[R.DesJardin]

Per a legitimate request from PappaBear10, let's throw in our ideas and comments here...........................

[thkoutsidthebox]

ATC's are certainly very cool, but I don't really know anything about them! Is the collet held into the spindle using vacuum, magnetic, or does it clamp in place, and if so, how is it released when time for a tool change?

A DIY ATC would be nice to add to my build! lol

[philbur]

Hi Gary, if you are there.

Here is Machinery's Handbook take on wedge forces. The equations are derived from the classical equation for mechanical advantage, which states that the work performed by the applied force is equal to the work performed by the produced force, or something like that. As you know work done is equal to the force times the distance moved by the force.

Have fun
Phil

[justgary]

Phil -

Thanks. I'll chew on that for a while. What I'm still not sure about is the force on a cone; your sheet shows the 2D case of a wedge, but not the 3D cone...

I have spent a little time surfing for R8 collet forces but have not yet found anything fully satisfying. I'll keep looking as I have time. I'm still thinking about just measuring mine, since I have had tooling slip and didn't like it at all. With real numbers, I could be more confident in any given setup.

Thanks also for posting it over here so we don't get in *trouble* for clogging up that other thread.

Regards,

- Just Gary

[philbur]

I also struggled with the 3D aspect. One possibility is to consider the cone rolled out into a 2D equivalent. The other is the fact that the first principle equation for mechanical advantage doesn't care about 3D.

I think at the end of the day these simple equations are only going to give you some insight into what is going on. Even the range of coefficient of friction (0.15 to 0.7) makes the calc very approximate. Then you have the multitude of different forces acting on the cutter. For example: the simplest procedure for removing a wedge is to wiggle it from side to side while pulling on it, which is exactly what the cutter is trying to do.

So I would agree, if you need real numbers, perform a test. At least some of the uncertainties are remove.

I think a direct acting air cylinder is always going to leave you at the low end on dawbar tension. So a lever or cam arrangement, or a geared motor, may be a better solution. Also if you go for big drawbar forces you need to consider if you need to move the reaction forces from the spindle bearings to the spindle shaft.

Finally I think a decent powered drawbar is of much greater interest to the average user than an ATC. An ATC with open loop CNC could be a very scary proposition unless it is extremely well executed. Imagine for example the consequences of losing Z-axis steps, due to say a blunt or broken drill bit, while you are at the pub.

Phil

Phil -

Thanks. I'll chew on that for a while. What I'm still not sure about is the force on a cone; your sheet shows the 2D case of a wedge, but not the 3D cone...

I have spent a little time surfing for R8 collet forces but have not yet found anything fully satisfying. I'll keep looking as I have time. I'm still thinking about just measuring mine, since I have had tooling slip and didn't like it at all. With real numbers, I could be more confident in any given setup.

Thanks also for posting it over here so we don't get in *trouble* for clogging up that other thread.

Regards,

- Just Gary

[R.DesJardin]

I plan on getting to disassemble the 5C collet closer soon. It uses a camming action to lock. In either postion there is no load on the spindle bearings, I really think this will be able to convert to the Tormach. It is adjustable to the desired tension on the material or on the tool holder in this new application.

[justgary]

Phil -

The open loop part can easily be solved by re-referencing at every tool change (or referencing to the change position, making it closed-loop). A few more seconds, yes, but much more likely to hit the target every time. I'll allow a few more seconds if I know it will work every time. Especially if I can be at the pub!

The more I think about a geared motor, the more I like it. At 5:1 reduction to a drive motor, we're only talking 4 ft-lbs to tighten the drawbar to 20 ft-lbs. If that is the upper limit (which we still don't know for sure, but somebody had mentioned that number as coming from Tormach), then we're not talking about a very big motor, and indeed it could even be a stepper.

The good part about numbers is that you can at least get a notion for the upper and lower limits of that which you seek. Whatever the numbers, I don't want slipping tooling.

Regards,

- Just Gary

[apeman88]

I've been wanting a power drawbar for a long time. I looked for one but none are available so last summer... decided to make one but due to time available, I have not touched it since end of last summer. Below link is to my design... it uses a lever to mulitple the force at the draw bar. I'm only using a 3 1/2" air cylinder that is rated for around 850lb but I can get double that with the lever. The biggest problem is the limited space available. 3 1/2" is the largest that will fit in the motor cage and still allow the motor belt tension to be adjusted. I have mounted it on the Tormach and tried it... but my mount points are at the 45 degree angree on each side of the motor mount plate and I need to add one or 2 more at the end. The way it sits right now, when the power drawbar is activated, it will tilt the entire air cylinder unit forward.

http://www.cnczone.com/forums/showthread.php?p=504052

Another problem as many of you have mentioned is the force needed at the drawbar. I am not an engineer so I did it by trial and error. I had a 1200lb spring washer in there and if I tighten it to 1/2 compression... it will cut with 1/2" Endmill with depth of 1/4"... I did not try any more than that. I think, and don't hold me to this number, around 1500 to 1800lb washers should do the trick. Seeing the recent thread... I got my intereste back and hope to have some time this weekend to start playing with it again. I bought a bunch of spring washers and will first try to see if I can use a 1800lb washer to cut 1/2" depth using 1/2" Endmill. This is probabily the max I will ever cut and if it passes this test... at least for me, I'll be happy.

If someone comes up with one that will work... I'll still buy it... I have very limited time to work on projects as it is and even less time to get my equipment working so if I can spend a little extra... I rather buy it.

[philbur]

Some photos of the important bits would be most appreciated.

Thanks
Phil

I plan on getting to disassemble the 5C collet closer soon. It uses a camming action to lock. In either postion there is no load on the spindle bearings, I really think this will be able to convert to the Tormach. It is adjustable to the desired tension on the material or on the tool holder in this new application.

[philbur]

You would have to re-reference all three axis each time. Is this easy/safe to do within a running program.

To be able to hold the tool holder during a "very" heavy cut you need the drawbar to be as tight as is safe for the drawbar. Which I believe is a lot higher than 20 ft-lbs. Wasn't it up at about 65 ft.-lbs or something like 6,000 to 9,000 lbs tension.

A simple cam operated large/small hydraulic piston might also be a way to achieve force multiplication.

Phil

Phil -

The open loop part can easily be solved by re-referencing at every tool change (or referencing to the change position, making it closed-loop). A few more seconds, yes, but much more likely to hit the target every time. I'll allow a few more seconds if I know it will work every time. Especially if I can be at the pub!

The more I think about a geared motor, the more I like it. At 5:1 reduction to a drive motor, we're only talking 4 ft-lbs to tighten the drawbar to 20 ft-lbs. If that is the upper limit (which we still don't know for sure, but somebody had mentioned that number as coming from Tormach), then we're not talking about a very big motor, and indeed it could even be a stepper.

The good part about numbers is that you can at least get a notion for the upper and lower limits of that which you seek. Whatever the numbers, I don't want slipping tooling.

Regards,

- Just Gary

[R.DesJardin]

We just bought a used Hansvedt EDM at work with a tool changer, it only hold 6 tools but is a very neat looking setup. It uses a double row chain with forks on it to hold the tools. There are ball detents to keep the tools in the holders and it looks like it rotates and extends using a worm drive an a eccentric camming opetation on a linear slide. I'll bet there are pictures on the net of it some where.

[JoeyGH]

This is a solid model of the ATC I been drawing up in my spare time to work with the Tormach PCNC1100 without any drilling on the mill. Most of the components will be made from aluminum. Just looking for feed back and what I could possibly redesign to make it better. I have seen Popabears design and think he's doing a great job and I read the one post about his costs and him saying if someone thought they could make it cheaper to try and design one themselves so I decided that was a good challenge. I have not yet designed the drawbar but I will be soon. Thanks in advance for any comments

[flick]

Nice re-use of the machine arm holes. I don't have a machine arm, so it wouldn't bother me, but some may object. It also seems that your design will not allow the machine to be placed as close to walls as the forward facing electronics cabinet otherwise allows. Perhaps re-orient the motor that sticks way out to the left (up? down? horizontal, but perpendicular?) and use a belt drive instead of a screw?

EDIT: Woops, just caught my mistake... that's not a motor, it's a cylinder. Mmmm... foots not so bad w/ a little catch-up.

[benji2505]

Why do you design it with two motions (rotation at machine frame and linear to insert tool)? Would it not be easier and sturdier to have a solid ATC on the side and reduce it to one linear motion? The rotation might make things unnecessary complicated and slow. And it is a source for failure, if the arm is not turned away in time you have a nasty crash.

Benji

[JoeyGH]

The only two motions that occur are the toolchanger can slide in and out using the pnuematic cylinder and the stepper motor that indexs the coursel I designed that swivel thing which is a 1.5'' inch square bar of aluminum and swivels both where it mounts to the column and where it mounts to the back of the slide. The purpose of that being that to adjust the whole ATC so it's directly under the collet. After you get that finished you'll have to bolt the swivel piece tight which will then no longer swivel and sustain it's position. But you need two things of motion one to rotate the tool coursel and one to pull the coursel out and in from under the spindle. Hope that clears it up a little. I'll be designing the draw bar this weekend and have it posted in this coming week.

[JoeyGH]

Also to respond to flicks concerns. The pnuematic cylinder doesn't stick out beyound the back of the main column so you'll will still be able to put the machine as close to the wall as you normally would. Thanks for the responses.

Joey

[JoeyGH]

Another concern I had myself is about how big of a cylinder would I need to open the collet. After having some bevel washers to lock the collet and still provide the crushing ability to open the collet. I was thinking of pushing directly on the R8 drawbar. Possibly a 4''-5'' bore pnuematic cylinder. If I use the lever idea with probably at most a 2 times advantage(due to the small amount of room left up in the head) what size cylinder would that get me down to a 3'' bore cylinder? Has any on here had experience with modbus or something similiar on Mach3. I have a new screen set already made for my toolchanger but I need more inputs and outputs from mach to add safety sensors to the changer to prevent crashes. Most of the inputs and outputs are already being utilized by the machine itself. Especially if you have the 4 axis option already on your PCNC 1100. Thanks in advance

[Guldberg]

I thought i just wanted to share my idea regarding tools for an ATC. I know some people have made their own tool holders, but i would really like to have the flexibility of ER collets without paying 92$ each. Therefore i looked at straight shank ER20 collets which can be bougth at http://www.ctctools.biz/servlet/the-...CK-20MM/Detail for only 12$ . Normally i would expect this to be some chinese garbage steel, but ive bought one before and was amazed by the quality. The idea is to turn a ring similar to Tormachs conversion kit and glue it on using epoxy.

Anybody can see why this shouldnt work? I know its a 20mm shaft, but for those who want to use the original Tormach collet there is a 3/4" version i think

[BlueFin]

Another concern I had myself is about how big of a cylinder would I need to open the collet. After having some bevel washers to lock the collet and still provide the crushing ability to open the collet. I was thinking of pushing directly on the R8 drawbar. Possibly a 4''-5'' bore pnuematic cylinder. If I use the lever idea with probably at most a 2 times advantage(due to the small amount of room left up in the head) what size cylinder would that get me down to a 3'' bore cylinder? Has any on here had experience with modbus or something similiar on Mach3. I have a new screen set already made for my toolchanger but I need more inputs and outputs from mach to add safety sensors to the changer to prevent crashes. Most of the inputs and outputs are already being utilized by the machine itself. Especially if you have the 4 axis option already on your PCNC 1100. Thanks in advance

That falls back to the original issue of pushing on your spindle bearings to release the collet. There has been talk of needing between 3,000 and 6,000 Lbs of force needed to hold the tool in place. If it turns out to be the higher number then you will be setting a full-size 1/2 ton pickup truck (Chevy or Ford) on top of your spindle bearings every time you change a tool (20- 50 times a day)

[Flat32]

A friend of mine runs a CNC Bridgeport without ATC and has been using an Ultron Snap Change tooling system. The tool holders seat directly into the R8 spindle and are retained by a gripper drawbar similar system to what modern CNC machines use. The drawbar is tightened by a toggle mechanism at the top that is "activated" by the spindle position. Fully up releases the tool while 1" downward grips the tool. No air cylinder and no power drawbar.

http://www.snapchange.com/default.htm

I bought one of these and was going to adapt it to my Tormach, but the Tormach TTS works very well for manual tool changing and I dropped the project. It could be viable for an ATC.

Flat32