What is the best way to make printed circuit boards at home?

[james_s]

Did you know that you can do toner transfer using a laser printer, printing out your design on the cheapest possible paper, (junk mail works best!) then scuff up your PCB with some steel wool to get it abraded pretty well, then use the iron to transfer the toner to the PCB, using a flat piece of wood as your ironing board. and for the etchant- use hydrogen peroxide, vinegar and salt.

I started out using junkmail paper for toner transfer and it worked ok, but not nearly as well as Press n Peel film. I often found that the ink on the paper would fill in the pores in the toner and result in perfect etches while the un-printed areas would get a texture to them. In many cases images and printing from the paper page would end up etched on the board. The blue film is consistent.

[SilverSolder]

+1 for using "proper" toner transfer paper  - I use the Pulsar toner transfer paper and it is extremely reliable, cheap enough that it isn't a big deal to use.

My printer will accept the "abuse" of taping a small square of transfer paper onto a letter sized sheet of paper, if I am doing a small board.  This way, the transfer paper is super economical to use.   Basically, I print the small board on a normal sheet of paper first.  Then I tape a square of transfer paper over the area where the print is on the paper, using very thin tape.  Then I run the page through the printer again, this time it gets printed exactly on the toner transfer paper.  Winner, winner, chicken dinner!

[cdev]

Come to think of it that has happened to me at least twice (where the ink from the junk mail behaved like toner)

.One time I cleaned it up and the other tie (with an antenna) where it just looked like it would be a PITA, I just redit it.

To be honest with you Ive never used any purpose made products intended for PCB toner transfer. I should try that.

Also, Ive never made a board that requires very very thin traces.

Text has turned out okay.

Did you know that you can do toner transfer using a laser printer, printing out your design on the cheapest possible paper, (junk mail works best!) then scuff up your PCB with some steel wool to get it abraded pretty well, then use the iron to transfer the toner to the PCB, using a flat piece of wood as your ironing board. and for the etchant- use hydrogen peroxide, vinegar and salt.

I started out using junkmail paper for toner transfer and it worked ok, but not nearly as well as Press n Peel film. I often found that the ink on the paper would fill in the pores in the toner and result in perfect etches while the un-printed areas would get a texture to them. In many cases images and printing from the paper page would end up etched on the board. The blue film is consistent.

[cdev]

What is the Pulsar papers shelf life?  (although I guess if it is super cheap it really doesn't matter.)

+1 for using "proper" toner transfer paper  - I use the Pulsar toner transfer paper and it is extremely reliable, cheap enough that it isn't a big deal to use.

My printer will accept the "abuse" of taping a small square of transfer paper onto a letter sized sheet of paper, if I am doing a small board.  This way, the transfer paper is super economical to use.   Basically, I print the small board on a normal sheet of paper first.  Then I tape a square of transfer paper over the area where the print is on the paper, using very thin tape.  Then I run the page through the printer again, this time it gets printed exactly on the toner transfer paper.  Winner, winner, chicken dinner!

[SilverSolder]

What is the Pulsar papers shelf life?  (although I guess if it is super cheap it really doesn't matter.)

+1 for using "proper" toner transfer paper  - I use the Pulsar toner transfer paper and it is extremely reliable, cheap enough that it isn't a big deal to use.

My printer will accept the "abuse" of taping a small square of transfer paper onto a letter sized sheet of paper, if I am doing a small board.  This way, the transfer paper is super economical to use.   Basically, I print the small board on a normal sheet of paper first.  Then I tape a square of transfer paper over the area where the print is on the paper, using very thin tape.  Then I run the page through the printer again, this time it gets printed exactly on the toner transfer paper.  Winner, winner, chicken dinner!

I am not aware that it even has a shelf life in practice - as long as it is stored in dry conditions, like any paper, of course.  I have about a dozen sheets here that are at least 5 years old, as good as new.

[cdev]

I was thinking about this clay coated inkjet photography paper I have (which might even work for toner transfer) It lasts a couple of years but not forever.

I am so ridicuously cheap sometimes.

What is the Pulsar papers shelf life?  (although I guess if it is super cheap it really doesn't matter.)

+1 for using "proper" toner transfer paper  - I use the Pulsar toner transfer paper and it is extremely reliable, cheap enough that it isn't a big deal to use.

My printer will accept the "abuse" of taping a small square of transfer paper onto a letter sized sheet of paper, if I am doing a small board.  This way, the transfer paper is super economical to use.   Basically, I print the small board on a normal sheet of paper first.  Then I tape a square of transfer paper over the area where the print is on the paper, using very thin tape.  Then I run the page through the printer again, this time it gets printed exactly on the toner transfer paper.  Winner, winner, chicken dinner!

I am not aware that it even has a shelf life in practice - as long as it is stored in dry conditions, like any paper, of course.  I have about a dozen sheets here that are at least 5 years old, as good as new.

Thats actually a good example of my cheapness. I very rarely use air conditioning, unless its over 80 or so or really really humid. Things don't get moldy, usually, its not that bad (I do have a dehumidifier in the basement)  but certain things dont do so well in high humidity.

[james_s]

The Press n Peel (or whatever) transfer film looks expensive but in practice a package of it goes a LONG way, I'm only on the second pack I've ever bought and that's only because I helped a friend etch about a dozen boards he needed right away and didn't have time to send out for.

What I do is print out my layout on ordinary printer paper, then I cut out a piece of transfer film just slightly larger than the layout and tape it along the leading edge right over the pattern I just printed. Then I feed it in the manual feed tray and hit print again, printing the same design on the same sheet, this time onto the piece of transfer film. By doing this I can get a dozen or more small boards out of a single sheet of film.

*edit: I just noticed that SilverSolder already said almost exactly the same thing. I've done the tape trick with a bunch of different printers and never had any trouble with it. As with most things though YMMV.

[SilverSolder]

We are all economical minds here, it seems! 

[Jwillis]

What really works well with toner transfer is the paper backing from adhesive vinyl used for signs . I could get that stuff for free from a local sign maker and cut it into 8 x 11 sheets for the printer.

[Whales]

I've had reasonable success using the waxy-paper used on rolls of book contact.  Cheap and abundant.  I just cut it roughly down to size and my printer is fine with it.

People wonder why I buy book contact only to throw all of the useful half out  

[Raj]

Oh, after you are finished etching there is a black gunk on your boards, which you need to clean off with acetone or nail polish remover, be careful not to get it on any plastics

I just drill the holes then use scotch brite with soapy water...polishes the copper and chamfers the hole, and sometimes it pushes the copper around the hole into the hole. The end result is nice and clean holes.

[KL27x]

Making your own boards is part of the fun with electronics.
Doing traces down to 0.5mm on 1 oz copper clad is easy. 2oz takes more etching time. 
Build the circuit and PCB with KiCad or other program.
I make negative of the pattern on acetate with the toner printer . Then use the Negative Photoresists which gives a much finer detail to the PCB art than toner transfer but not as fine as positive photoresist . Flash with UV and  develop .Then use the muratic acid / Hydrogen peroxide etching process. Lots of You Tube videos on the process.

That's 20 thousandths trace width. I do 8/8, 100% success rate over an entire 6x9" board. Toner transfer.

OP, check the manufacturing thread if you want details. I posted about it in a thread over there, earlier.

This method also practically eliminated any pinholes in ground pours and rough traces. Provided you use a proper etchant. That acid peroxide stuff is very fast and convenient, but the etch is very inaccurate. It gets under and lifts resist. And if you ever try sharpie, you can't get that to work well in acid peroxide.

[james_s]

I did few small double sided boards with 0.5mm TSSOP ICs back in the day when I was trying to see just how good I could get the process. I even developed a way to make decent vias, drill a hole just large enough to fit a thin solid copper wire, flush cut on both sides as close to the board as I could get it and then crimp it with a small C-clamp. That would mushroom out the nub of wire on both sides of the board resulting in a permanent connection then normally I'd add a touch of solder just for good measure.

These days I tend to only etch my own boards when I want a simple one-off for something without waiting, I don't bother to show off anymore. I still use quite a bit of copper clad but these days it's usually for dead bug/Manhattan or whatever construction. It took a long time for me to get past the ugly nature but it has grown on me and the RF performance is very impressive. Circuits I had difficulty with before, things like radio transmitters or high frequency switching regulators, mosfet bridges and SMPS stuff works with ease. It's still ugly, but dang it works well.

[KL27x]

It's not about showing off. 8-10 mil traces are the perfect size for a 0.5mm pitch IC. Fattening traces and then routing those cows means more work, even when it's not about space/size.

8/8 works every bit as well as 20/x, if you follow my process. There's no reason for me to fatten traces other than for current handling. It doesn't improve my yield. Yield is already effectively 100%. I use 8 mil traces because that is easier and less work to layout. And I use 8 mil traces by default for most of my professionally manufactured PCBs. (My main manufacturer charges more for under 7 mils, FWIW; I've never actually tried 7 with toner transfer, but 6 mils is a no-go due to failure in the print, with my printer).

I promise you I know something very useful about toner transfer that most people don't seem to have learned, otherwise they would also do it.

*This becomes more and more important, when you do larger double-sided boards. When part of the process can fail towards the end, you will waste a whole lot of time, a lot more often than you anticipate. It's like a complex part that requires multiple steps in the machining, and there's a tricky one at the end. You will mess some of those up, and it will be not much fun trying and failing, sometimes repeatedly, repeating boring rote things just to get back to that stressful moment of truth, once again.

[Wimberleytech]

Permanent marker works as well, probably not as good as laser printed heat transfer, but it is a cheaper option

Yup...did that when I was a kid (just a couple of years ago...lol)

[Wimberleytech]

I second the toner transfer method.
I use ferric chloride.
Played around with vinegar and hydrogen peroxide, but feric chloride is faster and gives better results.

I only do single sided, and I dont do them often.

It is kinda fun.

My comment to the "dont" crowd:  cool it.  Let the guy have some fun in his garage.

[cdev]

IMHO, its just great to be able to go from a design on your screen to a PCB in a short period of time.

[james_s]

It's not about showing off. 8-10 mil traces are the perfect size for a 0.5mm pitch IC. Fattening traces and then routing those cows means more work, even when it's not about space/size.

8/8 works every bit as well as 20/x, if you follow my process. There's no reason for me to fatten traces other than for current handling. It doesn't improve my yield. Yield is already effectively 100%.( I use 8 mil traces because that is easier and less work to layout. And I use 8 mil traces by default for most of my professionally manufactured PCBs. (My main manufacturer charges more for under 7 mils, FWIW; I've never actually tried 7 with toner transfer, but 6 mils is a no-go due to failure in the print, with my printer).

I promise you I know something very useful about toner transfer that most people don't seem to have learned, otherwise they would also do it.

*This becomes more and more important, when you do larger double-sided boards. When part of the process can fail towards the end, you will waste a whole lot of time, a lot more often than you anticipate. It's like a complex part that requires multiple steps in the machining, and there's a tricky one at the end. You will not mess some of those up, and it will be not much fun trying and failing, sometimes repeatedly, repeating boring rote things just to get back to that stressful moment of truth, once again.

10 mil traces are easy enough, but I found I had to be really careful to get the pads right for the 0.5mm pitch ICs. It's possible that it was mostly a limitation of the printer I had at the time but I found the pads would end up a bit wider than I intended once it was transferred to the board. Now if I need to do an IC with such fine pitch I send out for boards. SOIC is no problem at all though, I can etch those myself easily.

[KL27x]

...I found the pads would end up a bit wider than I intended once it was transferred to the board.

This sounds like the toner-squishing that I was describing in the other thread, in manufacturing section. The transferred art gets enlarged or smeared in a non-controllable/uniform way. This results in a loss of fidelity when you exceed the melting point of the toner. This is why you always aimed for just the right temperature/heat when you did a transfer. Or maybe you didn't, and you just smeared the board and took it at that.

Pre-etching the board removes this adverse consequence. (There is another, which is fuzzy traces; too hot with many papers, and the board turns into a chia pet; the right paper solves this problem). When you pre-etch the board, you need to melt the toner a little more to adhere even just as well as with a freshly cleaned copper board. So the minimum entry fee is a little higher than a clean board. But there is no distortion of the image; there's no upper limit to the temperature or pressure you can use during the transfer. You just have to get it hot enough. The board is the next point of failure, and that's way, way up there, in temperature. And when you completely liquify the toner, it sucks into the surface of the board, completely. It resists the acid, better, without fuzzy edges or pin holes.

[Miti]

I order from PCBWay for few years but when I made my own pcbs, I didn’t have much success with toner transfer method. The best method for me was using an old Epson inkjet printer, curb side pick really, and inkjet transparent film. I printed both sides, top side mirrored, aligned and taped them, put the photo sensitive board in between them and taped the film to the board. Then I exposed the board with a simple fluorescent light for about 10 min each side, a simple spiral bulb would do it, and developed with 10% NAOH, it is sold as developer in electronic stores. I etched with ferric chloride that I used for many years before disposing off property at a haz mat disposal centre.
The results were always perfect. Cold printing didn’t create any distortion, details were amazing, perfect alignment between top and bottom, perfect fine traces once I perfected the exposure time and distance.
What I hated was that the printer head got clogged and had to be unclogged before every use, up to a point where it couldn’t be unclogged anymore.
I still have the films that I created many years ago, it literally looks like a film and the ink becomes like a gel that comes in contact with the board and doesn’t allow the light to leak underneath. I even made boards with vias under the TQFP.

Edit: Some pictures attached.

[SilverSolder]

I order from PCBWay for few years but when I made my own pcbs, I didn’t have much success with toner transfer method. The best method for me was using an old Epson inkjet printer, curb side pick really, and inkjet transparent film. I printed both sides, top side mirrored, aligned and taped them, put the photo sensitive board in between them and taped the film to the board. Then I exposed the board with a simple fluorescent light for about 10 min each side, a simple spiral bulb would do it, and developed with 10% NAOH, it is sold as developer in electronic stores. I etched with ferric chloride that I used for many years before disposing off property at a haz mat disposal centre.
The results were always perfect. Cold printing didn’t create any distortion, details were amazing, perfect alignment between top and bottom, perfect fine traces once I perfected the exposure time and distance.
What I hated was that the printer head got clogged and had to be unclogged before every use, up to a point where it couldn’t be unclogged anymore.
I still have the films that I created many years ago, it literally looks like a film and the ink becomes like a gel that comes in contact with the board and doesn’t allow the light to leak underneath. I even made boards with vias under the TQFP.

Edit: Some pictures attached.

+1 for this method, which I have also used successfully.  The "secret sauce" is the gel coated transparency sheets that absorbs dye based ink.  It doesn't work well with pigment type ink, but is extremely good with dye type ink.  With a high resolution inkjet printer, the output is near perfect.  However; it is significantly more work than toner transfer (more steps...   the copper has to be coated/laminated with photo resist, exposed, and developed) so depending on the precision you actually need, it may make sense to be able to do both methods.

Note that HP printers have the print heads integrated in the ink cartridge, so when you change the cartridge, you get a new print head as well.  Over the years, I have learned to love this principle compared to printers that have "permanent" print heads (Epson).  There always comes a point where a print head doesn't respond to cleaning any longer, and has to be changed...   this is just so much easier when you have the "disposable" print heads, in that you just stop refilling the problem cartridge and use a fresh one instead.

[cdev]

I order from PCBWay for few years but when I made my own pcbs, I didn’t have much success with toner transfer method. The best method for me was using an old Epson inkjet printer, curb side pick really, and inkjet transparent film. I printed both sides, top side mirrored, aligned and taped them, put the photo sensitive board in between them and taped the film to the board. Then I exposed the board with a simple fluorescent light for about 10 min each side, a simple spiral bulb would do it, and developed with 10% NAOH, it is sold as developer in electronic stores. I etched with ferric chloride that I used for many years before disposing off property at a haz mat disposal centre.
The results were always perfect. Cold printing didn’t create any distortion, details were amazing, perfect alignment between top and bottom, perfect fine traces once I perfected the exposure time and distance.
What I hated was that the printer head got clogged and had to be unclogged before every use, up to a point where it couldn’t be unclogged anymore.
I still have the films that I created many years ago, it literally looks like a film and the ink becomes like a gel that comes in contact with the board and doesn’t allow the light to leak underneath. I even made boards with vias under the TQFP.

Edit: Some pictures attached.

+1 for this method, which I have also used successfully.  The "secret sauce" is the gel coated transparency sheets that absorbs dye based ink.  It doesn't work well with pigment type ink, but is extremely good with dye type ink.  With a high resolution inkjet printer, the output is near perfect.  However; it is significantly more work than toner transfer (more steps...   the copper has to be coated/laminated with photo resist, exposed, and developed) so depending on the precision you actually need, it may make sense to be able to do both methods.

Note that HP printers have the print heads integrated in the ink cartridge, so when you change the cartridge, you get a new print head as well.  Over the years, I have learned to love this principle compared to printers that have "permanent" print heads (Epson).  There always comes a point where a print head doesn't respond to cleaning any longer, and has to be changed...   this is just so much easier when you have the "disposable" print heads, in that you just stop refilling the problem cartridge and use a fresh one instead.

Miti, youre right, it does look extremely sharp. Better than what Ive been able to get with toner transfer.

[Miti]

With a high resolution inkjet printer, the output is near perfect.  However; it is significantly more work than toner transfer (more steps...   the copper has to be coated/laminated with photo resist, exposed, and developed) so depending on the precision you actually need, it may make sense to be able to do both methods.

Yes, it takes more work but:
1. The nearly 100% success rate once you master the light, exposure time, transparents type, distance, etc. makes up for the toner transfer mishaps.
2. I used to buy pre-sensitized boards, I don't laminate.
3. If you do two layers, being transparent, you can align the two films perfectly, you can't do that easily with toner transfer.
4. It's cold printing so there's no distortion. I tried printing on laser transparents, using laser printer, and the misalignment between top and bottom, once the film cooled, was significant. I'm sure that paper suffers some distortion as well.
5. You can reuse them over and over again and the result should be always the same, if you follow the recipe.

Note that HP printers have the print heads integrated in the ink cartridge, so when you change the cartridge, you get a new print head as well.  Over the years, I have learned to love this principle compared to printers that have "permanent" print heads (Epson).  There always comes a point where a print head doesn't respond to cleaning any longer, and has to be changed...   this is just so much easier when you have the "disposable" print heads, in that you just stop refilling the problem cartridge and use a fresh one instead.

I did not have the same success with HP and Canon. Apparently the ink droplets from the Epson head are smaller and better controlled than the HP. Again, this was years ago and I didn't have access to the best printers probably, but this was my experience at the time. Also, I remember that Epson print options gave me better control than HP.

[KL27x]

1. The nearly 100% success rate once you master the light, exposure time, transparents type, distance, etc. makes up for the toner transfer mishaps.

You can master toner transfer to 100% success rate, too. Sounds like a lot fewer things to master, even. What kind of resolution do you get with UV at 100% success rate, out of curiosity? Sure, PCB fabs get down to 3 or 4 mils. I have heard of people getting down to 6, maybe, using UV. And this took some special attention, doubling of transparencies to make up for defects in the print. I assume the PCB fabs have some equipment to make that easier, and probably have to use ammonium based etchant to get that resolution.

3. If you do two layers, being transparent, you can align the two films perfectly, you can't do that easily with toner transfer.

I can see that being easier. You tape the two together and then slide the pcb between, right? Even so, that is not going to be perfect. It will take a little bit of judgment, the thicker the board gets.

[Miti]

You can master toner transfer to 100% success rate, too.

No doubt about that with one layer but I have some doubts with two layers. I couldn't.

Sounds like a lot fewer things to master, even.

I'm not sure about this one.

What kind of resolution do you get with UV at 100% success rate, out of curiosity? Sure, PCB fabs get down to 3 or 4 mils. I have heard of people getting down to 6, maybe, using UV. And this took some special attention, doubling of transparencies to make up for defects in the print. I assume the PCB fabs have some equipment to make that easier, and probably have to use ammonium based etchant to get that resolution.

See that 18F4550 in the attached picture? The vertical part of letter F is about 5 mil, the horizontal is less. I was looking for my test board that I made to test the limits of this method but I couldn't find it, my best (thinnest) reliable trace was 5 mil, if I remember correctly. However, I avoided traces thinner than 10 mil.
I did not use UV, I used fluorescent but you may be referring to the UV content of the fluorescent.

I can see that being easier. You tape the two together and then slide the pcb between, right? Even so, that is not going to be perfect. It will take a little bit of judgment, the thicker the board gets.

Depends on your definition of perfect. As long as all the pads align "almost perfect" it's good enough to me. The drilling is way more imperfect than the rest of the process. How do you even align the two sides with toner transfer method?

Look, I'm not advertising this method, I'm just describing my experience.
I'm happy I don't have to do it anymore...