[webgeek]

Ok, anyone figured this out yet? Having an ATC is the last impediment to me achieving machining nerdvana - something I know you all lose sleep over . I just watched a video on the Tormach ATC and it is dead sexy but I'm not a huge fan of the belleville washer trick for a power drawbar. It SEEMS that the NM-200's slick drawbar design can do a better job. I'm thinking a small impact wrench or such that goes up and down. Khai told me a few months ago this was in the works but he also said to give him a ring if I figured out anything on my own which did not make me feel any better about it's chances

I've got a good bit of robotics experience (far more that than machining) and would love to collaborate with one of you experts on here. The initial goal is to make an effective power drawbar inexpensively for the NM-200 mill. If it works for other Novakon mills, the more power to us. Some requirements then:

1) Be cheap - we didn't buy Haas for a reason
2) Be easy to source parts - IE: no lucky finds required to make it work
3) Be 100% effective every time. If it ain't reliable, it ain't worth it.
4) Be totally digitally controlled (necessary for future ATC use)

To kick start the discussion, here is an electric drawbar I thought was pretty slick in it's simplicity:
http://www.cnczone.com/forums/bencht...o_drawbar.html

Here is a cheap and pretty solid electric impact wrench:
1/2" Electric Impact Wrench

This wrench in particular is of an interest to me because it's cheap enough that I'd gladly tear it apart to give it a vertical housing and lose the handle. By losing the plastic casing, it would be possible to use the already provided screw holes to attach it to a fixture. I imagine the "indents" used to recess the casing screws would serve double-duty as a way to restrain the tool to keep it from rotating. In short, it seems very doable, especially since we all have CNC machines at our disposal to mill up a fixture for it.

In case anyone is wondering, I'd rather not use an air impact wrench because I have enough crap in my tiny shop already and running a compressor just to mill is not ideal BUT with that said, this is clearly a working approach:
CNC Cookbook: Powered Drawbar for IH Mill
The thing to look at in that page is all of the tricks with the way to guide the socket down. Nice and simple solutions to a potential issue.

Another advantage to electric is that it's quite easy and cheap to control an electric system and a bit more of a hassle to computer-control a pneumatic system as good quality electronically activated air valves cost a lot more than relays. If the ultimate goal is total automation, I think we should stick with what's easiest.

Now I personally think the whole mechanism should "float" ~ 1/2 to 1" above the nut and assume about a 1/2 to 1" engagement on the drawbar to avoid stripping. The floating would be handled quite easily with simple springs on a pair of risers. This poses the biggest issue though. Sticking with the whole "electric" theme it means a few options - motors w/gears and the like, solenoids, or linear actuators. Motors with gears and such are likely a great choice but a pain in general for custom tasks like this. Linear actuators are incredibly strong but also generally dog slow. If you are willing to wait 10-20 seconds to release and re-engage they are great. Personally, I think the best approach is a solenoid. Sadly, I don't know where to source these cheaply so we have to work with something more reasonable like McMaster Carr. Part #7723K5 in particular seems reasonable with it's 21 lbs of force. The biggest issue with it is the depressing 1" throw but that's solvable.

Next we need the control electronics. I'm a whus here and would just stick with what I know. The EASY solution is an Arduino - the cheaper solution is a raw ATMega chip. Both would need to use some relays to control everything. The advantage to using a microcontroller is one of intelligence - this same controller can manage the ATC flawlessly too which gets away from all the nasty Mach 3 macros so many people need. Additionally it lets us do smart things in the future like detect if the drawbar didn't engage, etc. So let's assume Arduino but we can switch to the raw ATMega later to keep costs down if needed. This is a nice tiny one that is perfect for us:
Arduino Pro 328 - 5V/16MHz - SparkFun Electronics

And the relays:
Relay SPST-NO Sealed - 30A - SparkFun Electronics

We'd need some transistors, wire and the like but all too cheap to worry about for now. Additionally we need some sort of button to toggle engage/disengage and all that but we'll skip those too. So the pricing looks something like this:

Arduino - 19.95
relays - ~15
linear solenoid - 45.60
impact wrench - 49.95 (can get it for $40 with the monthly 20% off coupon)
Everything else - ~$70 (aluminum, wire, springs, etc)

Under $200-250 pretty reasonably. I think this is pretty realistic and very acceptable all told.

Here are some commercial options to see the range:
http://cgi.ebay.com/POWER-DRAWBAR-BR...4#ht_500wt_922
$135 w/shipping - this is still effectively a manual setup so not going to cut it for ATC but it is a good design and one we should "borrow"

http://cgi.ebay.com/Air-Power-Drawba...ht_1840wt_1139
$455 w/shipping - this appears to be very nicely done but it would need to be retrofitted with an electronic valve for ATC use. Still a clean and solid piece of work and if it fit my mill easily I'd very seriously consider it.

Any thoughts on this from the experts? Anyone willing to work with me on this? Thanks!

-Mike

[Hirudin]

I would really like something like this! I would also like to help but I don't really see how I could.

[webgeek]

Oops, fixed the auction links in my original post. It was this:
POWER DRAWBAR BRIDGEPORT MILLING MACHINE LIMITED TIME - eBay (item 120654916884 end time Jan-30-11 15:32:55 PST) and Air Power Drawbar MILL Bridgeport-Kurt-Maxi Torque A&T - eBay (item 300510113716 end time Jan-30-11 11:58:00 PST)

A little bit of R & D tonight and I've come up with a few things... The easiest way to make the ATC work with an Arduino driving it is likely to be a USB cable to the controlling computer and a Mach 3 plugin developed in C++. This is pretty doable and not too much of a problem, just a bit of work. The whole process would use Firmata (Main Page - Firmata) to tie it all together.

As for helping out, simply providing feedback as the idea and design progresses is more than enough to help out. Speaking of which, this is a view of the top of the mill with the motor removed (it would be towards the right.) The left most set of screws/holes seems a great way to secure the whole assembly to the top of the mill:
http://i920.photobucket.com/albums/a...v/IMG_1438.jpg

Thanks!

-Mike

[Hirudin]

OK, I am a total newbie (I've only had my mill for a couple weeks, only plugged in for a few days, and have only made one cut with it so far). I will try to help where I can...

$250 sounds reasonable to me. I'd gladly pay that much. If using something other than a solenoid reduces the cost that's great but I would say unless the alternative works better it's not really worth worrying about.

Electric impact wrench - I like that idea. That HF one is 120v. Using one that can be connected to the same power supply as the mill itself would be nice.

Arduino control - I also think this is a good idea. I might just be frivolous, but $20 doesn't sound like "too much" money to me. Again, to me, I'd only try to implement the cheaper control chip if it actually adds an advantage to the design other than cost.

A possible issue: stag53 (from your linked thread) says that he's thinking he would like to lock the spindle when using the impact wrench. Is this easily done with an NM-200?

[webgeek]

Cost is a slippery slope. It's easy to let the cost creep out of control quickly if you let it. In my case, it always seems to start when I price things out individually and not take into account the whole. I'm not actively trying to cut corners but I am trying to avoid the expensive options if I can. For instance, the Arduino is already the cheaper option vs. my initial choice of the Making Things controller (Welcome — MakingThings) which runs quite a bit more. The Arduino will be a minor amount more work but saves almost $100.

A possible issue: stag53 (from your linked thread) says that he's thinking he would like to lock the spindle when using the impact wrench. Is this easily done with an NM-200?

Hrm, good catch. There is no built-in spindle lock or break on the NM-200. I don't think it will be critical for the drawbar as you will be holding the spindle/tool holder when it runs. The ATC will need it though. We might end up needing to do something clever around this:
http://www.cnczone.com/forums/833330-post6.html

Thanks!

-Mike

[Hirudin]

Heh, I just realized that was a reply to a topic I started. I thought it looked familiar.

[JoeBean]

Oops, fixed the auction links in my original post. It was this:
POWER DRAWBAR BRIDGEPORT MILLING MACHINE LIMITED TIME - eBay (item 120654916884 end time Jan-30-11 15:32:55 PST) and Air Power Drawbar MILL Bridgeport-Kurt-Maxi Torque A&T - eBay (item 300510113716 end time Jan-30-11 11:58:00 PST)

A little bit of R & D tonight and I've come up with a few things... The easiest way to make the ATC work with an Arduino driving it is likely to be a USB cable to the controlling computer and a Mach 3 plugin developed in C++. This is pretty doable and not too much of a problem, just a bit of work. The whole process would use Firmata (Main Page - Firmata) to tie it all together.

As for helping out, simply providing feedback as the idea and design progresses is more than enough to help out. Speaking of which, this is a view of the top of the mill with the motor removed (it would be towards the right.) The left most set of screws/holes seems a great way to secure the whole assembly to the top of the mill:
http://i920.photobucket.com/albums/a...v/IMG_1438.jpg

Thanks!

-Mike

What exactly are you going for - a rotary toolchanger, or a stationary one?

In either case, shouldn't a simple mach3 macro, and using the mach3 outputs through the bob, be all that's needed to control the Arduino? 3 outputs should be able to specify 7 different tools (assuming at least one high output to trigger the event) in the case of a rotary, or just one output signifying a change start for a stationary toolchanger.

[webgeek]

What exactly are you going for - a rotary toolchanger, or a stationary one?

It won't be stationary as I want all the room I can get on the table. I use every inch of travel as it is. Wether it will be rotary or something else I've not yet decided. I do know that I want to avoid doing anything nasty with custom tool holders if I can. For instance, the ATC Tormach is working on requires grooved tool holders and I don't like that at all. Sadly, this might mean a more complicated arm system like Haas uses. A rotary device to hold the tools and an arm to switch them. The advantage is the "gripper" on the arm no longer has to be repeated for every tool slot in the tool bank.

In either case, shouldn't a simple mach3 macro, and using the mach3 outputs through the bob, be all that's needed to control the Arduino? 3 outputs should be able to specify 7 different tools (assuming at least one high output to trigger the event) in the case of a rotary, or just one output signifying a change start for a stationary toolchanger.

Hah, I feel like a moron. This is what happens with late night brainstorming. I looked at ModBus and the like and figured I'd want to do something where Mach 3 could send meaningful data to the ATC controller. I never thought to step backwards to something more simple. Your are right, 3 outputs controlled in a simple macro would give us 7 tools. Doesn't even need an enable, it's quite simple to just trigger it off the three. Four outputs would give us 15 tools. Then a single input in Mach would give us a "tool change complete" response that would be looked for before the macro completed. Piece of cake. The outputs would be brought low after this came back, ready for the next tool change.

I'm glad this means no C++ and plugins. That was a fair bit of work I wasn't looking forward to. Thanks!

-Mike

P.S. See? Community involvement is already helping

[kvom]

I have an air-operated Kurt clone tool changer on the Bridgeport. This wouldn't work on an NM200 since it relies on the drawbar going up and down with the spindle. To change a tool on the BP I raise the spindle to the top, lock the spindle, and press a button. The top of the drawbar has splines that engage the air motor.

There's also not that much room between the NM200 spindle and the front of the motor.

[ihavenofish]

ive got an ATC spindle setup in the works. were trying to keep it as cheap as possible. should be under $3000 for the spindle with drawbar, motor, inverter and fittings. there will be an all aluminium version meant for routers that will cost even less.

i was planning on having it take both iso25 and TTS style tools, but i dont know if theres a point to using the tts tools. they are only a little cheaper, and dont offer any other advantages if youre going to get rid of the r8 anyhow.

[webgeek]

I have an air-operated Kurt clone tool changer on the Bridgeport. This wouldn't work on an NM200 since it relies on the drawbar going up and down with the spindle. To change a tool on the BP I raise the spindle to the top, lock the spindle, and press a button. The top of the drawbar has splines that engage the air motor.

Sorry, I wasn't suggesting we clone that specific aspect of the design, just that it's a clean approach and worth inspecting.

There's also not that much room between the NM200 spindle and the front of the motor.

Yup, this is a concern of mine. I'm calling Khai for exact dimensions but a super rough measurement shows that we have roughly 6" wide, 5" deep and 12" tall of space to work with unless we modify the cover or remove the door. Since I plan on gutting the impact driver anyways I think we can get it to fit though it will be tight. I'm going to buy one in an hour or two to see what it will take. Removing the door seems totally acceptable but I'd really like to not modify the cover if at all possible. This means I need to keep things tight.

ive got an ATC spindle setup in the works. were trying to keep it as cheap as possible. should be under $3000 for the spindle with drawbar, motor, inverter and fittings. there will be an all aluminium version meant for routers that will cost even less.

I have no doubt of the quality and value of what you will produce but that's more than I'm willing to pay and I don't want to gut that much of the machine when I'm already quite pleased with the vast majority of it. My reason for TTS is that it's $28 per collet holder vs. ~$100 for ISO 25. TTS wont go super high RPM, true, but as it stands I'm happy with it and it's cheap/accessible. I didn't plan on getting rid of the R8 at all. My plan was to simply work with what's there. It won't be perfect by any means, but I believe it can be reliable and reasonably inexpensive.

Thanks!

-Mike

[ihavenofish]

yeah, not for everyone, just thought id mention it cause ive seen some people spend a ton of money trying to implement drawbars on existing spindles.

you dont have to gut anything though. the head pops off with a few bolts, then you just bolt on the new one.

[webgeek]

Update
I snagged the Harbor Freight impact wrench:
1/2" Electric Impact Wrench

It's a horse. Both physically large and heavy. It's also very strong. It will be reversible via the relays no problem as well which is nice as that's not always the case with AC motors. In addition to these good things, our first wrinkle has come up. Unlike your typical electric drill, the red plastic housing is actually required the support the motor. The coils are held in position as is the rear bearing all via the "unibody" red housing. The motor itself has no other housing. It also provides support for the brushes, which are replaceable.

This puts us in a tricky spot though, the design will include removing the handle but it won't be able to reduce the size of the wrench much beyond that as there is no where to go. It also makes mounting a bit more nasty as the four screws you see from the front just mount into the plastic. They are quite tight, but I wouldn't trust them long term. This means some sort of face mounting option.

Touch more news, I've reached out to Khai and he is providing me with all the exact dimensions for the various parts so we will be able to start machining up a prototype ASAP. Additionally, I've already started the CAD diagrams needed for the design. I expect to finish all the CAD work no problem this weekend. I'll post dimensions and some pictures later tonight. Thanks for all the input!

-Mike

[kvom]

Why use a 1/2" impact? A 3/8 should have plenty of torque and be smaller.

[webgeek]

The main reason was supply. Do you know of a cheap corded 3/8" impact wrench? I'm totally open at this point as the currently choice is large. Thanks!

-Mike

[JoeBean]

3/8 electric IW aren't usually made unless you want a battery, due to the volume required for the parts needed to get the relatively low torque 120V motors to be useful. it seems the 1/2" electric torque wrenches are really just really big 3/8 (or even 1/4) IWs with a 1/2" nose stuck on to keep it from looking stupid, considering the low torque ratings most have.

[rowbare]

Here is one made from a battery powered impact wrench: [nomedia="http://www.youtube.com/watch?v=G0VK3lQwLwc"]YouTube - Power Drawbar[/nomedia]

bob

[webgeek]

ooh, that one is VERY nice. Quite a bit more expensive though as you need batteries and the charger. Having to stop to recharge is a huge inconvenience too in a production tool... If there was a way to power it without the batteries, that'd be ideal. So, quick question to the community:

How many foot/lbs of torque is the "minimum" to consistently secure tooling in the NM-200? Is 100 ft/lbs enough? Is 50 enough? I ask because if the voltage drops to 12, I know how to make a good power supply inexpensively. Thanks!

-Mike

[webgeek]

Never mind about the power supply issue. I've tracked down both 12v and 24v power supplies quite reasonably - less than the battery + charger system and more than powerful enough. So the question is still:

How many foot/lbs of torque is the "minimum" to consistently secure tooling in the NM-200? Is 100 ft/lbs enough? Is 50 enough?

Thanks!

-Mike

[rowbare]

If you are planning to use TTS or a clone, you might find this interesting: Best Practices for Preventing Collet Slip « Milling Around

They mention torque values in the 30-40 ft/lb range.

bob