[bkinman]

I'm wondering what methods you folks used to make your own circuit boards for your CNC machines... I am very intersted in building myself a PICstep driver, but I have had some pretty poor results with many techniques of creating circuit boards.... There has to be an easier way for those among us who are broke to produce our own PCBs.

Following are the results i have had with different methods of PCB making.

Toner Transfer:
I have tried this methods staples brand picture paper, and have had very poor results. I'm not sure if the iron that I am using to transfer the toner is hot enough (Although I suspect it is). Also, there is a possibility that the laserjet xerox that i am using to produce my artwork is not fusing enough toner to the paper to get very good results... I have tried immersing the board in warm soapy water for ten minutes or so, and then proceed to peel the paper away from the PCB by using a toohbrush, and eventually, i end up with most of the paper off (except for the paper still sticking to the toner)... But I still have a problem in that some of my traces always end up damaged.

I have also tried transfering from kinkos' transparencies.... This method actually worked better than i expected... I have not heard many people speak of it using transparencies for toner transfer, but this method, if it can be proven to be effective, is a lot quicker than dealing with picture paper. However, the transparency has a tendancy to melt if the iron is kept in one place for too long, or the image can get warped and distorted if the transparency gets too hot as well.... There are a few tricks to even get this method to work at all. You have to pull the transparency away from the PCB while it is still very hot, so you may end up burning yourself. If for some reason the toner is not dense enough on the PCB, or part of the toner did not stick because it was too cool, it may come up, and you may have to use another transparency placed precisely over the top of the toner that did transfer to get it to work.... The upside to this method is that it seems to work with finer traces than the picture paper transfer method.

I wish that I could afford to use techniks' "press n peel blue" transfer paper... But this is way to expensive for me to afford (I am a poor college student). Also, they do not sell this paper near where i live.

Photo Etching
After messing with the toner transfer method for awhile, I decided that perhaps it would be a good idea to give photo sensitive boards a try... I purchased some boards that were pre coated with the photosensitive chemical. They were a bit more expensive, but i figured that if this method worked well, i would just bite the bullet and give up on toner transfer forever. This method requires additional consumables... I had to purchase the developer, that would make everything but the artwork water soluble, and I also decided to buy ammonium persulfate to use as my etchant, because it is supposedly better for close traces (as exist on the PICStep boards). Also, I purchased a 200W incandescant bulb, and a desk lamp with an adjustible arm which I intended to use to expose the photo sensitive PCB with.... It turns out this this whole method was a waste of time... I either over-exposed or under exposed the PCB.... Or perhaps the 2 transparencies I used stacked upon one another did not act as a proper mask (I doubt this). I exposed the board for perhaps 10 minutes under the incandescant bulb, which was 10 inches away from the board.... Another failing of this method may be the fact that i used the incandescant bulb. I have heard of other people just exposing their board under sunlight, but they never mentioned how long they exposed it for... I have also heard people say that incandescent bulbs worked just fine for them... Perhaps those people had higher wattage bulbs, or they exposed for much longer than the standard time.... In either case, i had to scrub the board and just use it as a standard copper clad board.... I bet this method would have worked perfectly if i had used the reccomended flourescent bulbs to expose the board with.... But this would have meant spending a $$$ on an UV exposure unit or building my own exposure unit using germicidal UV bulbs... Both of these cost quite a bit of money and were not really a good option. Especially considering the fact that i live in a small apartment, and the girlfriend probably wouldnt appreciate any more of my "mad scientist" junk laying all over the place.

Ammonium Persulfate
As sort of a side note, has anybody really had any luck with this stuff. I hear people say that it works a lot better than ferric chloride, but I could not get mine to work very well, even when it was agitated by hand and heated...

Xylene
I read somewhere that somebody was using colorless blending markers containing xylene to chemically transfer the toner from the paper to PCB... I have tried this method and am under the belief that the person could never have gotten this method to work well. First of all, the toner doesn't tranfer to the copper in a dense enough fasion that you could etch with FeCl, and second of all, I have a problem with the artwork smearing on the PCB... Perhaps there is something I do not know abou this method. I have tried using it in conjunction with toner transfer with no success. (By bleeding the blending marker through the back of the paper after using the iron)... Also, xylene is very powerfull stuff, and it is hard to avoid breathing it in. I am sure that it is not very good for your brain cells. ...

(Xylene would probably be a great way to print words on the top of your PCBs where the through-hole components are mounted!)

Ferric Chloride
Good stuff... Cheap and available in most every town... Something that actually works well!

Transparent Laquer
I have been concerned about how to prevent the copper traces from oxidizing... I cannot afford any of the expensive chemicals from MG, so I gave this a try (This is before I understood flux and tinning). The laquer that i bought from wal-mart works pretty well... Although you cannot solder through it (I hear you can do this with some laquers)... If you don't plan on making any more modifications to your board, and do not feel like 'goin through the effort of tinning your board, i would reccomend purchasing some of this stuff.

I have also just today heard of SMT screens which are used for applying solder paste before you send your boards in the reflow oven... I wonder if you can order stencil's like this for applying PCB traces to the board using wax, or perhaps some sort of spray enamel. If anybody knows that would be great.

I have heard that magazine paper works well for toner transfer. I hope to give that a try this weekend. Hopefully they won't catch me putting the magazine paper through the xerox machines at my local kinkos. :wee:

So... That is my experience with making circuit boards so far. Man! Talk about a pain in the a*$. I have learned a lot through trial and error... And for some reason I still cannot justify sending my art to a board house.... I guess i'm just a renaissance man at heart.

Perhaps some of you have some insight on the methods mentioned above, or have another exciting technique for getting the job done.

-bkinman
(first real post!) :banana:

[drawbar]

I, too, began experimenting with toner transfer. My best results with this method were with the clear transparency sheets you get from Staples, etc. Copy or laser print as dark as you can onto one of these and iron it on. Clean the copper very well before you iron. I used Scotch Brite and then rubbing alcohol. Most of the time this method worked pretty well. I almost always had to use an etch resist pen to fix traces that didn't transfer completely, though.

I also, like you, tried out some pre-sensitized boards. This was an easy switch for me because I was already printing onto transparencies for toner transfer. The most important thing I learned here is that exposure time is very very very very (did I mention very) important here. Under expose and you end up with copper remaining where you don't want it, over expose and the light begins to expose under the printed lines and your traces get thin. Experimentation is a must, and once you get the correct time for your setup, be consistent. Just so you know, I was using a 100W incandescent bulb in one of those clamp-on fixtures with the silver bowl-shaped reflector, straight down onto the board about 10 or so inches away. Yes, the pre-sensitized method is more expensive because of the costlier boards and the extra developer, but for me, when I got it right, the trace quality difference was spectacular. I don't think I will ever go back to toner transfer again unless this method becomes unavailable.

FYI, I have always etched with FeCl. Radio Smack still sells it.

And I have to learn how to properly coat my boards when finished. Unprotected copper always corrodes and looks cruddy after a while. Conformal coating seems to ring a bell as the thing to try.

Brian

[Rhodan]

I use the photo method with Ammonium Persulphate. I tried toner transfer with the fancy paper etc and could never get it right - though I do still use toner transfer after etching to put the printing etc from the silkscreen layer onto the board.

I started with Ferric chloride but found it quite unpleasant to work with and its also much harder on the environment than AP, so I tried AP and never went back.

I make double sided boards with 10mil traces and surface mount components regularly and most of the time a bad board is because of a bad design, not because of the etching process.

I don't use an incandescent bulb, I use a 2 foot florescent with a "full spectrum" bulb.

You mentioned using two transparencies - I cannot see how this would work. The most important thing is to ket the toner dead flat against the board so light cannot get underneath the traces. To do this I print mirrored onto the transparency for the topside of the board so I can place the toner side against the board then cover it with a heavy piece of plexiglass and check for shadows under the traces - a shadow means the trace is not flat against the board.

Plexi can be a bit curved so you have to check for shadows and decide which side is better. Plate glass is better as its usually dead flat.

I never have problems with over-exposure and I'm not all that careful with exposure times. Typically I start the process then go make coffee etc - I just make sure I give it enough time to not under-expose! I've gone as long as 25 minutes though I find 15 minutes works well - the shortest exposure that has worked for me is 10 minutes but I give it plenty of time to make sure.

What are you using to print the artwork? Inkjets are notorious for leaving pinholes in the artwork which light gets through. I use a cheap little Samsung ML-1200 laser printed set on "darkest" and its never let me down. An easy way to tell is just hold the artwork up to a light and see if you can see pinholes - if so, consider buying a different printer or, if you are using "bargain" toners/inks, try a brand name one and see if that improves it.

Although the photo method works well it does take quite a bit of time and its a real pisser when you find out the board is no good because you inadertantly scratched the resist is a critical spot.

This is the reason I'm building a CNC machine - to mechanically etch boards! If I have a bad design, I wasted $2 in copper board as apposed to $20 in pre-sensitized board (double sided of course). I'm hoping my first machine (which will be used to build parts for the second, more accurate machine) will be accurate enough to drill holes. One board I made has 280ish holes in three different sizes - that took a LONG time with an HSS bit on my little delta drill-press!

Anyway - to summarize the imporant bits:
- Use ONE transparency and print so that the toner is against the board to keep light from getting underneath the traces.
- Print using the darkest setting on your printer and check for pinholes
- Buy a cheap laser printer (I saw one for $80 at Future Shop (Best Buy in the States)
- Get a heavy piece of plexiglass to hold down the artwork.

Good luck!

Oh, and also join this yahoo group, they're great!

http://tech.groups.yahoo.com/group/Homebrew_PCBs

[bkinman]

What are you using to print the artwork?

I am currently printing out the artrwork on an inkjet printer, and then going to kinkos and photocopying onto a transparency.


I think that perhaps my UV method went wrong because I did not expose it long enough. It sounds like I could probably get better results if I exposed for 45 minutes or so...

How long did you expose for drawbar?

[bkinman]

This guy uses a pretty cool technique... He modified an HP plotter to plot the traces directly onto the PCB... It would be nice to get my hands on one of these old HP7440A plotters. Has anyone heard of or done this?

[BCwanderer]

There is a AVR (Picstep like clone) PC board and programmed chip for sale on Ebay today (Sept 9) at
DIY Stepper Motor Driver, microstepping, 55V, 3 Amp
Item number: 270025011340
Here is the sellers website
http://www.bright.net/~agarb/STMD/AVR_STMD.html
It looks to be a very nice board and chip combo for a good price.
He sells multiples on his website

[smarbaga]

i have used all the methods discribed above.
1. as for toner transfer the best result i had was the water soluable paper.
sorry i have no pics but some boards were used as 455 mhz transievers with surface mount stuff , lots of parts
http://www.techniks.com
2. i also got the stuff to make silk screens, this used razol diazol , photo imultion,
the screens were streched on a home made stretcher as tight as a drum skin
the stuff cost about $120 us including etch resist, screen ... everything.
this worked real good, no prob. getting traces between chip pads.
i could pring 20 boards in about 10 minuits with as much cleanup time.
never used amonium per........ only pre heated ferric chloride with a fish bowl air stone for agitation
etching would take about 5 minuits, for however many boards i had in the bath.
got some of that chloride stuff in my ete once, just a minute, minute speck.
it was instant blindness and directly to the hospital for a sailene bath,
doctor said it was close to burning my eye , but no damage , wheeew..

[kn6za]

I have been using my cnc machine to make circuit boards for about 3 years, and I like the process very much. I design the board in my cam system, mirror the design, assign tool path to include trace isolation, hole drilling and outside shape contour routing in one setup. Flip board over run trace isolation for 2nd side of board, and its done.

This process has its own share of dificulties. The first is, I am the "auto router" I have not been using board design software such as eagle, so layout can take a bunch of time.

I want to try using gerber and excelon files to generate the gcode for the board milling/drilling process, but have not gotten to it yet. I know it is being done, just need to find the time to get it figured out.

Second is keeping the board mounted flat in the cnc machine. Double sided tape, with two or more screws located inside the board has worked the best, I will be designing a vacuum system soon, so I hope this will help solve this dificulty.

I built the cnc machine and it works great for this application. I run 10000 rpm and 10 ipm trace routing, and 50 ipm drilling.

hope this helps,
Andrew

[bkinman]

I am going to try the toner transfer method using magazine paper this weekend, hopefully there will be pictures =).

As for using a cnc machine to route boards, that is the end goal, but I'm actually etching the boards for the driver of my first CNC machine.

[davesaudio]

Magazine paper works well for me, try different kinds.
Never lift, soak, sometimes rub the exposed side of the attached paper with a scotchbrite carefully to help water penetrate the clay coating,
rub sideways only with my thumb, finish w toothbrush.

[bkinman]

Well Sh*t! I tried doing the photo method one more time, and had terrible results.... This time I overexposed. I exposed for about 35 minutes in the hot santa cruz sun.... Perhaps my problem was that my transparency was not pressing close enough to the PCB, and sime light got in around the sides. I would have liked to sandwich the board between some glass sheets, but I had none around.


I then washed the circuit board off and gave the toner transfer method a try with magazine paper. I used the cover off of a magazine and had a bit of a problem with the ink from the cover transfering to the copper as well...

I think when i try the magazine method again, i will use a thinner piece of paper. I am noticing that thick stock papers gets bubbles....


This board layout that i am working on is a very hard board to etch... I am doing Joe's version of Pminmos 3977 driver. Has anyone else etched this board?

[ServoGuy]

Well Sh*t! I tried doing the photo method one more time, and had terrible results.... This time I overexposed. I exposed for about 35 minutes in the hot santa cruz sun.... Perhaps my problem was that my transparency was not pressing close enough to the PCB, and sime light got in around the sides. I would have liked to sandwich the board between some glass sheets, but I had none around.


I then washed the circuit board off and gave the toner transfer method a try with magazine paper. I used the cover off of a magazine and had a bit of a problem with the ink from the cover transfering to the copper as well...

I think when i try the magazine method again, i will use a thinner piece of paper. I am noticing that thick stock papers gets bubbles....


This board layout that i am working on is a very hard board to etch... I am doing Joe's version of Pminmos 3977 driver. Has anyone else etched this board?

You will want to fill those pads and vias, either with software or Sharpie marker.
The holes are much too big, and the traces around them are way thin, that will cause lots of pain for you when you are drilling and soldering. Just fill the holes and they will still be easy to drill out later. When I make my own boards, I use 10-12mil holes as a guide for the drill bit.

BTW: I am easily able to get 8mils for trace and 8mils space using the blue Tecniks film. I use a laminator set on high to fuse the film to the boards,and that seems to work just fine. The boards must be super clean and pre-etched for any of the toner tranfer methods to work well. Preheating the boards or a couple of passes thru the laminator help as well.

[bkinman]

Thanks Servoguy, filling in the pads is a really good idea that I had not thought of.... I think I will just have to break down and buy the press-n-peel. Also, I should see if i can get myself laminator.

[pminmo]

Heres an update with the pads filled in for drilling, that may help. I've always thought tis was to aggressive for toner transfer but those who have the process down seem to be able too.

[bkinman]

Thanks Pminmo! If I am able to get my hands on some of that techniks business i will give it another go using the pdf provided.... Perhaps I will try spacing things out so that i can build this board with the magazine method.

[pminmo]

I've tried the techniks, got better results with glossy magazine paper.

[bkinman]

Pminmo, Do you use the cover stock, or the thin pages?

[smarbaga]

i have tried both the press_and _peel and the water soluable method.
the water soluable method worked best for me.

1. print the image on a piece of plain paper.
2. cut and scotch tape the techniks paper over the image on the plain paper, about an inch extra on all sides. ( this saves expensive techniks paper ).
3. run the paper through the laser printer again.
4. used a regular house iron set on low to fuse the techniks paper to the pcb.
5. dropped it in a bowl of warm water.
6. after about 2 minuits the paper starts to float, the laser toner ink stays
on the copper pc board.
- this paper has an invisable water soluable film that disolves and separates the paper from the toner image that has been re-melted via the iron and stuck to the copper pc board.
- the trouble is the paper is about $1.00 a sheet (8.5x11) , i think min. 20 sheets per order.

[pminmo]

I find the glossiest page of paper in the magazine I can find. I had much better luck with magazine paper than any new "white" printer paper.

[bkinman]

i have tried both the press_and _peel and the water soluable method.
- the trouble is the paper is about $1.00 a sheet (8.5x11) , i think min. 20 sheets per order.

What is the name of this water soluble paper? Or is this the techniks you are talking about?