A friend and i decided today to master the art of transferring laser printer toner onto copper-clad boards.
I've had the boards and ferric chloride for years, but never had a clear enough idea of how it was actually going to work until just this week.
That, and i dug my old nixie clock out of the closet and set it up at work, and now i want to build another. But this time i want to print my own board.
So i went with this: Build a Nixie-tube clock
Aside from how cavalier the guy is about hooking circuits up directly to mains AC (I will be using a transformer), it's a pretty neat hack. No cool PWM fading like my clock built on the nixieclocks.de board has, but, no processing power at all other than a fistfull of cmos chips. No $45 pre-programmed MCU.
And, obviously, if you look at the artwork, about the worst layout to learn how to etch your own boards with.
But today was a day for jumping in with both feet.
The sticking point (pardon the pun) that made me wary of the toner-transfer process in the past was that none of the suggestions for substrate seemed like a 100% good idea or at all frugal, usually glossy photo paper stock, usually some very specific product at $1.60 per page that has just been discontinued.
Some people say to use inkjet photo paper in your laser printer, but this stuff sometimes has a plastic gloss on it that might melt in your laser printer, ruining the fuser.
The revelation that made everything mentally fall right into place? Junk mail.
Make PCBs at home with magazine paper and your laser printer
There's been a lot of talk about using ultra-premium inkjet printer paper for toner transfer, but the reason people like it is because 8 years ago it was often coated with a clay that would absorb the ink before it could run along the fibers in the paper. The problem with that paper was that it was necessarily dusty, and running a lot of it through your laser printer could destroy your developer drum, which is why you can't find that stuff anymore - paper companies don't want to talk to angry people who ruined their laser printer by running the wrong paper through it, and there are other ways to skin that cat.
And once again, it's expensive. It's not as expensive per page as the photo paper (which is now almost all waterproof, making removal from a copper-clad board problematic), you have to buy the stuff a ream at a time. I actually bought a ream of some very expensive inkjet paper before noticing that the label explained that it no longer contains any clay so it's safe to use in a laser.
Magazine paper provides the best of both worlds. Glossy magazine paper is cheap, often free. It can have a heavy coating of clay, but it's a glossy rather than dusty coating so it probably won't stick to your developer drum.
We spent more than four hours trying to do thermal transfers of that stupid nixie clock layout today, and i think Mr. Ricci has either achieved a mastery of the technique that i have a hard time even conceiving of, or he's overly optimistic about what paper he likes.
We tried a lot of different glossy junk-mail advertisements, but what ended up working best for us, appropriately enough, was pages from old issues of Wired. The best transfer i did today was printed over an article about Google's rumored browser. Nobody was going to ever want to read this again - it was essentially garbage. The best kind of raw material to need.
The trick is, the clay gloss on the magazine paper prevents the toner from bonding with the paper. You want that gloss to be laid on thick, so that when you remove the paper you take as little of the toner off the board as possible - hopefully none of it.
The toner doesn't stick all that well to the copper. It holds on just well enough. Paper with less glaze on it pulls more toner off the board.
But more about the process:
There were more than a few pages that suggested using double-sided tape to stick the board down to your work surface with double-sided tape, then taping the paper to the work surface after you've aligned the artwork.
This proved to be the biggest load of crap we'd read on the subject of DIY thermal transfer.
It's hard to align the artwork when it's on the other side of the paper, for starters. And the heat of the iron destroys the tape, and drives cooked adhesives into your work surface. This was a DUMB idea.
Trimming the artwork to the size of the board was a mistake too, as this seemed to make it easier for the paper to shift on the face of the board during the heating process. It was best if there was up to an inch of extra paper around the edges of the artwork - and hey, the paper is free.
Applying the iron directly to the paper carrying the artwork also proved to be a bad idea, and adding one layer of heavy paper over top of the artwork to protect it didn't help much either. In either case it was easy to apply uneven pressure that would cause the traces to squish out and run together, or to cause the paper carrying the art to shift on the board.
What worked for us was to unfold a section of a newspaper (actually a pack of grocery adverts in newsprint form) on our work surface, lay the artwork face-up on the newspaper, align the (pre-cut) copper-clad board on the artwork, fold the newspaper back over (gently, without disturbing the board or artwork), and then apply the iron (set to linen) to the top of the stack (heating through, oh 10 layers of newsprint) for 30 seconds or so - just to get the paper stuck to the board - open the newspaper, flip the board and artwork over (so the copper side is now up), fold the newspaper back again, and then apply heat and moderate pressure for up to a couple minutes - again through several layers of newsprint.
It was also a good idea to check for bubbles in the transfer paper while it was still hot - they could usually be ironed flat.
Anyway, after applying enough heat and pressure (but not too much pressure!) through several layers of newsprint for long enough, we could just pick up the board piece by the excess transfer paper and drop it directly into a container of water to soak for a few minutes.
When we were experimenting with junk mail, we generally had to rub the paper off with our fingers after a long soak. It was a mess and a hassle, and a lot of toner got lifted off the copper along with it.
With sheets of Wired Magazine, after 90 seconds or so in the water we could usually peel the whole swatch of paper directly off the board in one piece.
Some further thoughts on preparation for etching:
Your fingernails are more than capable of lifting toner off of the copper. Don't use anything more abrasive than the pad of your thumb to remove paper fibers from the board.
Traces of toner that have run together due to excess pressure beat the heck out of traces of toner that didn't stick. You can use any sharp pointy object (we used a tool called a teasing needle) to clean up accidental bridges between bands of toner. This beat the heck out of washing the toner off the board with acetone and starting over.
If you bought dry ferric chloride, mix it with COLD water. We saw where the label on my pack of anhydrous FeCl said that it releases heat when mixed with water, but we'd also read a lot about how using warm etchant speeded the process so we had a plastic container (a former jar of peanuts) with a couple inches of hot tap water in the bottom and began spooning FeCl into it with a disposable plastic spoon, and, it pretty much boiled briefly with each spoonful. The stirring spoon actually had burn marks on it. We badly melted the bottom of the plastic peanut jar.
Mixing dry FeCl with water was for the birds, and when i run out of etchant I'll be buying it premixed from a local supplier, and it'll be worth it.
That's all I can remember right now. We should have taken pictures for the edification of others.
Having dedicated about 5 hours to learning the process, I think this can be easy enough that nobody who has access to a laser printer and a clothes iron (and a paper recycling bin) should be shy about etching their own boards. Drilling them is another issue, but hey, you can always go SMT, right?
Thoughts for enhancing the process:
National Geographic pages are even glossier than Wired. We didn't try it, though, and NatGeo is a magazine someone might want to read in a year.
After we were done etching, it occurred to me that heat transfer through the blanket of newsprint could be enhanced by slipping sheets of standard kitchen grade aluminum foil between the sheets of newsprint, to act as heat spreaders.
I've had the boards and ferric chloride for years, but never had a clear enough idea of how it was actually going to work until just this week.
That, and i dug my old nixie clock out of the closet and set it up at work, and now i want to build another. But this time i want to print my own board.
So i went with this: Build a Nixie-tube clock
Aside from how cavalier the guy is about hooking circuits up directly to mains AC (I will be using a transformer), it's a pretty neat hack. No cool PWM fading like my clock built on the nixieclocks.de board has, but, no processing power at all other than a fistfull of cmos chips. No $45 pre-programmed MCU.
And, obviously, if you look at the artwork, about the worst layout to learn how to etch your own boards with.
But today was a day for jumping in with both feet.
The sticking point (pardon the pun) that made me wary of the toner-transfer process in the past was that none of the suggestions for substrate seemed like a 100% good idea or at all frugal, usually glossy photo paper stock, usually some very specific product at $1.60 per page that has just been discontinued.
Some people say to use inkjet photo paper in your laser printer, but this stuff sometimes has a plastic gloss on it that might melt in your laser printer, ruining the fuser.
The revelation that made everything mentally fall right into place? Junk mail.
Make PCBs at home with magazine paper and your laser printer
There's been a lot of talk about using ultra-premium inkjet printer paper for toner transfer, but the reason people like it is because 8 years ago it was often coated with a clay that would absorb the ink before it could run along the fibers in the paper. The problem with that paper was that it was necessarily dusty, and running a lot of it through your laser printer could destroy your developer drum, which is why you can't find that stuff anymore - paper companies don't want to talk to angry people who ruined their laser printer by running the wrong paper through it, and there are other ways to skin that cat.
And once again, it's expensive. It's not as expensive per page as the photo paper (which is now almost all waterproof, making removal from a copper-clad board problematic), you have to buy the stuff a ream at a time. I actually bought a ream of some very expensive inkjet paper before noticing that the label explained that it no longer contains any clay so it's safe to use in a laser.
Magazine paper provides the best of both worlds. Glossy magazine paper is cheap, often free. It can have a heavy coating of clay, but it's a glossy rather than dusty coating so it probably won't stick to your developer drum.
We spent more than four hours trying to do thermal transfers of that stupid nixie clock layout today, and i think Mr. Ricci has either achieved a mastery of the technique that i have a hard time even conceiving of, or he's overly optimistic about what paper he likes.
We tried a lot of different glossy junk-mail advertisements, but what ended up working best for us, appropriately enough, was pages from old issues of Wired. The best transfer i did today was printed over an article about Google's rumored browser. Nobody was going to ever want to read this again - it was essentially garbage. The best kind of raw material to need.
The trick is, the clay gloss on the magazine paper prevents the toner from bonding with the paper. You want that gloss to be laid on thick, so that when you remove the paper you take as little of the toner off the board as possible - hopefully none of it.
The toner doesn't stick all that well to the copper. It holds on just well enough. Paper with less glaze on it pulls more toner off the board.
But more about the process:
There were more than a few pages that suggested using double-sided tape to stick the board down to your work surface with double-sided tape, then taping the paper to the work surface after you've aligned the artwork.
This proved to be the biggest load of crap we'd read on the subject of DIY thermal transfer.
It's hard to align the artwork when it's on the other side of the paper, for starters. And the heat of the iron destroys the tape, and drives cooked adhesives into your work surface. This was a DUMB idea.
Trimming the artwork to the size of the board was a mistake too, as this seemed to make it easier for the paper to shift on the face of the board during the heating process. It was best if there was up to an inch of extra paper around the edges of the artwork - and hey, the paper is free.
Applying the iron directly to the paper carrying the artwork also proved to be a bad idea, and adding one layer of heavy paper over top of the artwork to protect it didn't help much either. In either case it was easy to apply uneven pressure that would cause the traces to squish out and run together, or to cause the paper carrying the art to shift on the board.
What worked for us was to unfold a section of a newspaper (actually a pack of grocery adverts in newsprint form) on our work surface, lay the artwork face-up on the newspaper, align the (pre-cut) copper-clad board on the artwork, fold the newspaper back over (gently, without disturbing the board or artwork), and then apply the iron (set to linen) to the top of the stack (heating through, oh 10 layers of newsprint) for 30 seconds or so - just to get the paper stuck to the board - open the newspaper, flip the board and artwork over (so the copper side is now up), fold the newspaper back again, and then apply heat and moderate pressure for up to a couple minutes - again through several layers of newsprint.
It was also a good idea to check for bubbles in the transfer paper while it was still hot - they could usually be ironed flat.
Anyway, after applying enough heat and pressure (but not too much pressure!) through several layers of newsprint for long enough, we could just pick up the board piece by the excess transfer paper and drop it directly into a container of water to soak for a few minutes.
When we were experimenting with junk mail, we generally had to rub the paper off with our fingers after a long soak. It was a mess and a hassle, and a lot of toner got lifted off the copper along with it.
With sheets of Wired Magazine, after 90 seconds or so in the water we could usually peel the whole swatch of paper directly off the board in one piece.
Some further thoughts on preparation for etching:
Your fingernails are more than capable of lifting toner off of the copper. Don't use anything more abrasive than the pad of your thumb to remove paper fibers from the board.
Traces of toner that have run together due to excess pressure beat the heck out of traces of toner that didn't stick. You can use any sharp pointy object (we used a tool called a teasing needle) to clean up accidental bridges between bands of toner. This beat the heck out of washing the toner off the board with acetone and starting over.
If you bought dry ferric chloride, mix it with COLD water. We saw where the label on my pack of anhydrous FeCl said that it releases heat when mixed with water, but we'd also read a lot about how using warm etchant speeded the process so we had a plastic container (a former jar of peanuts) with a couple inches of hot tap water in the bottom and began spooning FeCl into it with a disposable plastic spoon, and, it pretty much boiled briefly with each spoonful. The stirring spoon actually had burn marks on it. We badly melted the bottom of the plastic peanut jar.
Mixing dry FeCl with water was for the birds, and when i run out of etchant I'll be buying it premixed from a local supplier, and it'll be worth it.
That's all I can remember right now. We should have taken pictures for the edification of others.
Having dedicated about 5 hours to learning the process, I think this can be easy enough that nobody who has access to a laser printer and a clothes iron (and a paper recycling bin) should be shy about etching their own boards. Drilling them is another issue, but hey, you can always go SMT, right?
Thoughts for enhancing the process:
National Geographic pages are even glossier than Wired. We didn't try it, though, and NatGeo is a magazine someone might want to read in a year.
After we were done etching, it occurred to me that heat transfer through the blanket of newsprint could be enhanced by slipping sheets of standard kitchen grade aluminum foil between the sheets of newsprint, to act as heat spreaders.
