What kit do you need to run ATC spindles?

[ger21]

It's not that simple, for a variety of reasons.

1) How do you mount the toolholders in a rack? They don't have the notch that ATC toolholders have.

2) How would the spindle pick them up. A normal ATC toolholder is tapered, so you have a little room as the spindle drops over the tool. With a straight shank, you have zero clearance, and a few thousandsth of misalignment would cause serious damage. I guess you could grind a chanfer on the shank, to help with this.

3) Those tool holders are big and heavy. If you have ANY runout in your spindle, or in the toolholder, you'll get some serious vibration. We will occasionally get a small wood sliver caught between the flange on our spindle and the HSK tool holder. I'm talking about .002" thick. It'll throw off the balance so much that it'll sound like the machine is going to go bouncing across the room. And the machine is 25ft long, and weighs about 10,000 lbs.

Also, any runout will be magnified by the increased length.

4) How do you get a powered wrench to grap the nut. They use spanner wrenches, so you can't just use a socket or something similar.

With no sensors, you never know if the tool holder is held properly, which could be incredibly dangerous when the spindle starts spinning, and 2 lbs of steel with razor sharp edges comes flying out.


There have been various attempts by companies to make "inexpensive" ATC spindles, or ATC adapters. At the end of the day, they end up being not that inexpensive, and they don't work all that well either.

[Goemon]

It's not that simple, for a variety of reasons.

1) How do you mount the toolholders in a rack? They don't have the notch that ATC toolholders have.

2) How would the spindle pick them up. A normal ATC toolholder is tapered, so you have a little room as the spindle drops over the tool. With a straight shank, you have zero clearance, and a few thousandsth of misalignment would cause serious damage. I guess you could grind a chanfer on the shank, to help with this.

3) Those tool holders are big and heavy. If you have ANY runout in your spindle, or in the toolholder, you'll get some serious vibration. We will occasionally get a small wood sliver caught between the flange on our spindle and the HSK tool holder. I'm talking about .002" thick. It'll throw off the balance so much that it'll sound like the machine is going to go bouncing across the room. And the machine is 25ft long, and weighs about 10,000 lbs.

Also, any runout will be magnified by the increased length.

4) How do you get a powered wrench to grap the nut. They use spanner wrenches, so you can't just use a socket or something similar.

With no sensors, you never know if the tool holder is held properly, which could be incredibly dangerous when the spindle starts spinning, and 2 lbs of steel with razor sharp edges comes flying out.


There have been various attempts by companies to make "inexpensive" ATC spindles, or ATC adapters. At the end of the day, they end up being not that inexpensive, and they don't work all that well either.

They are certainly all challenges to be overcome. Off the top of my head though:

1: I wouldn't grip the toolholder in a claw like they do for tapered toolholders. I would have each tool holder in a precisely cast hole with a smaller hole in the center for the endmill. I would use the end mill to ensure that each toolholder was positioned properly on the tool cart. Each endmill size would have a specific hole.

2: The set-up would need to be capable of a certain level of accuracy and repeatability in returning to precise co-ordinates. I would place each toolholder in it's respective space using the spindle so I could record the coordinates. I would then cast the toolholder hole around it. If the spindle can not return to the specific coordinates it clearly won't work.

3: I can't disagree on that. It should probably be attempted with smaller toolholders and matched to the spindle. I.e. ER25 toolholders for ER25 spindles. I would also expect it to work better with short stub toolholders and only with good brand stuff. As for increasing run-out, that is an issue with any toolholder regardless if it's ATC or MTC. I am going to assume that anyone trying to do this on the cheap has to set expectations of precision to be in-line with their budget.

4: I have seen hex shape ER collet nuts which can be gripped with a regular ratchet. I thought about a belt drive but the hex nut with a direct grip seems like it would be more precise. In any scenario, there needs to be a method of ensuring that the nuts are properly tightened with enough torque and not too much. My assumption is that, if this can be solved with cheap power drawbars, it can be solved here.

If a digital torque measurement is too hard to figure out, my guess is that it might work by calculating how much pressure is needed to achieve the ideal torque and then delivering that specific pressure using an adjustable regulator. With a little trial and error with a torque wrench, it should be possible to figure out if it takes 30psi or 50psi to achieve ideal torque.

I am sure that some of the guys here could figure this stuff out (not me obviously). I saw that one guy came close without another ATC project but, imo, his design looked too complicated to be a viable diy method.

[ger21]

As for increasing run-out, that is an issue with any toolholder regardless if it's ATC or MTC.

No, it's not. With an ATC spindle, the tapered toolholder is inserted into the spindle, where there is zero runout.
You're putting a toolholder into a collet, which is where the may be introduced.

On our ATC spindle, With the toolholder in the spindle, the collet nut is less than 2" from the spindle body. What you are talking about is one collet nut ±2" from the spindle body, with another one 1-1/2-2" further down.

[Goemon]

No, it's not. With an ATC spindle, the tapered toolholder is inserted into the spindle, where there is zero runout.
You're putting a toolholder into a collet, which is where the may be introduced.

On our ATC spindle, With the toolholder in the spindle, the collet nut is less than 2" from the spindle body. What you are talking about is one collet nut ±2" from the spindle body, with another one 1-1/2-2" further down.

I see what you are saying. That's one of the things people complain about with R8 toolholders too as they protrude further down than BT30 or cat toolholders. I can't think of many ways around it though. I guess they could not use tool holders and have the spindle pick up end mills in collets instead (with a seperate collet nut for each one).

The shopbot aftermarket ATC solution for those Kress routers might be more viable. They just screw an HSK20 cone over the flange. I'm not sure it solves the problem you mentioned though. It still pushes the tool further down.

Either way, the real question is if the additional runout is going to be noticeable or problematic on your average desktop cheap hobby level machine cutting soft wood. Some of those Chinese ebay machines look like a gust of wind would effect runout...

We need someone with a decent lathe to make us a decent cone adapter.

[Goemon]

This looks like another option. For $150 these ATC control modules allow software like Mach3 to operate ATC functions like solenoid valves or relays.

Control Modules

It looks like something that could be found for 1/10 of the price if you remove "CNC" from the google search though. hard to believe these guys are capable of setting prices that are worth paying.

- - - Updated - - -

This looks like another option. For $150 these ATC control modules allow software like Mach3 to operate ATC functions like solenoid valves or relays.

Control Modules

It looks like something that could be found for 1/10 of the price if you remove "CNC" from the google search though. hard to believe these guys are capable of setting prices that are worth paying.

[ger21]

Get some relays like this:
https://www.amazon.com/SMAKN-SRD-24V...4v+relay+board

And some solenoids like these.
https://www.amazon.com/Baomain-Pneum...s=24V+solenoid

I'm using one of those relay boards to control the contactors on my machine.

[Goemon]

Get some relays like this:
https://www.amazon.com/SMAKN-SRD-24V...4v+relay+board

And some solenoids like these.
https://www.amazon.com/Baomain-Pneum...s=24V+solenoid

I'm using one of those relay boards to control the contactors on my machine.

I knew there would be a cheaper way if the parts didn't have "CNC" in the title.

BTW, did you see that "XATC" diy ATC method designed by that German guy for regular ER collet spindles? It's very clever. It's look so simple and yet still somehow well beyond my skills to ever build one. He uses a motorized carrousel and the spindles own motor to tighten and release the collet nut.

It looks like he has made the cad / cam files available for free download so technically, anyone with a CNC machine should be able to make the parts.

[ger21]

He actually uses the toolchanger assembly to tighten and loosen the nut.


And this only changes bit, it doesn't change the collet, so all bits need to be the same size.

I knew there would be a cheaper way if the parts didn't have "CNC" in the title.

That "controller" is an assembly of multiple parts, that need to be designed, sourced, and assembled. I don't think they are making a killing off of them. Businesses are in the business to make money. You can pay someone else to do all the work for you, or spend countless hours and do it yourself.

There isn't, and likely never will be a quality, affordable high speed ATC spindle. At least not less than $2000. It's much more complicated than you think. People have been trying to come up with one years now.

[arturod]

If you need a script, you can use this one for UCCNC:

https://cnc4pc.com/forum/projects/to...acro-for-uccnc

Arturo Duncan
http://cnc4pc.com

[Goemon]

He actually uses the toolchanger assembly to tighten and loosen the nut.


And this only changes bit, it doesn't change the collet, so all bits need to be the same size.





That "controller" is an assembly of multiple parts, that need to be designed, sourced, and assembled. I don't think they are making a killing off of them. Businesses are in the business to make money. You can pay someone else to do all the work for you, or spend countless hours and do it yourself.

There isn't, and likely never will be a quality, affordable high speed ATC spindle. At least not less than $2000. It's much more complicated than you think. People have been trying to come up with one years now.

That's a fairly safe assumption because you can't get a "quality" MTC spindle for much less than $2,000. Of course "quality" can mean different things depending on your expectations. The Shopbot aftermarket ATC is $700. It's no Colombo or Hiteco but it probably works for a hobby-level user.

https://stepcraft.us/product/automatic-tool-changer/

It's not a super unreasonable expectation that something similar might be made available to mount on a regular ER20 or ER25 spindle instead of being specific to that one router model.


There is also those eBay sellers offering those BT30 ATC spindles and some offer bearing upgrades to reach 12,000rpm. Again, it probably wouldn't meet your standards for "quality" but.... it's not out of the question that someone will come up with a cheaper ATC solution with quality levels similar to your average Chinese spindle. We'll see.

I didn't know that the XATC guy was trying to make a business out of it. I thought it was just a diy project for himself that he shared in case others wanted to copy it. It's not super useful if it only allows one shank size though. I guess you would need to use toolholders so you could use a broad range of tool sizes with a common shank size to grab.

You are 100% correct. I don't have much understanding of the difficulties involved in building an ATC solution. I see cheap aftermarket ATC upgrades for R8 and BT30 spindles and I don't understand why they couldn't be adapted to grip and turn an custom collet nut instead of a drawbar.

If it was truly easy then I have no doubt that people here would already be doing it though. So... I am assuming that it isn't easy at all.

My first ATC machine will most likely involve a used ATC spindle with a bunch of cheap tool forks mounted on a T-slot extrusion at the edge of the work space. Avoiding the need for precision indexing of a tool cart or carousel seems key to making an affordible ATC solution.

[ger21]

The Shopbot aftermarket ATC is $700. It's no Colombo or Hiteco but it probably works for a hobby-level user.

I guess you meant Stepcraft, not Shopbot? Anyway, I think that is made by these guys.

Werkzeugwechsler f

It's not a super unreasonable expectation that something similar might be made available to mount on a regular ER20 or ER25 spindle instead of being specific to that one router model.

USOVO.de


Blurry customs made one for a while, but I think it had issues, as they've released various versions over the years, and none have become popular.

Avoiding the need for precision indexing of a tool cart or carousel seems key to making an affordible ATC solution.

Carousels are actually somewhat rare with Router ATC's. A rack type is cheaper, and in many cases faster than a carousel. The main cost is in the spindle and tool holders.