Page 6 of 7

Re: DIY Mill Turn Center

Posted: Thu Oct 12, 2017 2:36 am
by Generic Default
Captain Starfish,

Does Kickstarter really eat 30%? They say 5% for them and 3-5% for payment processing on their website. My estimate was that I would need $90,000 to do this safely, any less than that and I would have to be careful at every step for cash flow.

Outsourcing all production would be prohibitively expensive in the US, and I won't even consider production in China or Taiwan for a bazillion reasons learned the hard way by other people who have done kickstarters. Skyfire is not an example I'd want to follow.....

My plan was to rent a small industrial space with a roll up door, and finance a large VMC. The majority of the build costs of this are in the aluminum parts. I would outsource the few turned steel parts (spindle tubes, B axis rotor and turret rotor) but I would mill all of the blocky aluminum parts before sending them out for anodizing.

The plan is 4 vises on the mill bed, aligned perfectly on the X axis. This would let me make the long boxways accurately in 2 or 3 operations, then do the more square parts 4 at a time (one in each vise). I already have the cycle time estimates for each part, but it's dependent on the machine that makes them. No more than 10 minutes for any operation on any single part.

I'm in the prototyping phase right now, and I only have access to weak, inaccurate machines by industrial standards. Having a full sized VMC with a powerful spindle, automatic tool changer, enclosure, dedicated tooling ect. would make easy work of these parts after the initial setup and programming. If a job shop did these parts for me, they would charge markup based on the material price rather than the cycle time. By having my own machine, the cost would be a monthly payment rather than a piece-wise rate.

Plus I've been getting quotes on suitable VMCs for months now, I have a list from $80,000 on the low end to $170,000 on the high end. Not upfront cash though, these are usually financed at 10-30 percent down and then monthly payments for the next four or five years. Shop would be ~2,500 per month as well. I crunched the numbers based on expected machining costs, and it turns out that making more than four machines per month justifies having my own shop and VMC.

About the electronics now...
I just checked the smoothie site and they updated it with bad news for me. There is no chance of the board being released this year, they say they are hoping to do a kickstarter in a few months which means the V2 will be a mid-2018 product at best.
The Duet board is my best bet, with the expansion it's 10 steppers total. But it would be ~300 dollars. I may have to re-purpose the azteeg board with 8 steppers that I already have, and maybe one of my unused 5x smoothies as well.

Sounds ridiculous, but I may have to wire the milling head and the turret to separate boards, and run two repetier-host windows on my PC to control each board. This would let me have 13 steppers, and I could do turret and mill ops separately or simultaneously.

Obviously not practical for a post-kickstarter product, since there is no communication between the boards. Eventually I will have to write some custom firmware for this no matter what.
Sorry for the excessively long post with no pictures, I'll get back to visual stuff at the next post.

BTW, I have reasonably accurate cost estimates now, tell me in your next post what you think the retail price of this thing would be.

Re: DIY Mill Turn Center

Posted: Thu Oct 12, 2017 3:41 am
by Captain Starfish
I'll admit I was only commenting on the basis of half heard comments about Kickstarter from the friend of a friend, so it might not be that bad. Interesting to hear another perspective on the Skyfire saga, I know Iain at least sounded supremely happy with how it all worked out. Other funding alternatives: need an investing angel or bunch of individuals more interested in seeing the product than making 99% returns :)

I'm a contracting nerd, my main customer for the last few years has three old Bridgeport VMCs and then last year added a Haas EC500 horizontal machine with a tombstone setup which cost 'em about half a million bucks and all we're hearing from the guy hired as the eggspurt on running it is excuse after excuse, latest is getting a shiny big VMC in next to it to make tooling so the EC can be productive. Meanwhile, the Bridgies (worth maybe $15k now) are trucking along happily. Just a thought, anyway, when considering production machines.

Electronics: for that many moving parts I think $300 is cheap. Hell, I dropped twice that on my router for NEMA 24 drivers and a Mach3 ethernet controller, and that's only 4 axes.

"What do you think the retail will be" is a tricky one because if I say X, and X > your expected retail price, I just got boned by my own big mouth. So I reckon about twenty bucks? :D

More seriously, I'd be thinking with steppers * 8 and drive electronics, that frame, bearings and spindles and screws and what not, you'd be doing very very well to get it under $2500 but I'd not be surprised if it was twice that. FYI at $2500 if I knew I was going to be able to use it (ie there was some halfway manageable software behind it) I wouldn't think twice, at $5k I'd probably need a job for it to justify the purchase.

Re: DIY Mill Turn Center

Posted: Thu Oct 12, 2017 6:10 am
by forrie
Even if you have built the best machine in the world it will be worthless without decent electronics and stable firmware, software \ workflow. Even if you market it at the low end $ range I wouldn't touch it with a barge pole without it. I'd be prepared to do some opensource community tinkering to improve the product, but unless you have the basic operation down pat, well...meh.

Im not so sure how well the duet would work as a turn center controller, maybe reach out to DC42 on this forum and ask some questions? Never know, he might give you a hand with the firmware side to sell more boards.

General rule of thumb in manufacturing is materials + labour + overheads + % profit. You also needs to factor in demand, market competition, blah blah blah economics is fun! I'm sure you will figure it out :D See how I non-committal I was :lol:

Re: DIY Mill Turn Center

Posted: Fri Oct 13, 2017 12:35 am
by Generic Default
turret block in machine.jpg
turret block on table.jpg
turret block on riser.jpg
Not quite done yet, I still have to put threaded holes in the bottom and counterbored holes through the riser plate. But it's coming together, I should have the turned parts going next week.

And don't worry, your pricing guestimates wouldn't affect the retail price in any way. My main concern once I launch the Kickstarter will be to reach the funding goal, and to do that I can't be pricing this thing high.

The 300 dollars for the electronics would be doable for the full machine, but I realized that using the boards I already have makes more sense for the prototype. It's mostly about the mechanical stuff for now.

The turret block in the picture will fit big tapered roller bearings; the smaller pocket is for the geneva crank bearings. On the V1 prototype, I used an acetal plain bearing to hold the rotating shaft. It worked, but there was more friction than I wanted which limited the turret to no more than 1 tool position index per second. Also, the bolt circle around the large bore will fit a 3d printed cover to keep chips out of the geneva mechanism. There are tapped holes on the back for a live tool motor and cover.

Re: DIY Mill Turn Center

Posted: Fri Oct 13, 2017 10:27 am
by Xenocrates
Honestly speaking, if you can get it under say, 4K, then it's beating the Grizzly tools Gunsmithing lathe, with CNC controls, and it's pretty sure to be actually accurate and stout, unlike a lot of the Grizzly tools. While I could buy an old Cinci Milacron for 3250, that comes with the baggage of an old machine and controller, plus the need for a large rotary converter, and would likely need specific wiring in my shop. So a tool that can plug into a wall socket I can install without a rotary converter (A 20A 110 or 208V turnlock would do alright by me), or stuff you can't buy in a standard hardware shop would be great.

Now, for me, I would prefer the Duet to the Smoothieboard V2, especially if you can get the web interface changed up a bit for mill/turn. Then I can have a nice large panel monitor to see what I'm doing on, (and also make it a lot harder to accidentally get the wrong file type uploaded to a given machine)

It's looking really cool so far. If you have to make it a premium product, do. Pocket managed to bring in 350K on a machine priced at 3500, which is less capable than this is, uses a vacuum motor for the spindle, and have plenty of deflection on the table. If you price a bit higher for something that's obviously better, you should be fine. Please don't cut corners to make it cheaper until you know the market won't buy enough. If you do, we might never see the machine you envisioned.

And please do check that your expected expenses include an air compressor, filters, coolant, and changing the later two with reasonable frequency. There's nothing worse than seeing a VMC with the tooling rusted into the spindle because the air was crap, or turning on coolant and getting a stringy bubbling strand launched out of the lines, or having the coolant pump go because it's clogged solid with algae. I figure you probably already looked into those, but I've found a few people who forgot to include them, and got burned on that.

Re: DIY Mill Turn Center

Posted: Sun Oct 15, 2017 7:06 pm
by Generic Default
Didn't quite finish the turret shaft, but I'll get it done in the next few days. Both the B axis and turret axis rotate on tapered roller bearings, shafts are 65mm where the bearings fit. Actually I designed the machine so all of the large bearing pockets are the same 100mm bores with 65mm shafts; the main spindles use precision angular contacts while and the low speed axes use tapered rollers. Most industrial CNC lathes use these things called curvic couplings and a pop-lock mechanism during the index, but mine uses a rotary plane from 2 heavily preloaded bearings and a tapered cam to locate the angle. This is just a larger version of what I did on the version 1 machine, and it seems to work great. BTW the shaft in the picture below is 3 inches in diameter.
turning turret shaft.jpg
Here's the machine with 3 major axes on. More progress coming this week, I'm hoping to get the turret octagon done and the main spindle turned.
oct 15 assembled 1.jpg
oct 15 assembled 2.jpg


There are quite a few old machines from the 80's and 90's that have depreciated under 10,000 at this point. Anyone can buy one, but the cost of shipping and rigging and wiring for 3 phase don't decrease with time. My machine will plug into any household outlet, both the 120v and the 240 appliance outlets. Since it uses hot-swap server PSU's, there is no need for phase converters or anything. The server PSU's output 12v 60 amps each, so in series they can do 48v 60 amps, which is just under 3kw. Eventually (long term) I'd like to get to the point of a battery augmented power supply for the whole machine. That would allow it to surge hundreds of amps for heavy cuts. But for now, the server PSU's are good enough.

It looks like I'll be doing two control boards for this prototype, I don't want to drop 400 dollars for a duet at this point since I already have surplus boards laying around unused. But I definitely want an on-machine touchscreen for the production version. I've done GUI interfaces before from the ground up, so I don't think it would be that difficult to program for a mill-turn as long as the base firmware is already done by someone else.

I'm not planning on cutting corners in production, which is why I want to do it myself. I would be offering several variants at different price points. The least expensive would be the base + headstock + turret model. Since the boxways on this machine are identical, extra axes can be added on later; so you could purchase the machine with just the headstock and the turret, then purchase the 5 axis milling column later on, then a 2nd headstock (subspindle) later on ect. It will be modular.

There are quite a few CNC kickstarters from the last few years, even a few from the last month! Nearly all of them that have a working prototype and a decent video get funded. Most get several hundred thousand dollars, some get millions. Only a handful of them failed to get funded, probably because they didn't have prototypes or business plans whatsoever.

Re: DIY Mill Turn Center

Posted: Wed Oct 25, 2017 10:37 pm
by Qdeathstar
CNC lathes aren't really accessible for dyi enthusiasts, software is defiantly going to be the largest challenge i think.

Re: DIY Mill Turn Center

Posted: Wed Oct 25, 2017 10:53 pm
by Captain Starfish
I dunno. With Fusion 360 now offering a fairly sane (and free for hobbyists) 5 axis CAM system including mill and turn operations, I think it's become a lot more readily available than it was before.

Of course, creating the post processor file for it will be a bit of a challenge. Users might struggle with "am I turning with live tooling or am I doing 4 axis milling" and similar decisions - and generally how to fit the machine's capabilities into the CAM mindset and strategies available. But I reckon basic mill-turn functionality is very, very achievable with pretty good free software.

PS - GD: send me a unit and I'll write a kickarse post for you for "free" :D

Re: DIY Mill Turn Center

Posted: Thu Oct 26, 2017 1:14 am
by Generic Default
Forgot about the fusion 360 thing, having free software for that kind of stuff makes complex parts viable. There are other options for CAM software that are free as well. I'd like to have a touchscreen controller and a full enclosure for this machine eventually. As long as the control board supports Gcode macros, making a conversational programming system is within my coding ability. For most functional mechanical parts, you can write gcode by hand or with a conversational system to get the parts done more efficiently than with real CAM software. For example, if you want to make a bore on the centerline of the part, you can either drill and bore to size, or you can plunge an endmill while rotating the spindle axis to make a helical path. I'll make a bunch of demo videos before the kickstarter.

Also, I have the turret mostly done; just need to make the toolholders that bolt onto the front, and turn the geneva pin so it can index. There will be a cover around the geneva mechanism and a box for the two motors on the back side. Having these two massive tapered roller bearings gives the turret great rigidity, and since there is no curvic coupling, there is no axial mechanism. No air cylinder needed. I'll be making the milling head and the main spindle tube next.
turret and b mounted.jpg
turret and b mounted 2.jpg
turret up close mounted 1.jpg
turret side mounted.jpg

Re: DIY Mill Turn Center

Posted: Thu Oct 26, 2017 3:09 am
by Captain Starfish
There are two CNC machinists at work. One on his old Bridgeports with their burnt-in CRT consoles, the other with shiny new Haas machines. The old-timer does EVERYTHING conversationally from dimensioned drawings using the Bridgeport language. He reckons he can set it up to trickle feed g-code from a computer but doesn't want to. The new boy does everything from step files generated out of PTC Creo by the mech.eng, into SolidWorks/SolidCAM and sends it down to the Haas machines.

What I've seen from these two is that the CAM approach is great for extremely complex pieces or ones with organic shapes (lots of NURBS surfaces etc) which are just too damned hard to do by hand. But, for most stuff where it's 2.5d sort of work with squares and holes and arcs, the Bridgeports are humming all the time where the Haas machines sit quietly waiting for programs way too much.

At home I tend to jump between the two approaches, and I strongly urge anyone getting into CNC to start with conversational g-code so they understand what the CAM is doing. Then get into CAM if you have to, either as a shortcut or when the parts get complex.

One thing - remember g-code ain't g-code. Some controllers will only support arcs in XY, some will support them in any of the three cardinal planes so long as you use the appropriate G-code to set the working plane first, some will let you arc in any defined plane. Some have better support for drilling ops. Some have better support for lubrication & cooling (flood, mist, air, air blast etc). Especially with a machine like this which is capable of such complicated coordinated moves, things like arc support are going to get quite important so be careful (if you have a choice) to get the best supported control you can!

Man, that thing is starting to take shape now. Looking really good!

Re: DIY Mill Turn Center

Posted: Sat Oct 28, 2017 8:35 pm
by Generic Default
With the control and support you're talking about, I have no real pre-made options. The electronic hardware will have to be something like the smoothie or duet board, and I'll have to write a branch of firmware specifically for my machine. There are no hobby level firmwares for a machine of this complexity and arrangement. Even in real industrial controls, there are only really 4 options; fanuc, heidenhain, mitsubishi, and siemens. They offer control electronics for mill-turn centers and swiss type lathes with many axes, but all of their controls are well over 10,000 dollars.

On to some little details of this build.
er16 holders with tools.jpg
I decided to use these little collet holders as standard tooling for the lathe tool turret. They're ER16 collet sockets intended to be put on motor shafts for hobbyist engraver machines. They're cheap, tiny, and compatible with with all sizes of er16 collets up to 10mm. They make these in ER20 as well, which can do up to 13mm (1/2").
These will go in tool blocks that bolt to the face of the turret. By having oversized bolt holes, the blocks can slide and rotate a little before being locked in position. This will make tool setting a lot easier, since you can clamp the drill/tap/reamer or whatever in the main spindle to align the zero position. It's a lot easier than having to indicate each tool in like on a full sized CNC lathe, and just as rigid.

Next is the power transmission. On the first V1 prototype machine, I tried to use rigid couplers with set screws to transmit torque for both the axis steppers and the spindle motors. It kinda-sorta worked, but it was unreliable over time and had problems with torque overload. On this new machine, I'm using these spider couplings for every axis and every spindle. They can take a lot more misalignment than rigid couplers, and they can clamp onto smooth round surfaces with much better holding torque.
shaft coupler on nema 24.jpg

Last thing for now are the encoders; these things are the least expensive, easiest to mount, and highest resolution of anything I could find. They offer several variants for both absolute and incremental positioning up to 4096 positions per revolution. The plan is to put one of these with absolute capability on both the main spindle and the milling spindle. If I end up using the O-drive controller with RC motors, these encoders fit on the shaft easily.
rotary encoder stuff.jpg
rotary encoder stuff 2.jpg
I probably won't bother putting them on stepper shafts even though the steppers have back-side shafts. Those will be for handwheels. I may end up mounting linear encoders on each linear axis since there are many DRO kits on ebay and amazon that are essentially just the working mechanism of digital calipers with 0.0005" resolution.

Re: DIY Mill Turn Center

Posted: Sat Oct 28, 2017 8:57 pm
by Captain Starfish
I wondered if you were watching the whole o-drive thing unfold when you were talking about batteries before - it's a tidy solution by the looks of it.

An ER-20 option would be great (says the guy who has already standardises on ER-20 on three machines in the shed :) )

Not sure how much added value you'd get, apart from maybe skip detection, by adding the encoders to steppers. Whilst glass scales etc give you a better feedback around not-quite-uniform lead screws, they do add extra fragility, complexity and expense. Personally, if I were rocking stepper motors I'd just count steps and allow a fine tuning function (measure 100mm of travel, compare it against step counts, store the progressive mapping along the length of travel). For servo drives the rotary encoder before any reduction gears/pulleys is going to be sufficient.

The firmware sounds like a fun challenge...

Can the Dynomotion (Kflop etc) not cope with large numbers of axes? Or even LinuxCNC running on a small quad core ARM card?

Re: DIY Mill Turn Center

Posted: Mon Oct 30, 2017 6:42 pm
by Generic Default
I don't think the firmware will be that difficult once I get used to the language, 95% of the work is already done for me by the people who made 3d printers mainstream. I just need to add a few tables and macros, and a correlation between B axis angle and Z axis position so the 5 axis head can be programmed easily. For now I'm still trying to finish the mechanical build of this machine with the limited time I have to do so.

I'd like to post some stuff about the milling spindle for your evaluation....
Below is the side view of the milling head and B axis, it's the part that slides up and down on the Y axis column.
milling head cross section 1.jpg
You can see the harmonic gear that will drive the B axis in the back (left). Two big tapered roller bearings in the middle, then the shaft that holds the milling head sticks out on the right side. In the milling head is the spindle. The spindle is driven by a pulley system at the top, with 6 round keys at the torque interface. The light blue shaft down the middle of the spindle is the drawbar.

To keep the milling head ultra-compact, I decided to go with a novel approach for the drawbar system. It's just an M12 socket head cap screw. At the top of the milling head, there is an enclosure box (not shown). In the enclosure box, there will be a solenoid that pushes a hex-key into the socket head cap screw hex socket, which effectively prevents the screw (drawbar) from rotating relative to the entire milling head. However, the pulleys can still rotate the milling spindle tube. By engaging the solenoid while rotating the spindle, the drawbar rotates relative to the spindle. This means that the top side of a toolholder can have the drawbar screwing into it to tighten/loosen the thread fit.

Also not shown is the spring inside the milling spindle cavity and the captive nut on the drawbar. The spring presses down on the nut, which keeps the drawbar at it's down position in the spindle. The drawbar can be pushed up in the spindle about half an inch before the captive nut hits the top of the spindle cavity.

This whole system allows toolholders to be changed on the fly using only a solenoid and the milling spindle motor. No compressed air needed, no pneumatic torque wrench, no wrench-from-the-back. I'll be building it over the next couple weeks, once the physical model is made I'll demo it to explain better.
milling spindle cross section 1.jpg

Now onto the actual tool holders. On the version 1 prototype from last year, I made an ER25 spindle. This one will have an ER32 spindle. But there's a catch...

The spindle nose for the ER32 is almost the exact same size as a BT30 spindle. ER32 has a 32mm taper entrance with a 16 degree slope (8 degrees per side). The BT30 has a 31.75mm taper entrance with a 16.26 degree slope.

That 0.26 degree difference means that ER32 collets and BT30 tool holders look identical, but don't fit quite right.

I want the best of both worlds for my prototype; the ability to use ER32 collets directly in the milling spindle but also the ability to have preset tool holders that I can swap out quickly with an ATC.
So the tool holder below is what I came up with. It's the same size as an ER32 collet/BT30 holder, but the taper is 16 degrees for the collet system. This means I can fit both in the spindle. I would have gone with a BT30 taper, but the tool holders cost in the range of 100 dollars each, and I need a bunch of them. So for the prototype I'm making my own holders out of soft steel. If they don't work as expected, I can still use ER32s as a back up.
I'll make my tool holders with ER20 noses for tools up to 13mm.
32-8 toolholder er20.jpg
Hope this post wasn't too confusing. Let me know if I need to clear anything up, I'd like input on this subject because I'm pretty much making a new standard right here that may be important if/when this goes to production.


Posted: Mon Oct 30, 2017 8:30 pm
by Captain Starfish
Not too confusing at all. My cheapo chinese 24000rpm spindle has an ER-20 collet chuck and, like pretty much everyone else, I've thought about a compact ATC mechanism that doesn't cost a fortune. Having a tapered entry (for self alignment) "spanner" form that I could lock the collet nut into, then spin the spindle, seemed like a good way forward. No way was that going to work though, not if the nut was going to be tight enough to be effective.

I know with a pulley and reduction setup there'll be more torque for the spindle on yours. But consider: DIY mechanical setups out there at the moment often use a rattle gun on the drawbar. Not for fun or to be cool, but because they need to setup (and overcome) that much tension and stiction on the drawbar thread for sufficient clamping force to prevent pull-out.

If you can make it work, I will tip my hat to you. But I wouldn't make any grand plans or promises until you have a working prototype of just the ATC spindle because I seriously doubt (sadly) that it's going to work. Good luck though, be great if it does!

Looking at the first picture I have a little concern over head nodding based on the relatively thin cross section of the dark L form on the right just above where it clamps onto the light grey part. Just sayin ;)

I like where you're going with the tool holders. Worth considering something else, though. I'm not sure whether interchangeable collet vs ATC is really an advantage. If you wanted to rock hand changed tooling, you could just as easily have a couple of BT-30 tool holders with ER-20 collet holders. Whilst one was on the job, change out the next tool on the other and drop it in the rack. At the end of the current operation, swap out holders and do the tool height set then get on with it. Arbitrary tool changes without needing to stop the machine operation during the changes. Even if you had only one holder - you're not really any worse off doing manual changes in a BT-30 holder.

So, if it's just for that one spindle, here's the question: are there other advantages I'm missing that make it worth creating a whole new non-standard tooling setup? If you don't like the price of BT-30 holders (and you can get them cheaper, AFAIK, out of China) is it any harder to make a BT-30 in soft iron that it is to make your custom taper mongrel holder?

I'm not giving you grief, or not intending to anyway: sometimes we get an idea and chase it down the rabbit hole, and it takes fresh eyes to spot an issue with our crazy schemes!

But I really hope the spindle powered ATC plan works man, that would be freakin awesome.

Re: DIY Mill Turn Center

Posted: Mon Oct 30, 2017 10:57 pm
by Generic Default
The black L shaped part that connects the B axis rotary shaft to the milling head block only looks too thin from a cross-section view, since there is a circular cut for the motor. It's actually 1 inch thick of solid 6061 aluminum at the narrowest point, and 2 inches for the rest. It won't bend or break!

I'm familiar with the impact wrench drawbar method, and the pneumatic actuator/belleville method. I also considered the spanner-socket thing you're describing where the spindle nose screws directly into the collet nut with the collet and tool in it. Turns out you need tens of foot-pounds of torque to get a regular collet sufficiently tight, although that number can be reduced a lot with bearing nuts or DLC coated nuts. The friction between the threads and the collet angle is what drives up the required torque so high; most of the torque you put into tightening a collet is wasted on the friction.

30 taper spindles use between 600 and 1000 pounds of axial drawbar force in most cases. Since all of the industrial machines use a pneumatic cylinder and belleville washers, they have no real friction losses.

This method I'm doing will have the collet already tightened in the tool holder, and the drawbar only has to provide enough torque to suck the taper into the spindle. Since the drawbar on mine is already M12x1.75, it is compatible with BT30 tool holders, and it requires no pull stud!

At four foot pounds or torque, the M12x1.75 threads will put out between 450 and 1000 pounds of axial force in this system. The huge discrepancy is caused by the coefficient of friction. For now, I'll be using a plain bolt as a reference, which has a high COF and therefore lower axial force. If I get these drawbars made in quantity, you can bet that they will be stronger, harder, and smoother than regular bolts, and the entire screw will be PVD or CVD coated with an ultra low friction material like DLC. Backup plan is tungsten disulfide like I did with the Tri hotend.

By dropping the coefficient of friction below 0.1, the axial clamp force of the drawbar increases upwards of 900 pounds, which is plenty.

Even on the low end, a few foot-pounds of torque from the sensored BLDC spindle motor will give several hundred pounds of clamp force. Some links below if you want to verify ... teners.pdf ... e_calc.htm ... Calculator

And I checked out what you were saying about BT30 holders. Plenty of Chinese manufacturers selling in the 20-40 dollar range. I think for now I'll stick with my custom tool holders, but I'll likely just use the standard BT30s for when I do the production run. Mainly I was worried that my tool holders might not work out very well, which is why I wanted to be able to use ER32's as a backup plan. For industrial users, spending loads of money on tooling is no big deal. For hobbyists and home-shop types, it's unreasonable to expect people to buy expensive holders from 3rd party vendors. Right now my ATC can hold 8, 16, or up to 32 of these 30 taper tool holders, so the cost of each one needs to be low.

Re: DIY Mill Turn Center

Posted: Mon Oct 30, 2017 11:11 pm
by Captain Starfish
I was hoping it'd be something like that on the "thin" cross section :)

I'm seriously wishing you all the best with the ATC design, most reservations come from the primary thought "If it were that simple, why's everyone messing around with pull studs and claws and bellevue washers and pneumatics instead of just doing that?" If you can crack the hidden obstacles, power to you!

Kickstarter extra tier, perhaps, of ATC rack and 6 BT-30 => ER-20 holders?

Sorry can't remember now are they R5 main and sub spindles on the C axis behind the chucks? If so, you could later add one or two extra X carriages with bar feeders and some kind of collet clamp for a swiss style operation... ok shutup now starfish.

Re: DIY Mill Turn Center

Posted: Mon Oct 30, 2017 11:23 pm
by Generic Default
No I get it, I know the vast majority of ATC attempts on DIY machines end up as a failure or disappointment. I know that the method I'm choosing now isn't guaranteed to work, and I have a good idea of the factors that would make it fail. It's still a gamble at this point of whether or not it will work fully automated, but I know for sure that it will work if I stick a hex wrench in the top and crank it manually! ;)

The main turning spindles can fit a 4, 5, or 6 inch front mount chuck, and also a 6" back mount chuck. I already have a 4" ER40 fixture from last year's machine, luckily these things have standardized mounting dimensions so swapping is easy. It can also use 5c and 16c collets with the right nose fixture, I know 5c nose fixtures are fairly cheap on amazon and ebay. ... 1495809729

Oh look, they have a scroll mechanism just like a chuck. Makes it easier for me, no drawtube needed! Can't remember if I mentioned this a while back, but the milling spindle on my machine can hold a square key to actuate the scroll mechanism, so the machine can effectively tighten and loosen it's own workholding for automated production. At least that's the theory, I still have to put this thing together.

Main spindle has a 2 inch bore now, I turned it last Friday and I'll see if I can finish it tomorrow. Big bore for a machine this small.

Re: DIY Mill Turn Center

Posted: Mon Oct 30, 2017 11:29 pm
by Captain Starfish
Sorry, R5? I meant 5C. Derp.

This beast just keeps getting more and more fascinating. Keep up the great work man!

Re: DIY Mill Turn Center

Posted: Wed Nov 01, 2017 12:10 am
by Generic Default
Another progress update, it's movin along.

This is the harmonic gear mounting block, it's on the back side of the B axis. The flex-spline cup gets bolted to the bolt circle on the B axis shaft, while the rigid internal spline fits in the block. A motor aligned on the back will deform the flex-spline cup to rotate the B axis shaft. I printed out a test model over a year ago, and it seemed to work pretty well.
harmonic mount backside.jpg
harmonic mount backside 2.jpg

Also here's the milling head mostly finished on top of the main headstock for size reference.
milling head size reference.jpg

Re: DIY Mill Turn Center

Posted: Sat Nov 04, 2017 3:40 pm
by Generic Default
Spindle, faceplate, and 6 inch chuck finished and mostly assembled.
spindle face plate iso.jpg
spindle face plate back.jpg
spindle chuck iso view.jpg
spindle chuck side view.jpg
The back side of the spindle tube has 9 round keys, 3 are screwed up because of an indexer fail. The other 6 seemed to machine fine, they will take the torque with dowel pins as round keys from the pulley/worm gear combo.

This spindle is about the same size as what is used on most 40 taper VMCs, so it should be very rigid once installed. 2" through bore with 65mm ID bearings. Compared to most manual lathes like the grizzly ones, this spindle has larger flange face contact with bolts rather than threads or camlocks.

Re: DIY Mill Turn Center

Posted: Sat Nov 04, 2017 7:04 pm
by Captain Starfish
Looks good and meaty. Guess it will always be a tradeoff between convenience / speed of switching chucks vs strength and runout.

I kinda like my SEIG's screw on chucks just because it doesn't become a major operation every time I change over. I do change frequently, from 3 to 4 jaw and less often to centre and dog or to faceplate. I guess three cup heads ain't that big of a deal - as long as I can reach them with my allen key without having to remove the spindle! :)

Re: DIY Mill Turn Center

Posted: Sat Nov 04, 2017 7:16 pm
by Generic Default
Yeah, it would be nice to have a quick change system. Can't do threads because the mill-turn machine can put torque on the chuck in both forward and reverse, so it would unscrew itself and fall off. Camlock systems are pretty convenient but there's no way to adjust runout, and they sacrifice some rigidity for the mechanism gap.

Having 3 screws from the back or front of the collet nose / chuck allows runout adjustment, and it's the most rigid.

The backplate I have on mine allows mount 4, 5, or 6 inch standard chucks/collet noses. 6 inch from the front or back, 4 and 5 inch systems from the front only since there is no way to reach the screws.

If you go back and look at the headstock block, you see the big circular slots on the front side. These allow you to easily reach the back-mount M8 screws that fasten the chuck into place.

The 6" chuck I bought was less than 100 dollars including shipping. Hope there's no catch......

Re: DIY Mill Turn Center

Posted: Sat Nov 04, 2017 7:25 pm
by Captain Starfish
As always, you've thought it through. And yeah, compromises :-/ Like I said 3 screws isn't a major PITA though and it keeps things simple, rigid and good.

Looking forward to reading the magic runout numbers when it's built.

Re: DIY Mill Turn Center

Posted: Mon Nov 06, 2017 2:31 am
by Generic Default
A few more pictures to show how it's coming together. Still have quite a bit of work to do but most of it is done.

The turret tool blocks are for 3/8x3/8 stick tools for now, I'll get a few variants in the next few days.
turret with 4 tools nov 6.jpg
mill turn iso left nov 6.jpg
mill turn iso back left nov 6.jpg
mill turn iso top right nov 6.jpg

Re: DIY Mill Turn Center

Posted: Tue Nov 07, 2017 11:58 pm
by Generic Default
Here's what the tool holders look like. Two of these are aluminum, just wanted to test my program before running the steel one out of hex bar. ER20 nose on these, so it can't hold tools with a shank larger than 1/2". I'll run a bunch more later on, I still have to make the milling spindle so I can test the taper to taper fit of these.
toolholder protos 1.jpg
toolholder protos 2.jpg
toolholder protos 3.jpg

I'll definitely go with a standard BT30 compatible spindle in production but for now I need to have ER32 as a backup in case these tool holders suck.