making pc boards

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Finally I had some success with the paper/iron method. I now have 2 almost perfect pcb’s for my aleph 3.
My problem was that the iron I used wasn’t perfectly flat, resulting in poor transfer on some spots. For those interested here is what I did:
First I placed the paper (HP glossy photo paper) with the toner facing up on the table. Then I put the pcb, copper side down on the paper, carefully aligning it. Then I took some heavy-duty tape over the back of the pcb and paper so that it would not move. Then I turned it over, paper facing up. Then I proceeded, with the iron set to max, rubbing back and forth, laying my weight on the iron. I carried on like this for a few minutes. Then I quickly cooled the board in cold water. And finally I soaked the board in soap and water.
There was very little touch-up to do. :)
 
I have had good success with Olimex. I recommend using eagle for the design. They charge $21 for 100x160mm single sided or $26 for double sided with plated though holes. Shipping is $8. They are in Bulgaria so you get your boards in about 3 weeks. I havn't found a US place I like yet. The boards I had done had TSSOP's and TQFP's and it was no problem.

http://www.olimex.com

http://www.cadsoft.de

Download the freeware. The only limitations are 100x80mm board size limit and 2 layers.

Darrell Harmon
 
:att'n: Newbie alert :att'n:


Figured I'd post my method for making homebrew PCB's and etching them. :) I do schematic and board layout in CadSoft's Eagle. It has its limitations, but has sufficed for me so far. When I've got everything done, I create a ground plane and pour it over the bottom of the board, and hatch it with a .125in spacing and 10mil width with .25 isolation. This step seems to be important when doing toner transfer ...

I then print the page, one pass, through my HP LaserJet 1100. For a printing medium, I use pages torn out of a magazine. Specifically, I use some Microsoft Tech Journal, but it's just regular magazine paper. The layout is printed right over the words and everything -- the ink on the page does not affect the layout at all.

I pre-heat a standard WalMart iron to "Cotton" (hottest) with no steam, and next to it I have a bowl full of water. I also use a bamboo cutting board (just what I happen to have) -- a firm, heat-resistant, and insulative surface is necessary. From this point on, I wear thin latex gloves. Fingerprints are my enemy in this case. The circuit board is cleaned with a paper-towel scrub of acetone until no more color comes off. I use RadioShack boards, they seem to work the best.

I pre-heat the board for 30 seconds with the iron. I remove the iron, lay the cut-to-size magazine page down, and immediately tack it in place with the iron. The page will try to curl instantly, so tacking it in place is important. From there, it's just constant pressure and constant rubbing around. You should push about double the weight of the iron against the board. Any more than that, and you run the risk of smushing your traces wider than you wanted. After about two minutes of ironing, lift the butt end of the iron up and gently (weight of the iron) scrub the entire board. The idea here is to get one last higher-pressure pass to ensure everything is down.

I quickly put the board into the bowl of water and let it soak for about 5 minutes. After it's cooled completely down, and the entirety of the paper is soaked, I then put it under a stream of running water and begin to rub the paper off the back with my (now ungloved) thumb. Rubbing the paper off takes about 5 minutes, and does not require a great deal of care -- provided you use your fingers only to remove paper. They seem to be stronger than the paper but weaker than the toner. Make sure to get down between all the pads and traces ... the pour from earlier will serve to make your life easier during all this process. It seems to help even out the heat from the iron while at the same time not making giant planes that are difficult to transfer.

Then, the board is set aside to dry as I prepare the etchant. I use 2 parts hardware-store muratic acid mixed with 1 part 35% food grade hydrogen peroxide (thank you, internet) and 1 part plain ol' tap water. Enough of this is prepared to fill a Pyrex baking dish about 1/2", and stirred around with a plastic straw.

The board is dropped into the etchant, and immediately agitated with aforementioned scientific bendy straw. This is a VERY powerful etchant (but still controllable), and can remove all the copper from an unmasked 4x6" board in 5 minutes or less. The pour from earlier will again help you, as the etchant has less work to do.

Once all the copper is etched (feel free to brush the bendy straw over unetched parts to speed it up), immediately remove the board and rinse for 30 seconds under running tap water. Neutralize the etchant with a few tablespoons of baking soda and wash down the drain.

I then tin plate the board with TINNIT. It's a straightforward (albeit annoying) process.

The holes are drilled out with carbide bits and my Dremel in my Dremel drill press thingy. I got a few grab-bags of carbide bits from Harbor Freight. I typically use the .85mm bit, and keep the speed about 20k RPM.

I currently don't know how to solder mask or silk-screen. For the silk screen, I've just done the toner-transfer method on the topside of the board to label components, but I'd like something more official if someone knows one. A solder mask / greenscreen on the bottom would be awesome if anyone has ideas -- that would save me the tinning step.

Anyway, wall of text and such. Hope it helps someone :) I haven't tested it on boards larger than my iron (4x6" and smaller).
 
Re: Question:

my 4p worked very well for this the last time i tried making a board using this technique. i used the back of an avery labels page - it worked better for me than the blue press and peel. also, at the time, i was using an original hp toner cartridge - that may make a difference

mlloyd1

LAJ said:
Do you think a Laserjet 4+ printer will work?
 
I use commercial laser transfer paper for hard surfaces, a colour laser printer and an Iron :) Etch in the kitchen sink...

Colour lasers seem to help as they lay down composite black (4 layers of polyester I think) which makes a nice resist. The Transfer paper peels clean with good definition.

I won't post the product as I work for a representative of the maker. If the Mods say its OK then I'll link the core maker.

Alan
 
Yes, it will work fine. Crank the toner density, turn off all "Draft" and "Saver" modes, pump the DPI to the max.

Experiment with a few types of magazines -- you'll find one that works. If you mess up on the ironing part, acetone will take the toner off and let you try again and again. It's a heck of a lot cheaper than Press'n'Peel and gives better results.


Also: I've been using Linux/GNU systems for a few years. Feel free to PM or start another thread in here if you'd like help and guidance! :)
 
It is one thing to make a functional PCB that reliably conducts signals as needed at expected impedance. It is another thing entirely to have a really good-looking board to present to colleagues, friends, or potential employers, investors or customers. “Home made” boards may also not stand the test of time, with inherent corrosion processes that occur from residual chemical traces and moisture intrusion.

While I am not associated with Advanced Circuits in any way, I have successfully ordered from them in the past. Their deal has changed slightly from what it was traditionally. Check their website for details. I gather they currently offer a double sided PCB with resist and silk screen on fr4 equivalent material for $33 each if you buy [4] of them, ($132 total), and their special adds a [fifth] one for free? It used to be [3] for $99, but it’s still a screaming deal. This includes all NRE and setups charges, and in my case there was some design verification for net lists and pad terminations for Gerber files at no additional charge. The resulting product is gorgeous, long lasting and approaching mil spec for fabrication dimensional accuracy and material quality.

I also saw mention of student discounts. Not sure on any details.

http://www.4pcb.com/

http://www.4pcb.com/index.php?load=content&page_id=130

I used to go to Parks Medical for my proto runs, and there it was around $600 for a flat of boards that yielded about 20 smallish cards. Even that wasn’t too bad a way to go, especially for a limited production run. I don’t know if they are still offering their small run service, but you could give them a call in Oregon.

http://www.parksmed.com/
 
I have a circuit board I did with the dry transfer method in 1985 for a tube amp high level section that still looks fine. I used Tinit to coat it. A test board I did not coat shows some greening around the edges. I've seen boards with solder completely covering the traces. Actually my old HP test equipment has boards like that. Also heard that Deoxit or Cramolin would prevent corrosion.
Dave
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.