[Goemon]

I came across this while searching for an entirely unrelated part. I thought I'd share it as I know that diy ATC mods are of interest to some people here.

This is apparently a screw-on ATC adapter to turn a regular ER11 MTC China spindle into an SK15 ATC spindle:

USOVO.de

I knew they had these for Kress routers but I haven't seen them for other spindles before. The adapter and tool cart look affordable - espiecially if you already have the gear for the pneumatics.

I wonder how well they work...

[1stPrototype]

I bought one and mounted it on 1.5kw Chinese spindle. It's not plug and play, and the documentation is sparse. The unit feels very well constructed and precise, but it has some critical gotchas.

First off, you have to get your spindle taken apart at the top, just to access the pin spanner holes at the top of the shaft. Without that, there is no way to immobilize the spindle to thread the changer on -- and nobody tells you this before you buy it. The Chinese spindles have no spindle lock built-in like the Kress routers that the ATC was originally designed for so they require surgery. With water-cooled spindles, extra care needs to be taken when reassembling the top portion, as that is where the water/electrical barrier is crucial. There are two tiny o-rings counterbored into one side of the mating faces, and they are covered with a slopass dollop of some kind of sealant. In my case, one remained glued to each of the separated faces, and I didn't even notice them the first time I took it apart and reassembled it -- I got lucky, because on subsequent teardowns, one o-ring popped out and rolled across the bench.

The second gotcha - a probable design flaw in the ATC unit. There is a cylindrical collar that screws over the tapered shaft that mates to the Chinese spindle. That collar is about 30 mm in diameter, and is held in place (at 24000 rpm) by about 3 threads, engaging the bearing preloader beneath it only about 2mm. During my first attempted cut, the collar separated from the preloader and that caused the top and bottom air seals to self-destruct. The manufacturer seems to not believe this, and tried to convince me it was my fault for ejecting the tool while spinning...but that did not happen. My theory is that aggressive spindle braking worked the collar loose and exposed its insufficient thread engagement flaw.


The replacement seal kit is about $10 and the shipping is 20 -- for a 40mm x 1mm and a 30mm x 1mm silicone washer. I sourced some 1mm silicone sheet and machined a set of punch dies to cut my own seals -- since I don't believe these seals can last very long as designed. The top seal, in particular, is subject to constant friction, as its inner diameter rides in a groove machined into the spinning collar while its outer diameter is held down by a pinch plate and o-ring. The collar is now held in place with a fair application of Loctite, and we'll see how it holds up. So far, so good in dry runs...not cutting.

I've actually made some modifications to the Chinese spindle to mimic a spindle lock, so now I can install and remove the ATC without giving the spindle another lobotomy. I'll make some test cuts later this week and share anything that is interesting.

[ScotiaVisuals]

I bought one and mounted it on 1.5kw Chinese spindle. It's not plug and play, and the documentation is sparse. The unit feels very well constructed and precise, but it has some critical gotchas.

First off, you have to get your spindle taken apart at the top, just to access the pin spanner holes at the top of the shaft. Without that, there is no way to immobilize the spindle to thread the changer on -- and nobody tells you this before you buy it. The Chinese spindles have no spindle lock built-in like the Kress routers that the ATC was originally designed for so they require surgery. With water-cooled spindles, extra care needs to be taken when reassembling the top portion, as that is where the water/electrical barrier is crucial. There are two tiny o-rings counterbored into one side of the mating faces, and they are covered with a slopass dollop of some kind of sealant. In my case, one remained glued to each of the separated faces, and I didn't even notice them the first time I took it apart and reassembled it -- I got lucky, because on subsequent teardowns, one o-ring popped out and rolled across the bench.

The second gotcha - a probable design flaw in the ATC unit. There is a cylindrical collar that screws over the tapered shaft that mates to the Chinese spindle. That collar is about 30 mm in diameter, and is held in place (at 24000 rpm) by about 3 threads, engaging the bearing preloader beneath it only about 2mm. During my first attempted cut, the collar separated from the preloader and that caused the top and bottom air seals to self-destruct. The manufacturer seems to not believe this, and tried to convince me it was my fault for ejecting the tool while spinning...but that did not happen. My theory is that aggressive spindle braking worked the collar loose and exposed its insufficient thread engagement flaw.


The replacement seal kit is about $10 and the shipping is 20 -- for a 40mm x 1mm and a 30mm x 1mm silicone washer. I sourced some 1mm silicone sheet and machined a set of punch dies to cut my own seals -- since I don't believe these seals can last very long as designed. The top seal, in particular, is subject to constant friction, as its inner diameter rides in a groove machined into the spinning collar while its outer diameter is held down by a pinch plate and o-ring. The collar is now held in place with a fair application of Loctite, and we'll see how it holds up. So far, so good in dry runs...not cutting.

I've actually made some modifications to the Chinese spindle to mimic a spindle lock, so now I can install and remove the ATC without giving the spindle another lobotomy. I'll make some test cuts later this week and share anything that is interesting.

Hi,

I have the same or similar ATC and a 1.5k water cooled spindle. I've had bearing issues with the ATC and I'm trying to get the ATC removed from the spindle (again) but it's jammed on this time. Can you advise what modifications you managed to make to your spindle to be able to lock it, and how effective was it?

Thanks SV.

[jkkmobile]

This one should be better:
https://www.sorotec.de/shop/Spindles...ange-adapters/

[1stPrototype]

That design immediately solves a problem with Usovo model -- it doesn't use the collet lock threads to engage the changer. If my guess is right, that diagram is depicting a hex profile shaft being held in the Chinese spindle, and that in turn engages with a hex mating hole in the changer. Far more elegant on every level. That would also erase the limitation of only running in CW direction. All that goodness adds about $500 more to the total price, but if you plan to use your system beyond hobby level, that is trivial. A better designed critical component like an ATC should not be fragile or limited to one direction.

[Goemon]

That design immediately solves a problem with Usovo model -- it doesn't use the collet lock threads to engage the changer. If my guess is right, that diagram is depicting a hex profile shaft being held in the Chinese spindle, and that in turn engages with a hex mating hole in the changer. Far more elegant on every level. That would also erase the limitation of only running in CW direction. All that goodness adds about $500 more to the total price, but if you plan to use your system beyond hobby level, that is trivial. A better designed critical component like an ATC should not be fragile or limited to one direction.

One thought.... that store is in Germany - not exactly the cheapest place to import goods from. I wonder if it is something that could be sourced directly from China (where it is probably made). If it was $300-$400, it feels less pointless (again, if it actually works properly). If it isn't made in China, I', sure a Chinese manufacturer will "borrow" that design soon enough.

[Goemon]

To be fair, if you buy a real ATC spindle, they aren't exactly "plug in and play" but, any product that requires users to mod their spindle themselves would be a no go for me. The fact that they don't warn you about that before you buy doesn't give me a lot of trust in the vendor....

I'd like to see an affordable aftermarket ATC design that focuses on opening and closing the existing collet nut without any internal mods. I noticed that Datron offers (or at least supports) direct ATC designs where only end mills are switched out instead of tool holders. That would make more sense for a low cost product as toolholders are expensive.

To open the collet nut, you would just need something to hold the shaft still and something similar to the power drawbars you see on eBay with a hex driver that is large enough for an ER20 or ER25 collet nut. Or.... a couple of motors with the right size gears....

With any aftermarket ATC design, I think there is a hard price limit above which the product becomes pointless. Once you get above $1,000 or $1500 for the whole package (spindle, ATC adapter and tool rack), most people would probably just invest in a real ATC spindle.

[ger21]

I noticed that Datron offers (or at least supports) direct ATC designs where only end mills are switched out instead of tool holders. That would make more sense for a low cost product as toolholders are expensive.

I bet that spindle costs at least $10K-$15K. And it's only available in small spindles (600w-2Kw), which makes sense as it probably can't clamp tight enough to handle higher power. I've never seen one of those before.

[Goemon]

I bet that spindle costs at least $10K-$15K. And it's only available in small spindles (600w-2Kw), which makes sense as it probably can't clamp tight enough to handle higher power. I've never seen one of those before.

I bet you are right. They don't list spindle prices on their site but I have seen those Jager and GMN spindles sell for upwards of $30,000 on ebay so god knows what Datron charge....

The ATC adapter I posted a link to is limited to 2.2kw spindles for the reasons you suggested but I think that covers most of the hobby market it is aimed at so I don't see it as a barrier for most people (although it was for me).

With that said, they make regular MTC spindles with far larger motors than that. If an ATC adapter used a regular collet nut with direct connect end mills, I can't see why one couldn't be made to work with more powerful spindles. It would just be a case of ensuring that enough torque was used for the size.

My spindle's manual has a table listing the torque requirement for each size of spindle and collet nut. It would be part of the design spec for the ATC device. Whether any of us would trust a Chinese made ATC adapter with a 10hp spindle or not is a different matter....

For now, I think the most cost effective route to ATC for hobby use is still buying one of those Tormach spindle cartridges and one of the many cheap (but functional) power drawbars.

[ger21]

If an ATC adapter used a regular collet nut with direct connect end mills, I can't see why one couldn't be made to work with more powerful spindles. It would just be a case of ensuring that enough torque was used for the size.

You'll never see something like that, as there's just no market for it.

For now, I think the most cost effective route to ATC for hobby use is still buying one of those Tormach spindle cartridges and one of the many cheap (but functional) power drawbars.

That's not really an option for router users.

[Goemon]

You'll never see something like that, as there's just no market for it.




That's not really an option for router users.

do you mean because the speed is limited to 10,000 rpm? I was thinking that could be somewhat fixed with a bearing upgrade. It's not ideal but it seems like all of the low cost ATC options have limitations. For me, it would be the most appealing option if I wanted to go the ATC route for hobby use. I use my router at 12,000 and 15,000 rpm fairly regularly.

I am going to respectfully disagree on the point about there being no market. I think there is as much of a potential market as there is for any hobby CNC products and, just like other hobby CNC products, I think it entirely depends on the price point. I don't think there is any market for a hobby-level ATC adapter at the price point some of these companies are charging now. However, If there was a product that was as affordable as the Chinese spindles and it worked properly.... and.... it was easy and intuitive to set up, it's an obvious upgrade. I'd buy one.

It's logical that an ATC adapter that is plug in and play will be more appealing to hobby users than any requiring internal spindle mods. One that doesn't require expensive toolholders is also more appealing (again, assuming it works).

I guess we will see.

[1stPrototype]

do you mean because the speed is limited to 10,000 rpm? I was thinking that could be somewhat fixed with a bearing upgrade. It's not ideal but it seems like all of the low cost ATC options have limitations. For me, it would be the most appealing option if I wanted to go the ATC route for hobby use. I use my router at 12,000 and 15,000 rpm fairly regularly.

I am going to respectfully disagree on the point about there being no market. I think there is as much of a potential market as there is for any hobby CNC products and, just like other hobby CNC products, I think it entirely depends on the price point. I don't think there is any market for a hobby-level ATC adapter at the price point some of these companies are charging now. However, If there was a product that was as affordable as the Chinese spindles and it worked properly.... and.... it was easy and intuitive to set up, it's an obvious upgrade. I'd buy one.

It's logical that an ATC adapter that is plug in and play will be more appealing to hobby users than any requiring internal spindle mods. One that doesn't require expensive toolholders is also more appealing (again, assuming it works).

I guess we will see.

Yeah...the SK15 tool holders are a cool $65 per copy. Total cost for a 5-tool setup is pushing a grand as is...

Sent from my SM-G900V using Tapatalk

[ger21]

I am going to respectfully disagree on the point about there being no market. I think there is as much of a potential market as there is for any hobby CNC products and, just like other hobby CNC products, I think it entirely depends on the price point.

No market because the price will never be what you want it to be. An ATC spindle requires a level of precision far greater than a $200 chinese spindle, which results in a far greater price. If you are doing production work, where the ATC spindle makes you money, then $2000-$3000 for a chines ATC spindle is cheap. Even if they could make one for $1000, it's still 5x more than a standard chinese spindle, and most hobbyists are not going to pay for it.

So yeah, there would be a market for a $400-$500 adapter, if it worked as well as a real ATC. IMO, that's highly unlikely. And your still looking at $1000 all said and done.
I'd be really surprised if you ever see an ATC spindle any cheaper than what you can find now from China, and I'd bet money that you'll never see an adapter for larger spindles.

do you mean because the speed is limited to 10,000 rpm?

That, and many other reasons. There's a reason, that I've never seen anybody use that type of spindle on a router.

[Goemon]

No market because the price will never be what you want it to be. An ATC spindle requires a level of precision far greater than a $200 chinese spindle, which results in a far greater price. If you are doing production work, where the ATC spindle makes you money, then $2000-$3000 for a chines ATC spindle is cheap. Even if they could make one for $1000, it's still 5x more than a standard chinese spindle, and most hobbyists are not going to pay for it.

So yeah, there would be a market for a $400-$500 adapter, if it worked as well as a real ATC. IMO, that's highly unlikely. And your still looking at $1000 all said and done.
I'd be really surprised if you ever see an ATC spindle any cheaper than what you can find now from China, and I'd bet money that you'll never see an adapter for larger spindles.



That, and many other reasons. There's a reason, that I've never seen anybody use that type of spindle on a router.

Haas uses a 40 taper cartridge spindle on their gantry router. There is no reason why not if an acceptable speed can be achieved. Plus, all the old overhead or pin routers use belt drive spindles. An ATC "router spindle" is just a BT30 cartridge spindle with a direct drive (or pin drive) high speed motor.

I think we agree on the ATC thing except that I am more optimistic that, one day, someone will figure out how to make a lower cost ATC adapter. They need a whole new design because very few will feel that it makes sense to invest $1,000 or even $600 in a $150 Chinese spindle. I can understand your lack of confidence in the manufacturers for ATC adapters though. Previous and current attempts haven't exactly "nailed it".

They need to stop comparing their pricing to the industrial market because hobby users aren't writing off the cost against profit. The value of a hobby ATC product is only in convenience.

They need a design that doesn't use pneumatics at all. It needs to use a (fairly cheap) servo or stepper motor with some gears and a mechanical method of stopping the spindle turning while the change happens.

[Goemon]

This is a clip of the Datron direct ATC spindle:

https://m.youtube.com/watch?v=b1_aB4CXmrI

[TTalma]

I really wanted an ATC for my machine. I looked and researched and tried just about everything for around 2 years. I made the tormach style changer (Mock up for feasibility) using a head stock from a lathe, a MT2 1/2" collet, and homemade draw bar, actually worked pretty well. Until I tried to run it past 5000 rpm. then the vibration was pretty scary. We build spindles at the company I work for (large ones, 120mm is the smallest). I showed my design to the engineers who design the spindles. They said it was defiantly do-able. The issue was that you would need to do high precision machining and balancing, which are expensive. To make a prototype I would be looking at a few thousand.

So it's not that it can't be done, it's just that to get what we need it's expensive. In the end I dropped around $2k for a used columbo, and a new Delta VFD. and another $300 on tool holders (Got lucky got 12 used for $300). I would do this again in a second. I see used ones come up on eBay every once in awhile, just need patience.

One other thing I looked at was using a router, and actually swapping out routers. This would be the cheapest method, but take the most space. You can buy porter cable routers on eBay for less than $100. Or make your Y axis extra long and put 2 or 3 routers on the same set of rails with different drives, for Y1, Y2, Y3 etc. I know routers aren't the best spindles, but changing brushes and bearings is actually pretty easy.

[shujaa]

HI,

My spindle is ER16, 1.5 kW. Can i use this to use as quick change

What i am thinking is that a hardened precise rod will be tightened in the existing ER16 collet chuck . Th other end with 2 flats will go into the back end of this fitting. The allen keys will be used as quick fit mounting. one of each will be used for a cutter and ready for allen keyeing.

will it work at high rpm ?
will it be a good idea for manual ATC ?
will it work at all?

Regards

[shujaa]

HI,

My spindle is ER16, 1.5 kW. Can i use this to use as quick change

What i am thinking is that a hardened precise rod will be tightened in the existing ER16 collet chuck . Th other end with 2 flats will go into the back end of this fitting. The allen keys will be used as quick fit mounting. one of each will be used for a cutter and ready for allen keyeing.

will it work at high rpm ?
will it be a good idea for manual ATC ?
will it work at all?

Regards

here are the photos of the 2 components to be joined with a 10mm rod.

[skrubol]

HI,

My spindle is ER16, 1.5 kW. Can i use this to use as quick change

What i am thinking is that a hardened precise rod will be tightened in the existing ER16 collet chuck . Th other end with 2 flats will go into the back end of this fitting. The allen keys will be used as quick fit mounting. one of each will be used for a cutter and ready for allen keyeing.

will it work at high rpm ?
will it be a good idea for manual ATC ?
will it work at all?

Regards

here are the photos of the 2 components to be joined with a 10mm rod.

Will it work at all? Sure.
Is it a good idea and will it work at high speed? No.
You'd be reducing rigidity quite a lot by hanging something out on the end of a 10mm rod, even if it's a short extension. Having a set-screw type connection is generally a no-go at higher RPM's, and definitely not something you want to have a whole collet chuck hanging off of.

[wizard]

There is no perfect approach to the spindle speed nor the tool change problem.

If you look at all of the solutions to either problem that have been applied to R8 (Bridgeport as an example) milling machines the problem of tool changes and spindle speeds has been solved in many ways.

One approach to high speed was to mount a high speed spindle to the side of the mills head. Another was to mount a sub spindle in the R8 spindle driven by a small motor. There are many others.

Like wise fast tool change has been accomplished in many ways over the years. Today one popular approach on R8 is the Tormach Tooling System.

In any event what I’m trying to get at is that one needs to consider what will work best for a specific need. For example multiple routers can be a good idea and frankly there are several approaches to this. You could swap out the whole router quick change style. You could mount two routers side by side. Or you could implement two Z axis.

I’m actually surprised that more users don’t go the two Z axis route. This is doable with routers or real spindles requiring only minor adjustments in machine size.

With respect to dual spindles on one Z axis this was a common solution to upping productivity all the way back to the early 1900’s when manual profile mills where common. It’s an idea worth considering for small scale manufacturing even today.

I really wanted an ATC for my machine. I looked and researched and tried just about everything for around 2 years. I made the tormach style changer (Mock up for feasibility) using a head stock from a lathe, a MT2 1/2" collet, and homemade draw bar, actually worked pretty well. Until I tried to run it past 5000 rpm. then the vibration was pretty scary. We build spindles at the company I work for (large ones, 120mm is the smallest). I showed my design to the engineers who design the spindles. They said it was defiantly do-able. The issue was that you would need to do high precision machining and balancing, which are expensive. To make a prototype I would be looking at a few thousand.

So it's not that it can't be done, it's just that to get what we need it's expensive. In the end I dropped around $2k for a used columbo, and a new Delta VFD. and another $300 on tool holders (Got lucky got 12 used for $300). I would do this again in a second. I see used ones come up on eBay every once in awhile, just need patience.

One other thing I looked at was using a router, and actually swapping out routers. This would be the cheapest method, but take the most space. You can buy porter cable routers on eBay for less than $100. Or make your Y axis extra long and put 2 or 3 routers on the same set of rails with different drives, for Y1, Y2, Y3 etc. I know routers aren't the best spindles, but changing brushes and bearings is actually pretty easy.