around you might get it down to $400. It'll cost you that for that fat toroids and all the electrolytics, fuses etc.

Just a thought after using 2 VFDs'. They rock.

[epineh]

Russell. DIN rails are the way to go. I built a complete 4 axis controller system, with VFD in 3 days last week. You build it all on a plate, then put the plate in the box. You can use AL plate to double as a big heatsink if you need one. The removable backplate was predrilled on by CNC for 10 plugs, all wired up with the rest of it, then put into the box with the finished assembly.

Look at the pix all you, but ask no questions. I have no answers! (They are not mine to answer)

Hi Neil, I may use DIN rail yet but need to get all of the components together and see how it will all fit into the enclosure, I will be doing as you mentioned and using the backboard as a chassis plate, mounting everything with it on the bench then dropping it in. About the only thing I have that has DIN rail mounting is a 24V power supply I was going to use for my limit/homing/estop switches, but I guess an RCD and MCB or two would also be nice.

Also the enclosure is way bigger than it needs to be, I think it is about 700mm square, the PC will take most of the room but I want to use one of those tiny MB's, about 100mm by 100mm, just so long as it has a PCI slot for the MESA card.

Hi,
Only just picked up on this thread, reminded me I've work to do in a direction that might be of mutual interest!
I'm thinking of using RC "outrunner" motors for home-brewed spindles, a) monster power/weight, b) big bang/buck!
I'm going to trial a drive using a VFD & 3 big toroids as an alternative to the RC controller, reasoning that the motor simply wants to see 3-phase at big amps, but low volts, and the stock controllers look like electronic grenades if anything goes amiss!
The only predictable downside is that injection braking simply won't work, & regen braking may be iffy, but neither issue is problematic (to me anyway!)
I need to do some more "research" (ask clever people questions!) before kick-off.
If anyboy else has already tried this, I'd be grateful for any info, especially if it helps me avoid silly mistakes!!( I was only supplied with the one life, and I've used most of it already!!)
Regards,
John.
"Ploughing your own furrow" gives you a ready-made ditch to drive into!

Hi John, I don't know enough about this to add much, I had thought of doing the same with an alternator conversion and VFD but the smell of electronic smoke really affects my hay fever

Russell.

[epineh]

Chich has been doing some fine tuning to the Z axis and it is really starting to look the part, should be pretty sweet once it is all up and running.

I am saving for a spindle now so things will be quiet for a while, once I have it the toolchanger build can start, several drawings have been done in chalk and at least two beers have been drunk so that takes care of the planning stage

The design has changed a little, in that I am going to make this a cantilevered gantry, I have a couple of reaons for this, the main two are simplicity of building and ease of job loading, with no side gantry slide at the front it is all nice and unobstructed.

It will look a bit weird but hopefully it goes to plan and works for my needs.

Russell.

[epineh]

Well time to brush the cobwebs off this build, I have received my spindle today, 0.8 KW air cooled with er11 collet, I already have a VFD for this, 1.1 KW so I have a tiny bit up my sleeve so hopefully I can get reasonable spool up times, not the end of the world but always nice to push things along.

I have a friend with a really big VMC who has kindly offered to cut out the pieces for the router for me out of 16mm plate aluminium, I have purchased about half a sheet and hopefully that should be enough to make the whole machine.

Bed size is still a respectable 400 by 800 and will consist of a solid sheet of 16mm plate ally with a grid pattern drilled and tapped for job clamping (probably 6mm), and also I am going to try Rod's method of cross clamping as I am doing something similiar with my first machine, though not as pretty looking as Rods setup

Still unsure of the servo drives, at this stage I am going MESA 7i43 anything IO card with "Dumb" H Bridges with EMC closing the loop to the PC, this is a little daunting for me but I hope to plod along and give it a go.

Plan B is to build standard step/dir servo drives and I was always going to try build a version of the DSPIC servo drive, it originally used an obsolete op amp for the power stage but now has a standard PWM output to drive discrete power stages and works very well by all reports.

Unfortunately I have decided to put a hold on the toolchanger until the machine is finished, I will still have a go at it but for now the rest of the machine will take priority, besides I am thinking of starting from scratch rather than adapting an existing spindle which should give me a few advantages.

Enough of my ranting, I will post some pic's once I have some plate aluminium cut out.

Cheers.

Russell.

[chich2]

Is it Made Yet? Is it Made Yet? Is it Made Yet? Is it Made Yet? Is it Made Yet? Is it Made Yet? Is it Made Yet? Is it Made Yet? Is it Made Yet? Is it Made Yet? Is it Made Yet? .......................... How about NOW???? ................. Now maybe?????
Can't wait to see your build Russ. It's going to be a sweet machine!

Chich

[epineh]

Lol Chich, no not yet

At least I am back on my feet after being crook for a few weeks, not quite 100 percent but almost !

I am tempted to try machine the Z axis servo mounting plate with my old router, never did try cutting aluminium with it, would be interesting to see how it goes, sounds like a good weekend project.

Cheers.

Russell.