DIY CNC-Router from RotoZip
Ardie Scott Powell
got the RotoZip.
http://www.instructables.com/id/Easy-to-Build-Desk-Top-3-Axis-CNC-Milling-Machine/
Paul Larson
I had seen this a while back and considered if I could use my craftsman rotozip-like tool with it. Looks like a useful, and relatively cheap project.
-Paul Larson
> You received this message because you are subscribed to the Google Groups "ATX Hackerspace" group.
>
> To post to this group, send email to
> austin-ha...@googlegroups.com
>
> To unsubscribe from this group, send email to
> austin-hacker-s...@googlegroups.com
>
> For more options, visit this group at
> http://groups.google.com/group/austin-hacker-space
Jerry Rutherford
Servos are better than steppers... but steppers are cheaper and well suited for small projects.
A RISC based serial/USB controlled processor is better than a parallel port of a PC, but sometimes a cheap PC and a home-brew circuit can get the job done.
EMC2 is free and well supported... worth looking into, but there are also other companies out there with pre-built packages.
It all depends on how deep you want to go.
Having said all that...
I have an X/Y table, power supply, two gecko drives, and three steppers. One more gecko drive and you have 3-axis. I had it on my lathe for a bit... but it wasn't beefy enough... and I'm going to use servos on the final design. It's set up to use a printer port for the step-direction input. Someone want to play?
Jerry
Askjerry... everyone else does.
Visit me online at http://askjerry.info
See my projects, video links, tutorials, and blog today.
Danny Miller
Brushed rotary tools are really problematic for the job and I would not recommend spending time or money on them. Why not just get an actual milling shank, like an ER8-ER20? Those are dirt cheap, and a million times more accurate. Slap it in some bearings, spend $20 on a brushless motor vastly more powerful than the Rotozip, $10 on a driver, add a power supply and you'd have something actually awesome. This is nice because you'll actually mount the bearings in bearing blocks here instead of trying to hold it via the plastic case. Also quiet enough that you can actually use it.
And a milling machine isn't that expensive. An actual high-performance mill with NEMA23 mounts and tables and all can be bought OTC for $1100. A lot of the massive knee mills show up here and there at auctions or For Sale ads for just about nothing because they're hard to transport.
Danny
David Mitchell
How about doing this yourself and showing us how awesome and cheap it really is? We can have a CNC-off and compare the Rotozip CNC vs the Millermatic?
I guess my point is, the original question was, "does anyone want to help work on this?", it wasn't "how perfect is the idea and how much better could it be done some other way?"
I hereby appoint Danny "Chief Pessimist".
-David
Martin Bogomolni
Ardie Scott Powell
EBo
> damn...am I good at stirring up trouble, or what!
It is not you. This is Danny's high horse, and it is aggravating most
of the time. No, make that all the time, but here's the deal - he knows
alot on the subject, and if you can get past his snide and condescending
remarks then you will likely learn a thing or two about power systems,
steppers, servo motors and the like. So, if you want to start learning
and want to build that little CNC, then go ahead. There are plenty of
us who will cheer you on. When you get to the place where that machine
will no longer do what you dream of it doing, ask Danny to explain to
you motor drivers and why you might want 5x the rated voltage and limit
the current. With a better driver, you can get the same machine to have
a maximum traversal rate of 3 to 5 times you can get with the type of
circuit you are planning to use on your machine. On the other hand that
design is simple and understandable for a first wack at CNC. I'm so
tired at the moment I cannot even remember the proper terminology, so
please go ahead and play and have fun, and I for one will give you a
thumbs up for trying.
And for Danny, not all of us have an extra $1100 lying around to spend
on a machine. Who knows, maybe the machine he is building is fast and
precise enough for his needs. Now that is the heart of one of our
arguments - you can build a machine that meets your needs without bowing
to the almighty got of ultra precision and have it do what it is
supposed to do.
Laters,
EBo --
Ardie Scott Powell
1. Why use a RotoZip (You called it a brushed motor) instead of the
one you've mentioned? Because I already have the RotoZip. This is
about making use of something I already have on hand. As to your
question about why there is this "drive to hack together from really
inappropriate parts", you hit the nail on the head with one of the
words in your question. HACK. Perhaps they'll be appropriate,
perhaps they'll be inappropriate. Either way, something will be
learned in the process...and that's the goal. I would welcome your
assistance & instruction in the endeavor.
2. Stepper motors vs servos: Since I know absolutely nothing about
either of these things, I'm not the one to make such a decision
without at least being shown what the differences are in terms of
capabilities & costs. Hell, half of why I asked for "collaboration"
on this was so I could start learning some of this stuff. I can go
about as far as understanding the physical assembly of the unit as
described. Connecting the electrical components may or may not be
beyond me. Programming something to control the unit? Definitely
beyond me.
3. Size: As mentioned by Jerry. I had actually already thought of
that, and planned on scaling it up anyway...24" was about the size I
was thinking of shooting for.
4. Use: As for what my "needs" are? To learn, and no other reason.
I have nothing I can think of that needs to be put through a CNC
machine. I want to do this because I saw it on Instructables, thought
"damn that's cool looking", and also thought "hey, I have one of those
RotoZips, I wonder how difficult this really is". I imagine if we can
get it up and running, it would be a handy item to have around The
Space. The fellow who introduced me to the concept of HackerSpaces
(he was looking for one near him in SC) recently posted a video of a
rather extreme CNC performance. Cnc cutting 1018 steel with a 1"
endmill, no coolant at 2750 rpm, 45 ipm, 0.1" deep at 0.75" stepover.
Hopefully that makes sense to someone else, cuz I don't understand
more than half of it. Here's the link, it's pretty friggin cool to
watch.
https://www.youtube.com/watch?v=SHd-KWi4Ung&feature=youtube_gdata_player
EBo
>
> ... Slap it in some bearings, spend $20 on a
> brushless motor vastly more powerful than the Rotozip, $10 on a
> driver, add a power supply and you'd have something actually awesome.
You have mentioned these cheap drivers several times. Would you list
some part numbers? I would not mind playing with a brushless drive and
control if they were reasonably priced. Double that for a brushless
servo that had positional positional control.
EBo --
Danny Miller
12v * 1700KV=20400 RPM.
80A peak=960W.
60A continuous would be a reasonable guess, that would still be about
1HP. Since the losses are primarily I^2*R, running at 60A is only 56%
of the I^2*R heating of 80A inside the motor. Sounds conservative enough.
For the most part, as far as I can tell, the max continuous power rating
is gonna go down proportionately with RPM. If you lower it to 5KRPM,
it'll be down to 180W. That big Taig AC motor only has a 1/4HP
continuous rating (186W).
Got that motor right here if you wanna look at it and see if it'll fit
or whatever. I might use it for a mill conversion, probably not anytime
soon though. These things sometimes take a long time to order and it
wasn't really expensive.
80A speed controller:
http://www.dealextreme.com/p/mystery-speed-controller-80a-esc-for-brushless-motors-on-r-c-helicopters-41165
A scrap ATX supply should be able to provide the absurdly high 12v
current needed.
You'd need a microcontroller to turn on the speed controller, emulating
an RC receiver, but it's basically a trivial job. The pulse width
determines the throttle given to the motor. Speed regulation is
open-loop unless you wanna get complicated, but the motor's probably not
gonna bog down much, not with this sort of overkill.
That motor will handle up to 24v, 40K RPM. However, there's an Io
(no-load current) loss which will probably become very significant by
the time you get to 24v. The Io loss is a source of heat inside the
motor, and deducts from the ultimate power handling capability. It
would be a 2KW motor at that point, but might lose like 30W in Io loss
inside the motor. But it's beside the point unless you have a power
source capable of delivering like 24V 80A (a 110v circuit could barely
even deliver that). But it doesn't have to be 80A, the motor could be
driven faster just because you need it to spin faster even without a
high HP loading on it. If you wanna use the a 12v, 50A PC supply for
milling aluminum then change to a 24V @5A supply for high-speed,
low-torque engraving, that makes some sense.
I also have one of these looking for a job:
http://cgi.ebay.com/Power-36-KV600Brushless-outrunner-C4260-motor-New-/320737709397?pt=Radio_Control_Parts_Accessories&hash=item4aad750d55
<http://cgi.ebay.com/Power-36-KV600Brushless-outrunner-C4260-motor-New-/320737709397?pt=Radio_Control_Parts_Accessories&hash=item4aad750d55>
Lower RPM/V. Still has a 1100W continuous handling capability. Still
cheap!
On a large router, the concept might fall through, because the gauge of
wire needed for carrying high amp/low voltages over many feet, with low
loss, can become unreasonable.
Danny
EBo
MaxNC. I'll need to find some collet that will attach to the motor, do
you have suggestions for that?
John Dolecek
--
You received this message because you are subscribed to the Google Groups "ATX Hackerspace" group.
To post to this group, send email to
To unsubscribe from this group, send email to
David Mitchell
To unsubscribe from this group, send email to
Danny Miller
Collets, they do SELL collets. It depends on what diameter you need to attach to. In the RC world, we've got prop adapters and sometimes shaft couplers or flex couplers for boats, designed to grip a flex cable.
IMHO it'd be simplest to just use a lathe and some aluminum bar and just machine out a fixed coupler, install set screws, and machine a flat onto the motor shaft. The spindle shaft probably already has a flat. It doesn't have the capacity to flex and that could be troubling, but, if it's done on a lathe, alignment should be good. Typically the machine world doesn't do anything but join them up with a rigid collar so I bet that'll do.
I did some work trying to machine the motor shaft itself on the smaller motors. Usually the shafts are available as replacements. However, it was dicey and didn't work. The bell- the rotating outer body which holds the magnets- is attached in back with a set screw and the front of the bell just sits there over the stator without bearings, but low clearance. In my motors, they pressfit the bell onto the shaft pretty tight and I ended up trying to tap it out- later progressing to "banging". The mounting hole on the bell was damaged enough that it couldn't be reassembled, the bell magnets collided with the rotor. But, that might be due to just how narrow they made the shaft collar on the bell. With a larger motor, it's probably not undersized like that. Plus, well, that motor could probably have been disassembled without so banging it apart like I did.
The motor itself could be mounted through rubber, well, it was an idea at least. Other than that, I imagine there's any number of ways to machine a coupler with some flex, but I just doubt it's all that necessary.
Helical flex couplers used for connection with leadscrew/ballscrews won't work, they'd wind up, or unwind.
Or just belt drive it. With variable reduction pulleys, it'll deliver better power at low speeds than electronic speed reduction.
Danny
To unsubscribe from this group, send email to
--
You received this message because you are subscribed to the Google Groups "ATX Hackerspace" group.
To post to this group, send email to
To unsubscribe from this group, send email to
Jerry Rutherford
http://www.grizzly.com/products/16-Piece-Quick-Change-Collet-Set-MT3/SB1351
The trick is that you are going to need to have a shaft mounted in a bearing block... then bore it to accept an MT-3... it might be cheaper to buy an inexpensive mill and just use the milling head.
If you don't need something in this scale... look at the MT1 size... likely an adapter could be found or made. It's a good weekend lathe project. (I made my own MT3 half-dead center... 'cause I was cheap.) http://www.toolmex.com/new/products/product.cfm?id=16155
Jerry
Askjerry... everyone else does.
Visit me online at http://askjerry.info
See my projects, video links, tutorials, and blog today.
Danny Miller
Not a Morse taper, straight shaft. So you'd just need bearings and a bearing block, hopefully some sort of shielding against dust. The ER systems are pretty great IMHO, easy to use. Self-extracting collets, for one.
ER collets come in system diameters from 8mm to 40mm, the shank diameter is not tied to the ER diameter so you can get whatever works for your bearings. The shanks go up to at least 150mm.
They do sell collets piecemeal, but the set of 10 has a far better price:
http://cgi.ebay.com/FULL-ER16-COLLET-SET-10-PCS-CNC-MILLING-LATHE-NEW-A03-/150645421219?pt=BI_Tool_Work_Holding&hash=item23132ab4a3
So it's far more expensive to get individual collets ($20/ea a lot of the time). Also means that buying a shank like the one I showed above, with a collet, isn't worth paying extra over just getting the bare shank if you're just gonna get a full set and toss the collet it came with as a "spare part". But, 1/8" shank IS a standard size for small bits now. All PCB work will be done with 1/8". I've used larger, like a 1/4" bit. Also used the other collet sizes for drill bits because the hole depth was too high for an endmill. But all carbide PCB drills, engravers, and routing cutters have the same 1/8" shank regardless of end diameter.
Danny
John Griessen
> Not a Morse taper, straight shaft.
The purpose of using collets is concentricity... straight shafts
mount with clearance and inherent wobble away from concentricity.
So, dismissing Morse tapers as a way to mount collet sets is
missing the point.
Ardie Scott Powell
that plan (the physical construction) adaptable to the aternate driver?
Jon Kelly
assist in the construction I have donated a Dremel Press for permanent
use to the space. It is located in it's box underneath the tool bench
(under the mural).