Community supported PCB milling machine
PLJack
Posts: 398
This thread is a continuation of this thread:
http://forums.parallax.com/showthread.php?p=538861
First some thread background:
Ok, then, things we need to do to make it real.
The design must be very cost efficient so that everyone can build one.
With that in mind:
The communication interface should be serial.
USB would be nice, but expensive to implement.
The IC's should be SX chips.
Material types should only be specified if needed.
For instance, the enclosure can be made of anything but the rods should probably be steel.
Do we need rods? Maybe something cheaper.
Design with varying motor quality in mind.
SX source code is free of course.
PC software free as well. Distributed as a ready to run package.
Develop a serial protocol for milling. Maybe use an industrial standard?
I'm willing if you guys and girls are.
Here is what I am willing to contribute:
Develop all the PC side software.
Research material.
Design and CAD work.
Anything else I can think of.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Jack
Post Edited (PLJack) : 6/10/2005 8:28:20 PM GMT
http://forums.parallax.com/showthread.php?p=538861
First some thread background:
PLJack said...
I would kill for a circuit milling machine.
What a great tool for prototyping. I think everyone should have one.
I wonder if the SX community as a whole could design a milling kit.
I'd be happy to write the Windows interface.
KenM said...
That is a darn good idea. I would be willing to chip in where I could.
Let's see where this goes. Maybe start a new thread to see if there is interest.
The basic concept as I understood it is
1. Create the artwork
2. Download to the machine
3. Set the milling depth and longitudal speed (too fast and you break bits)
4. Set the milling machine at a known XY corner
5. Hit a button and let it rip.
6. Once the milling was done, it would then wait for a drill bit to be installed, in a certain order for diff size holes.
If this actually goes anywhere, maybe Bean could take some pics of the machine at his work and add details that I left out.
Bean said...
It is basically a fixed table about 20" long by 12" wide. There is a carriage above the table that travels along it's length. On the carriage is a head that moves back and forth along the carriage that has the bit and motor. Then there is a seloniod on the head that controls up/down of the bit.
It would definiately be a great project. Ken Gracey has also expressed interest in a PCB milling machine.
I'll see if I can get some pictures of our machine. It's a "LPKF C60".
Ok, then, things we need to do to make it real.
The design must be very cost efficient so that everyone can build one.
With that in mind:
The communication interface should be serial.
USB would be nice, but expensive to implement.
The IC's should be SX chips.
Material types should only be specified if needed.
For instance, the enclosure can be made of anything but the rods should probably be steel.
Do we need rods? Maybe something cheaper.
Design with varying motor quality in mind.
SX source code is free of course.
PC software free as well. Distributed as a ready to run package.
Develop a serial protocol for milling. Maybe use an industrial standard?
I'm willing if you guys and girls are.
Here is what I am willing to contribute:
Develop all the PC side software.
Research material.
Design and CAD work.
Anything else I can think of.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Jack
Post Edited (PLJack) : 6/10/2005 8:28:20 PM GMT
Comments
That should be "Community supported PCB milling machine" not "IC milling machine"
I'll sign up for the SX control software.
Bean.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"SX-Video·Module" Now available from Parallax for only $28.95
http://www.parallax.com/detail.asp?product_id=30012
Product web site: www.sxvm.com
"If you keep doing what you always did, you'll keep getting what you always got."
·
-dave
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
This is not a sig. This is a duck. Quack.
Here are my thoughts:
Drill = Variable speed Dremel
X-Y Axis= Stepper motors
Z Axis = Bang-Bang (Up or down)
Able to route a 0.010" trace with 0.010" spacing.
Able to route a 12"x12" board
You get the idea.
I think someone could build it for $500 ?
Bean.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"SX-Video·Module" Now available from Parallax for only $28.95
http://www.parallax.com/detail.asp?product_id=30012
Product web site: www.sxvm.com
"If you keep doing what you always did, you'll keep getting what you always got."
·
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Jack
$500 might be out of reach for most.
Why buy the entire Dremel? Would just the motor be cheaper?
There is probably no way to avoid using stepper motors. Hmm
Actually, around $500 might be more realistic.
I am hoping to develop a system that could be of varying costs depending on how accurate you need it to be.
Just drilling the through holes would be a great time saver.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Jack
$500 might be high to some, but I think with reasonable scrounging skills and an adaptable system (generic stepper driver, etc) the cost to the resourceful will be lower.
I can't recommend the Dremel for this application. I can't specify why, but I wouldn't use one in an electronically controlled milling machine that needed high acuracy, precision, and repeatability. A belt driven system with ball bearings would be much better in my opinion, and would allow the builder to pick their own motor, provided it fit within a small set of parameters.
I'm thinking specifically of a standardized ball bearing supported spindle and chuck for the milling bit, with a belt take-off at the top to whatever motor is driving it. Belts provide a great amount of adjustability, as well as shock absorbtion for the motor in case of failures. There's also no backlash with belts. As an added bonus, if built with heavier duty components, the system could be used for milling things other than copper PCB traces, like thick PCBs, thin aluminium, or acrylic, which would be useful to a lot of folks.
My idea right now:
Timing belt connection between all motors and driven shafts (ratios as needed for absolute positioning requirements)
Limit switches for automatic calibration
Stepper motor drive for positioning, DC motor drive for spindle
12"x12"x6" work space (X-Y-Z)
I'm still looking into dovetail positioning slides, high precision drawer slides, leadscrews, etc. Remember, that the drive and positioning systems must be protected from or tolerant of very small pieces of copper and fiberglass. Hmm..maybe an integrated vacuum connection?
-dave
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
This is not a sig. This is a duck. Quack.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Sid Weaver
Need a bezel for your LCD?
Newzed@aol.com
·
I would have no problem paying $500 for a home brew PCB milling machine.
Ryan
What do you measure to get the length of a timing belt? I do not recall seeing any belts that were long enough to go around two small wheels that were 12 inches appart, but can't say that I payed that close attention.
I will not be in the market for a milling machine, but will be glad to help with finding parts from the stores here.
Russ
Great stuff so far guys.
I was thinking belts too. Once we nail down a basic design we can start cad-ing up some test images.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Jack
Sid, the Rotozip isn't a bad idea if we can keep the distance between the cutting tip and the collet bearing as short as possible. I have similar insight on them as the Dremels, and they do have much greater promise. They still suffer from early bushing death when given a side load, so we should investigate ball bearings as a retrofit for the busshings if the mill grows to include board outline cutting at any kind of speed. They are a lot more durable than the Dremels however. I'm suprised I forgot about them. Silly me...
I still like the idea of an uncomittted motor better, but this is a democratic effort, so let's go where the majority heads.
I have access to a 3-axis manual mill, so I may be able to do some proto work as well as board layout. I don't know when the last time the mill at work actually ran however, since it's under a pile of boxes about 9' high.
-dave
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
This is not a sig. This is a duck. Quack.
I hope this project gets off the ground - I'd love to have one before I get too old to use it.
Incidentally, I've used my RotoZip quite a bit.· There is no exploded view in the little instruction manual, but· I'm sure the adjustable chuck screws right into the motor shaft, and I would think the motor shaft runs on ball bearings.· At any rate, I can detect absolutely no endplay·at the tip of my cutter.
Sid
I would like to start the layout of the machine but I don't know too much about them. Yet.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Jack
This has the makings of a great think...but at the same time it could suffer the set back of too many cooks in the kitchen. At least with someone that stamps the final word, the project will get done rather than be an endless debate on why this or why that.
Can I ask that the unit be gun metal blue (of course there'll be arguments on this!! lol...) [noparse];)[/noparse]
Also, the captain of this ship would be able to set timelines on what different 'volunteers' are offering!
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·
Steve
http://ca.geocities.com/steve.brady@rogers.com/index.html
"Inside each and every one of us is our one, true authentic swing. Something we was born with. Something that's ours and ours alone. Something that can't be learned... something that's got to be remembered."
First if you live near Illinois, they are having a CNC workshop on builting your own machine or modifing a manual machine. Go to this web site www.CNC-Workshop.com and check it out. I personally do not how good the work shop will be. But a know of some of the vendors.
I do not want to throw "cold water" on the idea, but here goes my thoughts.
The rigidity and the smoothness the X-Y-Z tables has all the world to do with machining. Builting these tables is not easy task. There are alot of articles on the WEB about builting CNC machines for the Home shop. I have known a few in the "Metal Munchers" group here in Georgia that have tried it and have had little success building an inexpensive CNC machine.
The second thing is the controls that control the stepper or servo motors that move the tables.
Straight lines are not hard to do, but circles or curves are a whole different ballgame. Most of the time the code to control CNC is written in a form of "G" code. This in turn is converted into step pulses that are sent to stepper drivers that control the motors. There are alot of programs and controllers that are relatively inexpensive, but to do all for under $500 is going to be rough.
They are more than $500, but "Sherline" and several others make some nice small machines. Check out the magizine "Home Shop Machinist".
Joe Fishback
Fishback Designs
check out this one... http://cq.cx/pcb-router.pl
and this one too... http://www.crankorgan.com/plans.htm
There are a bunch of good homebrew pcb milling machines out there
Bean.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"SX-Video·Module" Now available from Parallax for only $28.95
http://www.parallax.com/detail.asp?product_id=30012
Product web site: www.sxvm.com
"If you keep doing what you always did, you'll keep getting what you always got."
·
As steve_b mentioned, there should be somebody to oversee the project, coordinate between members etc.
Any volunteers?
If Ken Gracey lets us use some of his workers' time, I'm sure it'd be a great PR thing.
In any event, someone somewhat neutral; that has as much to gain/lose from this as everyone else interested. (that way there's no bias)
What's the difference in a $500 and a $1000 machine? Workable area or accuracy?
I think the end product should be affordable.....with maybe 'options' for the masses (cup holders
What's the end result here? To have a product to sell? or a design to share for others to build for themselves? (and download the software to work it)
how do you compensate those that are 'truly active' and are out of pocket $$ buying bits and pieces to test (this can't totally be a "paper" project...at some point you have to build it).
Anyhow....definately need a moderator to organize the flow here! Of course, Parallax's permission should be sought as we plan on using their forum for this.
Many many questions.....
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·
Steve
http://ca.geocities.com/steve.brady@rogers.com/index.html
"Inside each and every one of us is our one, true authentic swing. Something we was born with. Something that's ours and ours alone. Something that can't be learned... something that's got to be remembered."
That is why I started this thread.
I thought if we could build a stable platform, develop the SX code
and protocols, that we could all have an affordable prototype milling machine.
Since I have the expertise and am willing to do the work it seemed like a great opportunity for us all.
I will happily volunteer as moderator.
Although it should be made up of a team.
Myself as software designer.
An SX / circuit designer.
And a mechanical designer.
All three people taking suggestions from the community.
In any case I would need final word on any software designs.
Steve is right about the "too many cooks" statement.
Life is too short to write software for a committee.
If our SX protocol is designed correctly I can write software to get that
table to do anything we need.
We could start the SX protocol and software right now while we are designing
the mechanical platform.
So what is it going to be people.....
Should we get our act together and go for it.
I think we should. Sounds like a good time.
p.s.
We could all learn from this site that Steve Joblin posted:
cq.cx/pcb-router.pl
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Jack
Post Edited (PLJack) : 6/12/2005 3:37:46 AM GMT
I will do what I can to help. I can do some basic autocad work however I am self taught but have been at it a couple of years.
I also think I might have a good platform to build from as there is a mill in my garage that just needs steppers, brackets and controls. There appears to be the needed bracket holes·ready to go (at least for the X and Y). Will need to do something for the spindle speed as the current max is 1500
Ryan
Actually, I'm talking about the axial load on the bushings. The further you put the cutting tip form the bushing, the more lateral load is developed against it (basic lever math here). The bushings are designed for limited, shot term loading no more than 1 3/4" from the bushing. If we're going to be using a rotozip or something similar to cut metal (copper) and fiberglass (board outlines) and want a long, happy life from our tools, we should really look at either minimizing the lateral load on the end bushings or replacing them with ball bearings.
-dave
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
This is not a sig. This is a duck. Quack.
Sid
Dave
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Dave Andreae
Tech Support
dandreae@parallax.com
[noparse][[/noparse]url]Http://www.parallax.com[noparse][[/noparse]/url]
·
I mean etching a PCB isn't that hard to do, but getting all the holes in the right spot is a pain.
The software would be pretty easy, go to the right spot, move the head down, them back up and go to the next spot.
We could then build on that to make something that will mill boards.
You can etch small boards as a multi-panel just as easy as a single board, it's the drilling that is the pain.
If you have 50 holes per board (not hard to do if you use an SX28 or 24 pin stamp) and make a 3x3 panel thats 450 holes to drill.
It's just a thought,
P.S. I have some Turbo Pascal code I wrote a long time ago to read DXF files. I'll search for it.
Bean.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"SX-Video·Module" Now available from Parallax for only $28.95
http://www.parallax.com/detail.asp?product_id=30012
Product web site: www.sxvm.com
"If you keep doing what you always did, you'll keep getting what you always got."
·
Sid
z travel 2-3/4.· Too small, but its a start.
Sid
also I would like to help out.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
DTQ