Slashdot Mirror
Build Your Own PCB Milling Machine
mwandel writes "It used to be that one off amateur printed circuit boards were all etched in acid. A lot of companies nowadays use a special form of milling machine to mill them out of solid copper clad circuit boards. This guy Jonathan Westhues built his own PCB milling machine out of various parts, with a laminate trimmer as the milling head. Lots of other neat hacks on his Webpage as well."
Comment removed based on user account deletion
More people need to be able to make their own parts out of steel and plastic. The problem is cost. It's curious that a country that is inventive as ours doesn't have some type of affordable CNC(computer numeric computation) milling machine.
Affordable metal cutting lathes are expensive too.
/. really needs a mirroring system or at least ask guys like this and/or give them fair warning to prepare.
2 comments so far and its already down. What's the point? The article might as well be yanked all together.
..mork
Seems like a lot of trouble to go to when it's pretty cheap to get small-quantity custom boards done.
They're still called PCBs even if they are milled. It's a naming convention that has stuck. It is in no way incorrect.
Also, It is not difficult to make a double sided PCB with a milling machine. There are many times in circuit board design that you don't need more than one or two layers. This is a solution that solves a lot of smaller problems every day.
"Also since these would be milled, they would be called MCB's. Get it straight. "
Yeah and traditional boards are ECBs right?
"Milling machines, while a marvel of modern technology in their own right, cannot create multi-layered circuit boards, unless you ignore the obvious extra steps involved."
Yeah and etching a board obviously couldn't either, could it??
Troll alert!
If you want to build a machine to mill circuit boards, do it right. Build a machine designed by John C Kleinbauer. The Brute is designed to make PCBs cheaply. I recently bought some of John's plans (well worth it) and they are quite nice. They are very well done, easy to understand, don't need things that are exotic and hard to get (if your in the US). He includes a booklet on how to mill PCBs with The Brute, or you can order it seperatly. He even maintains and activly participates on his forum, Hardware Store CNC.
I've started to build a brute, and things are going pretty well considering I'm doing this in my spare time with only some time to work on it. If you guys are like me (I really like to build things with my hands) this is a ton of fun. I can't wait to get it running so that I can make PCBs, robot parts, a wooden clock and more.
Comment forecast: Bits of genius surrounded by a sea of mediocrity.
My dad wrote some CNC control software that would work quite well for this application. It's designed to run anything from table top machines (such as this) all the way up to large scale CNC retrofits, where the iron is good but the control is shot. Very competitively priced as well. Even has a free demo version with no time limit for those that want to check it out. Requires a dos based machine to run it on though. FreeDOS works fine, of course so does MS-DOS.
:)
I'm sure this will kill his pipe, but here is a link: www.cnczeus.com
It's listed in google as well, so you may want to check that out if/when the pipe goes dead from the load.
Good article for those that don't already have access to PCB milling equipment. There really is no reason to do the old-school etching method anymore, in fact, I don't even know any hobbyists that do that anymore. Milling equipment can be found, borrowed, or made pretty easily these days. I've even seen a working setup made from Lego Mindstorms and a cordless Dremel! Hey, it works and beats the heck outta the mask-and-acid roll of the dice method.
I admit that there are some CNC machines that have more degrees of freedom, but for sign cutters and circuit boards, it's just a large plotter.
I'm surprised that as the old school plotters get decomissioned, more people haven't snatched 'em up.
Build it, and they will come^Hplain.
It seems dabbling in electronics is a dying hobby for the younger crowds... I hope projects like this spawn new curiosity and interest.
i tcellar.com/
For those new to this hobby... here are some publications that could be of great value to you:
http://www.nutsvolts.com/
http://www.circu
http://www.poptronics.com/
Anyone know of any others?
- Preferences: Solaris 10 (servers), Ubuntu (desktops), Solaris 11 (personal servers) -
for general mcb construction, but it really depends upon the application. Most of the stuff I would use it for would be RF, and 'ugly' construction is a heckuva lot faster, and usually performs better.
For computer hacking stuff, or for audio, this might be a handy thing to have.
Milling is the process of cutting by movign a quickly rotating cuting head in relation to the item being cut down. Most machine shop milling machines have a stationary head, and move the item around, while some cnc machines have a head that moves.
Of course, questions like this are where search engines come in handy -- http://www-me.mit.edu/Lectures/MachineTools/mill/i ntro.html
Build it, and they will come^Hplain.
Opensource Hardware!!!
3000 dead over past 2 years, still no free Palestinians, still
I can't get to the site (slashdotted). But I used one of such machines. Nowadays they are absolutely, unconditionally useless. They can't make the fine traces that are required for modern chips. Instead, it is cheaper now to order your PCBs from a board house (such as Advanced Circuits). $30 per board is not that expensive.
Great idea there, I went to the mirror and was able to read it with pictures and all. I bookmarked his site though and will check back tomorow as it looks pretty interesting.
Thank you, I would love to see that done more often. If it is worth listing in Slashdot, it should be worth helping the guy out with bandwidth.
A mill is sort of like a drill-press that can go sideways.
.001", you can quite easily.
In reality it's much more complicated than this, but you get the idea.
The bits are not pointed, they have flat ended cutting flutes. There are exceptions, of course, like ball-end-mills (but that's just adding to the confusion).
A mill in a machine shop can be used to make quite complex parts out of a billet (big chunk) of metal. It's table can move in X and Y axis, and the mill-bit moves in the Z axis. They can get much more complicated than this, but you get the idea.
A PCB mill typically operates more like a flatbed pen-plotter with a dremel attached... as it doesn't have to have much Z axis travel.
Unlike a drill-press, a mill will have micrometer controls, which means if you want to make a cut that's accurate within
Here's a quick search that should help: http://www.google.com/search?q=bridgeport+mill
Hope this helps.
- Preferences: Solaris 10 (servers), Ubuntu (desktops), Solaris 11 (personal servers) -
At the risk of sounding like an idiot, how exactly do you get the electrically conductive metal in the etching?
-------
"In times of universal deceit, telling the truth becomes a revolutionary act."
-- George Orwell
Back in my high school electronics class, we used to etch our PCBs in acid, and when the teacher was out of the room, we'd put things like pennies and copper wire in the tray just for kicks. Milling the boards is probably a lot more efficient and safer, but doesn't sound like as much fun.
Whatever happened to the guys that were going to print out boards on inkjet printers?
Maybe they statred doing this. Modified Printers Used For Tissue Engineering
PCB killing machine? I guess more beer less sleep isn't always better but it can be more interesting.
The best thing about this milling machine is that it built itself. The first version was controled by electronics on a breadboard. The second version uses much more sophisticated electronics on a PCB that had been milled by the first version of the machine. That is unbelievably elegant.
Slashdot monitor for your Mozilla sidebar or Active Desktop.
I used a small CNC miller to produce two complex alumnium plates that I used to produce a winch. It was complete overkill, but I was bored, and it did a very good job.
It had dodgy software and the sheet I was using was too large (X and Y) to use in the machine, so I had to make a jig using MDF and steel rod to locate the plate, so that it could be flipped, and the machine used to cut material much bigger than it was designed for.
The software and PC controlling it was updated last year, and now it can accurately profile 3D shapes. There are obvious restrictions, such as no undercutting (unless you use special bits, which is awkward). You can use opaque acrylic and etch photos onto it, with the darkness being converted into depth. By then holding it up to the light you see an image. Quite a neat idea.
I could see it being used for PCB production, but to be honest, it wouldn't be much use for anything small. It is very accurate, without a doubt moreso than the homebuilt one, but even with the smallest tips, I can't see SOIC outlines being milles with enough precision to be used. And most PCBS have some sort of surface mount components now.
I also see problems with raising the edges of the copper tracks slightly with milling. This would make surface mount components hard to get flat on the board.
Isn't Ferric Chloride (the stuff you buy in bottles at Radio Shack, or at least I did when I was a kid) actually a salt? FeCl... looks like a salt to me!
1.) United Auto Workers 2.) Gun control zealots
The college I go to has a milling machine. I've milled about 4 boards from it. That would be great if I could just have my own though and use it on my own time. However, could a home built milling machine achieve an accuracy of 10 mills for track width? Also, what software (opensource) could be used? I couldn't imagine a homemade machine achieving the same functionality as a $15,000 device, but if it could, a low cost home milling machine for hobby projects would be a great product.
Now they just need to make a low cost device for placing and soldering SMD components on the boards.
Over here, (in school at least) we just print out the circuit diagram onto a transparency sheet (inkjet printer of course) and then just use a UV light to destroy the appropriate parts of the circuit.
Use some chemicals (stored in a safe location and brewable in your own home) and you've got a nice PCB for you to use.
ahh....so if you pay...
;oP
does that still make you a karma whore?
thnx for the link/mirror.
Milling machines are not an overkill or a solution for amateurs, they are a very good solution for inexpensive prototypes or production needed in small-mid volume.
:)
Acid etching produces a functional board, but it simply does not look good, and you can't make 100 boards that are exactly alike. The lines are almost straight, the edges are not perfect, and if you are on a contract to deliver a product, this is not an option. Examples of where a milling machine is one of the best solutions: a university lab where the researchers are under government/industry contract and are supposed to deliver a working prototype or a small business with a military contract (small volume products).
I worked for a startup company for a while, and part of my job was to work on a QuickCircuit milling machine, and that thing had milling bits that were 4 mils thick (comparable to hair). You can hardly reach this precision with acid etching. I also adapted the machine to dice wafers, which replaced the company's practice of using an exactoknife
If you are in the IC design business, testing cheaply is of primary importance. You can get a full setup for producing boards for less than $10K. How's that compared to billions of $$$ for setting up an IC production plant? And if you are in the RF design business, you need the precision so that a crappy board does not screw up your high-frequency measurements.
Of course, milling is no option for producing high-volume PCBs with many layers, but don't think that ASUS spent months to design such a board to test their new motherboard design. First they have to verify that design works, which is done with a cheap PCB design, one that would hook up the ICs. Only after that stage can the final PCB design begin.
As long as the Z80 or 68K processors are still in use, simple PCBs will be here, and we need a cheap and fast way to design and make them.
Milling machines are not an overkill or a solution for amateurs, they are a very good solution for inexpensive prototypes or production needed in small-mid volume.
:)
Acid etching produces a functional board, but it simply does not look good, and you can't make 100 boards that are exactly alike. The lines are almost straight, the edges are not perfect, and if you are on a contract to deliver a product, this is not an option. Examples of where a milling machine is one of the best solutions: a university lab where the researchers are under government/industry contract and are supposed to deliver a working prototype or a small business with a military contract (small volume products).
I worked for a startup company for a while, and part of my job was to work on a QuickCircuit [t-tech.com] milling machine, and that thing had milling bits that were 4 mils thick (comparable to hair). You can hardly reach this precision with acid etching. I also adapted the machine to dice wafers, which replaced the company's practice of using an exactoknife
If you are in the IC design business, testing cheaply is of primary importance. You can get a full setup for producing boards for less than $10K. How's that compared to billions of $$$ for setting up an IC production plant? And if you are in the RF design business, you need the precision so that a crappy board does not screw up your high-frequency measurements.
Of course, milling is no option for producing high-volume PCBs with many layers, but don't think that ASUS spent months to design such a board to test their new motherboard design. First they have to verify that design works, which is done with a cheap PCB design, one that would hook up the ICs. Only after that stage can the final PCB design begin.
As long as the Z80 or 68K processors are still in use, simple PCBs will be here, and we need a cheap and fast way to design and make them.
As for drilling the PCB holes, I found that a good set of second-hand carbide bits, a dremel drill press, and a thin center punch could provide nearly perfect results. You *MUST* center-punch with the carbide bits, though, or they might start to "walk" across the board, which due to their brittleness could get quite expensive. Presumably the engraving machine can cut a nice pit in the surface to serve the same function, even though they often lack the force to actually drill the hole.
Hardware, software, and blinking lights!
Scanners are essentially a flatbed plotter with only one axis of travel - cheap USB scanners can be had for under $30.00, buy a couple to get a two axis system, third axis could be a simple solinoid config. There would still be a lot of work left to do to get it all to go together, but it could be done (whether you stuck with the USB stuff, or just kept the stepper motor and drivers, then added you own custom interface). Also, back in the early 1980's there was a BYTE magazine article on building your own plotter (not that difficult)...
Reason is the Path to God - Anon
bit tense aren't we...oh wait... AC...too scared to come out from under the bed. PMS alert!
Thankfully, the pin pitch is often a little bit easier to deal with, so it's not QUITE as difficult for making boards.... As long as you're willing to make multilayer boards.
The problem is SOLDERING the damn things! The balls are UNDER the package, facing the board, with no accessible places to work with. Some sort of area heating is required to solder them at all (normal hobby methods include hot-air heat guns and little temperature-controlled ovens).
Worse still, you have to hope that you have the chip aligned correctly. This is normally done via xray, but most of us don't have access to xray machines (although there was that "radio tube xray machine" thing around...).
As electronics get more and more integrated, they become less and less friendly to amateurs.
Hardware, software, and blinking lights!
Expense, both monitary and temporal. Most of the damned kids figure "why bother spending months learning basic electroncs theory, hundreds on breadboards and LCDs and Microcontroller programmers, when I can just go buy something that will do what I need for $10 at RadioShaft? Sure it'll break in two months and I won't have a clue of how to fix it, but so what? by then I'll be bored with it and be on to something else. Anyone want to buy a meow-chi?"
Another part of the problem is that pretty much anything you buy today is made with propritary custom BGA ICs. Even if you could get a hold of the chips to do some of the really suh-wheat stuff, have you ever tried to solder a BGA? right, doesn't work to well unless you have a half million dollar IR reflow oven. Hell, can you even buy DIP form factor for half the IC's out there anymore? Sure, you could do the same thing with 3,000 descrete components, but again, why would they bother?
--Spoken as one of those little whipper-snappers, and a Nuts and Volts subscriber.
--I don't want the world, I just want your half.
Sorry if this is slightly off topic, but seeing that it seems that alot of people responding to this story know about board design and such, about how many layers does a normal PC (like an Athlon/P3 if that matters) motherboard have in it?
(Score:0, Interesting)
The most serious problem comes in from the really high-end etching systems. Photoresist is fine with ferric chloride, but when you start using something like an acid-peroxide etchant, it will actually eat many "resist" inks. The resist pens become TOTALLY useless, for example.
And so this is where the old technique of serigraphy comes in. This is the "screen-printing" that makes the name Printed Circuit Boards. The resist ink used on the actual board is usually just some form of lacquer, which holds up well to even the harshest etchants. A stencil is prepared on a screen, usually through a photographic process. Since the screen resist doesn't have to deal with the same harsh chemicals, it is generally much more environmentally conscious and cheaper than PC-Board photoresist.
I actually set up to do this method myself, in a semi-hobby context. I had some previous experience with serigraphy, and the acid/peroxide system was cheaper, far more effective, and much easier to dispose of than the ferric chloride. On the flip-side, it will eat virtually anything metallic and smells aweful (it isn't much of a health hazard except in the sulphuric/peroxide system, but should be ventilated nonetheless), so it's not for the faint of heart.
Hardware, software, and blinking lights!
I dont know about milling, but I've got a PCB mining machine right here: a fishing pole dipped in the East River.
The nice thing about it is that, unlike Ferric Chloride, there are ways of reclaiming acid/peroxide systems. I used hydrochloric/peroxide, since it was the cheapest and safest (concentrated HCl hardly itches if you spill it on yourself, the 25% peroxide is a lot worse). In this system, if you dumped in some standard Lye (sodium hydroxide), the copper all precipitates out as a brown sludge (copper hydroxide), and you're just left with a mixture of salt water and peroxide, which decomposes to salt water when left open. The copper hydroxide isn't as hard to deal with as it sounds. If you're running a copper sulphate/sulphuric acid plating system, you can just dump it into the plating system.
Hardware, software, and blinking lights!
You might want to see if you can pick up an old X-Y plotter somewhere and use that as the basis of a DIY design.
Replace the pen with a high-speed cutter and then you can interface to the plotter quite simply through the RS232 port or whatever the plotter supports.
Here's a plotter on eBay that might be a good starting point.
Realistically, PC boards are made by sending output from a board design program to a service that makes boards. It's reasonably cheap, turnaround is good, and you can get double-sided with plated through holes, which is what you want.
Toner transfer methods are for people who like looking at their boards under a magnifier and doing rework. Do it yourself photographic methods work better, but few people bother any more. Anybody still have one of those small etching tanks with the aquarium bubbler?
A related business is making front panels. Send out your design and let a computer-controlled punch make all the connector holes. They also etch the lettering and anodize.
homecnc
A little P.C wouldn't hurt you either Ps that don't mean personal computer go to the redneck board
Diplomacy is the art of saying "Nice doggie" until you can find a rock. Will Rogers
Agreed at one time I ran a shop doing engineering modifacations for a high end workstation mfgr .001" that we used lasers I told them that I used a chainsaw
did it all with saws and files High tech is not always necessary or even the right way to go
by the way they thought since we did the cuts to about
Diplomacy is the art of saying "Nice doggie" until you can find a rock. Will Rogers
/. a sympatico cable modem. Next time use your head :) - If he caught it in time before his poor innocent home pc burned out he would have changed the ip on dyndns.org... if not you've just used his bandwidth quota for the month.
Then, with your own two hands, build somthing better than this. Show me a better tool for building short-run products, or prototypes. Hell, just build a CNC mill. If you can make it all by yourself, i will be impressed.
If you can, show us.
If you cant, then maybe you should try to be a little more polite.
What? Me? Worry?
http://www.taigtools.com/
This is one of the cooler "News for Nerds" articles I've seen in a while and its completely unreachable at 4 in the morning because of the slashdotting.
this is no longer humorous. As much as I often enjoy the +4 comments on certain articles reading slashdot is pretty much no longer worth the frustration of not being able to RTFA.
Can't wait till this article moves down the frontpage.
-
Last time I check the T-tech, the startup miller is about $8,000. That's the heavy for my pocket, is there any other work-around?
This is quite cool.. I wonder if you could hook up a drill to a normal flatbed plotter.. You can find old Roland plotters quite cheap on ebay and places now..
"Never let the truth get in the way of a good story..."
I absolutely have to put in a plug for TurboCNC.
It really works on machines down to 486-SX33, it has a HUGE and growing user base and support network, it is actively being developed (version 4 almost out!). Best of all, it is free. As in, you can download a the full working program and use it as much as you like. If you want the source code, it's $20. That's a much better deal than any other functional CNC program out there, and you think you'll ever see the source?
EMC is out there, but...forget it. It's clumsy, fragmented, and requires a fast computer. TurboCNC can directly drive steppers at 20KHz...there's even phase drive available, which makes driving bipolar steppers as simple as wiring up a bridge chip and a power supply.
Check it out at DAK engineering. Also take a look at the Yahoo! group at turbocnc.
...
I'm not sure if it's chain wide but the Radio Shacks around here are selling off the last of their stock in electronic components. The staff tell me that they'll no longer be carrying them.
So, if the shack is your only local source of simple electronic components, you'd better stock up now. (Or find a good mail order source)
I'd consider dumping $1000 and a couple months of "spare" time into building my own miniature end to end fab to serve me the rest of my life, but it seems kind of pointless without two/three features which seem to becoming exponentially more important: bga capabilities and multilayer.
What would be required for BGA work? what sort of accuracy? does BGA automatically imply multilayer, or can you route between the contact patches? what sort of "fill rate" do the solder balls form? do you need to add more solder, or do the solder balls themselves suffice?
Also, does anyone have thoughts for a BGA capable DIY soldering system? you'd have to hold the parts in place, i assume? heating elements: what's used by the big boys? how can you get a system that would provide the appropriate ramp up in temperature from that system? pizza ovens come to mind, but maintaining the constant and low delta-t has been described as an absolute requisite.
moving on to multilayer...
*gulp*.
i'd think this would vaguely just be like solder masking - with which i happen to have no experience either. just mask a nonconductive coating, then glue. what adhessive is used? what is done for interboard connectivity?
and where do you find thin PCB for use in such a system? somehow those 6 layer nvidia boards (random guess) are just as thin as my single layer ISA serial cards (bless them both).
Myren
The CNC system here http://www.cocopuff.net/cnc is hopefully going to be an inexpensive cnc system aimed at enthusiasts, hobbyists, and educational institutes. The photo's located at the above URL are a prototype of our cnc machine, with the final version being slightly smaller (12"x12"x5" working area). We're planning on offering it at sub 2k prices, including all motors, controllers, and software. Should begin production sometime early this summer. Questions and comments can be directed to mbreitba@shanje.com -Flash
Once in a great while, we got a board with a funny pattern for some traces, and he pointed out those were actual circuitry. He said that at high enough frequencies, an engineer could play with the trace pattern and fiddle with impedence or frequency attenuation.
Lastly he said of course we can call them whatever we want - but whatever the customer says, goes. The customer was always right. :-)
"The most sensible request of government we make is not, "Do something!" But "Quit it!"
I've always thought that CNC/rapid prototyping was pretty cool. The ability to take a design from CAD and manufacture a prototype that is true to the original design is cool in and of itself, not to mention the time it saves. I also like the artistic side to it. You can take complex shapes and lines, or a picture used as a depth map, and create a "carving" from it with great precision.
My current CNC mill is a MaxNC NC-10. At around $1,300, it was the best deal I could find. Its a little slow, and not the best thing by any means, but works nicely. I'm going to be selling engraved desk signs, I'm going to use it for engraving. I'll probably eventually do 3D image engravings(depth mapping), but at first I'll just do lettering.
I've made a couple keychains of Tux, some 3D statuettes, several more keychains, some lettering on wood, some robotics parts, etc... They all look really nice!
Actually, my biggest complaint about that machine(the NC-10) is that the effective X work distance is only about 6 or 7 inches. Its big enough for small work, like engraving signs, but I'm thinking about building a much bigger CNC router with a 5x10ft(or something) table sometime.
One thing I have to say to everyone who is going to buy or build a mill. Software for these can cost over $1000! You need the CAD program to design the item, the CAM program to create G-code from it, and the actual CNC program which will come with the mill, or be freely available. MecSoft did offer a free "lite" version of their CAM program, but don't seem to anymore(Luckily, I have it archived), but that is what I'm using.
Have tried vinyl and the material is not destroyed by the normal RS etching solution.
Found a printer late last year that has a strate paper path. A Lexmark i3. Cramed a CD and it actualy printed on the top of it. Did try a test with paper and printed test in some etching solution and it seam to pass ok. Now if I can get some time will cram a pc board and print out a circuit! Be nice if I could replace the std ink with a copper based and just print the layout.
I have an old XY plotter that seems to have enough clearance for PCB. I just ran it over a piece of crappy perfboard that looks to be about 1/16 of an inch thick.
You could either draw the traces and etch away the rest, or you do as my grandfather did in building a computer back in the 60's and that is to deposite a 1/64" x 1/64" silver trace onto the board. It was one of his achievements that his company gave him an award for.
Would it make sense (for strictly home use, no ultra-high precision required, just enough to attach normal surface-mounted chips) to build such a machine without step motors, just with manual dials, driven manually - so you follow a drawing on the board while etching, lift the drill manually, move the board by turning the dials etc? I have to admit my etched circuits with tracks thinner than some 3mm were a complete failure, the paths either shorted or broken (or both).
45 5F E1 04 22 CA 29 C4 93 3F 95 05 2B 79 2A B2
Another consideration is that circuits dealing with very small signals (i.e. RF or nanovolt amps) will be all over the place with an etched board (remnants of the etching process). For several years, I worked for an electronic scale company where we would build custom load cell trimming boxes. The leakage from a hand-etched board relegated most of these to the "this is what it will look like" collection, and not something you could actually install at a job site for beta feedback.
In contrast, milling the board would give you similar turnaround time (less than an hour or so) but without the leakage. At the time I was doing this (about 10 years ago), I was printing out negatives on laser transparencies, photo-etching directly from the printer output, and then straight to the drilling machine. In one case, I was able to prototype a new trimming box for a customer within 2 hours, including layout & board stuffing time.
Being able to produce a milled board in an hour or so would be pretty groovy, but I have to think a pen plotter & Dremel tool would be more accurate.