DIY PCBs-- worth the trouble?

[SABorn]

The answers are simple.

If using toner transfer with laser print, then the shiny waxy paper that printer labels come stuck on works a treat, and peels off clean without the need to soak it in water, just pop it in the fridge to cool after ironing and it will peel off cleanly.

I bought a cheap pack of envelope labels for a laser printer and pulled the labels off and used the surface the labels was stuck onto, (the shiny side) if you want to be conservative and cut the sticker paper up, then do a print on plain paper and use the labels to attach the cut paper over the printed area and reprint it on the cut paper.

Dont use stick tape through the laser printer.


Bryan1,

Well i just cant teach some people can i??

UV exposure.

Dont use transparencies for printing artwork onto as it will never print dense enough, its designed for light to pass through as a transparent image on the projector.

Use good old plain white paper, printer set to high quality, print in black only.

Now here is the trick...... wipe vegetable cooking oil over the back of the paper, and it will go translucent, wipe off any excess oil with a soft tissue (toilet paper works)

I have made 1000s of boards this way without fault.

2 things to remember.

After exposure check the board for any oil smuges and remove with a soft tissue before developing as it will slow the developing in the smuge area.

Dont keep the prints longer than a day after oiling as they dont work as well once left for sometime.
I always print a fresh one if its past 24 hours, hell plain paper is cheap.
This method also works well with bubble jet printers, if not better than lazer.

Pete.

 

[DirtyLude]

... also, if it is of any help to anyone here, I found that sharpie marker actually works reasonably well as a substitute for an etch resist pen, although it is not perfect.

Try to pour a ground plane. It really helps save etchant since you're not etching so much copper.

On an old web show (SYSTM) they did a test comparison of Sharpie as an etch resist and the actual Etch resist marker that's sold. I believe it was a marker by MG Chemicals, but the Sharpie won hands down.

I'm going to try the a Staedtler that KISS recommended. There are fine Sharpies, so I'm not certain if he's saying that the etch resist is better, or it's just a finer line you can draw.

I haven't had to make any corrections with a marker in a long while, but I still will marker the edge of my boards where I've cut the board a little larger than the PCB dimensions, so I got copper right to the edge.

 

[Veraxis]

...Dont use stick tape through the laser printer...

oh...oops. my printer seems fine, but I'll keep that in mind from now on.

...Try to pour a ground plane. It really helps save etchant since you're not etching so much copper...

Good point. I did use up a lot of etchant for those big boards. What is the best method for doing that? Is there an easy way to do it using Eagle?

 

[mvs sarma]

Perhaps it is cheaper to loose etchant, as toner appears costly , at least in our parts of world (India).a viamedia is to use hashing type ground plane instead of complete copper. the ground plane looks like a mesh.

 

[SABorn]

If the cost of toner is too dear for you then best you forget about electronics, as the cost is far greater.

The purpose of a ground plane is important to sensitive circuits, as well as if you dont need to remove the copper it is best not too, just etch off the mimimum, this gives a better board design and a longer life to your etchant and faster etching times.
As the etching time is much faster it helps to prevent under cutting of tracks and pads.

Your first attempt of the board you posted shows signs of over etching where it has under cut some tracks and pads.

This can be a result of the board not being cleaned well enough prior to applying the toner transfer or the transfer not being heated evenly to create a good bond to the copper or just too long in the soup.

It is only too common that people jump into a large project first up and have no prior experience of making a pcb, it is far cheaper to do a few small test boards of lesser importance to perfect the procedure prior to attempting the final board.

Pete.

 

[QuietMan]

I picked up a couple of tips here at this site, one was from Ron using wax paper (it does work). I shopped long and hard for a printer before buying one for $60 (a month later I found one for $10 at a yard sale). I've been doing it for around 35 years off and on. The real improvement for my technique was a laminator, before that it was a PITA. It made a dramatic difference overall in the quality of my boards. I went ahead and wrote a tutorial on it.

How I make PCBs

My experience is PCBs are always worth it. The work in making a PCB is easily compensated by the reduction in work in making the circuit.

 

[Veraxis]

...The purpose of a ground plane is important to sensitive circuits, as well as if you dont need to remove the copper it is best not too, just etch off the mimimum, this gives a better board design and a longer life to your etchant and faster etching times.
As the etching time is much faster it helps to prevent under cutting of tracks and pads.

Your first attempt of the board you posted shows signs of over etching where it has under cut some tracks and pads.

This can be a result of the board not being cleaned well enough prior to applying the toner transfer or the transfer not being heated evenly to create a good bond to the copper or just too long in the soup.

It is only too common that people jump into a large project first up and have no prior experience of making a pcb, it is far cheaper to do a few small test boards of lesser importance to perfect the procedure prior to attempting the final board.

Pete.

Do you find that having a ground plane prevents traces from being lifted up when peeling off the paper? I seem to recall reading something online about tracing a sacrificial ring around the layout to prevent lifting of the traces.

As for the flaws in my boards, I can attest that most of them were caused by bad transfers rather than over-etching. heating the traces evenly seems to have been my main problem. I was pretty careful to keep the boards clean, going over the whole board with steel wool before starting and wearing gloves through the entire process. the unetched spots at the top and bottom of the boards were caused by the residue from the tape I used to hold the printout to the boards during the transfer. I had cut the printouts to size, only to realize that the boards were actually somewhat larger than the specifications stated.

p.s. Just wondering, but when you say "sensitive circuits", are you referring to things like RF circuits, where you need to worry about stray inductances?

 

[3v0]

I first print the artwork on a regular sheet of paper. I then cut enough transfer paper to cover the artwork and tape it on the top edge of the design and run the paper through the printer a 2nd time. Should work with any paper.

It may have been mentioned but if your printer has a difficult time delivering enough toner use a hash fill pattern for the fills instead of solid.

Regarding Pulsar TRF (Toner Reactive Foil). Of late a member has been trying to use the white TRF which is intended for lettering in place of the green used for etching. They are not the same. The green TRF has a finer grain. The white may not work for etching as it wants to fill in between traces. It is a pain to remove, the green comes off clean.

 

[SABorn]

Just wondering, but when you say "sensitive circuits", are you referring to things like RF circuits, where you need to worry about stray inductances?

Partly yes this is correct, and i would have used it for your board as well, i recently have been using a adc to 24 bit data chip and have noticed how important it is to use good design practices or the lower 8 bits of data is so unstable they are unuseable.

I was not attacking your work but implying you could improve your design by retaining more copper.
Many multi layer boards have a ground plane sandwiched in the board so you dont see it, but it is there.

It do take some skill with a hot clothes iron to apply the toner evenly, and you need to work every square of the transfer with the edge of the iron and not trust just rubbing the iron flat over top will do.

I done a dozen practice toner runs before i could produce a perfect result, it was just a matter of wiping the toner off with acetone and start again.
Each time i leant a new problem and a solution, but better to understand and get it right than make a mess of your board.

I also found if i printed with the high ink setting on the printer it smudged when i ironed it down, so the standard ink setting was best for me.

Another hint is if you want silk screen text on the top of your board then you can reverse the text in print and iron that onto the top of the board before you fit any components.
I have done this many times and it looks neat, but best to keep it simple like names for switches or plug names etc and not a full component layout as that can look messy.
If you have a colour laser then colour text can be added.


Pete.

 

[mvs sarma]

If the cost of toner is too dear for you then best you forget about electronics, as the cost is far greater......
Pete.

Ha ha,
I was discussing hobby viability. How ever, thanks for your suggestion, Pete !!!

 

[Veraxis]

I was not attacking your work but implying you could improve your design by retaining more copper.
Many multi layer boards have a ground plane sandwiched in the board so you dont see it, but it is there.

Pete,

I apologize if I came across as being defensive: I merely meant to clarify what I thought my problem had been so as to more efficiently discuss solutions. I do greatly appreciate your advice, and this was one of my main reasons for starting this thread. I think that now that I have the materials and the beginnings of a process for making my own boards, my next goal is to collect as much advice as I can on how to improve my methods so that I can achieve a relative degree of consistency. My hope is to avoid, to some extent at least, the unnecessary trial and error in perfecting my methods that I would normally have to go through if pursuing this entirely on my own.

JLNY

 

[SABorn]

No probs,
I too can be a little acid tongued at times (often)

If you really want to produce good boards with repeatable results than ditch the transfer method and set up some simple equipment and move to photo resist method, it is not expencive and after a few trial runs you will be producing boards of a high quality with ease.

I have used most methods and would not go back to dicking around with anything other than photo resist now days.

Trust me you will thank yourself for taking the step, many times as the years pass, as it is so nice to be able to reproduce perfectly on a board, what you see on the computer screen.

There is alsorts of horror stories about using photo resist, with comments like..my printer is not good enough and it dont work for me......etc, i say ********, get your act together and fix the human error and it will work perfect, i have posted so many tips of how to use photo resist and the ones who have followed my lead have perfect results, the ones who cut corners end up with crap boards and complain that it dont work.

Pete.

PS:- have a look at the links i posted in the thread below titled ... Pcb etching equipment

 

[3v0]

@SABorn

I too have used most methods too.

Both toner transfer and photo resist are valid. People should use what they are comfortable with, or is the best fit for their needs. With either process resolution is limited by the printer. Both are 'processes' and you have to get it right.

People doing toner transfer using paper that releases 100% and a laminator can get results that rival if not equal photo resist. This has been true since Pulsar started providing TTF (Toner Transfer Paper).

 

[SABorn]

3v0

Yes that is very true, but if i was to give advice and steer someone in a direction i would still go the photo resist route as i said above.

1 print in photo resist can produce 100 boards or more where as with TT you have to dick around with each board and produce a new print each time.

Time wise i know which method wins hands down and is why i say P/R is a much more reliable method.

From printer to etchant it 5 minutes with P/R and 1/2 an hour for each TT board.

I have nothing against the TT method and it can work well, but think it sucks compared to the P/R method.

But its what you have and what works best for you.........or as we say......whatever lights your fire.

Pete.

 

[3v0]

Gosh I am glad you have no bias.
Seems most everything you know about TT is 20 years old.

Most people are only interested in doing prototypes at home. Anything more then a few boards can be better made by the board houses.

TT can be as fast as PR. With both processes it takes longer to setup the equipment then to do the work.

TT is cheaper then PR

TT does not use chemicals with a limited shelf life.

TT does allow the use of inexpensive surplus copper clad stock.

If you screw up the TT transfer you can clean it off and try again. Not so with a messed up PR exposure.

 

[HarveyH42]

Most people doing boards at home, are only looking to do one or two. If you want to do larger quantities, after you got a working board, you can send off to a board house, confident your design is good, and you aren't ordering a hundred or so, that you are stuck with, if you missed something. But on a hobby level, usually on board is all you need. I've been rushed a couple of times, and managed to go from printing to testing a completed board in under an hour, programming the AVR chip included. I usually stretch it out, and take my time, take a few breaks, not really into production type work, just want to get it right the first time.

 

[lilimike]

There's always 2 sides to everything, I normally do Toner Transfer and after reading Pete's post #34 he convinced me to try out Photo Resist but then after reading 3v0's post #35 he convinced me to continue using Toner Transfer!

Anyways I think both methods are valid and for someone who does multiple boards P/R is the way to go.

For beginners who might screw up and need to clean up and start over or for those who only make one or two boards T/T is the preferred method.

In my case I make one board with T/T, if I need a few more I normally get a cheap prototype house to make a few more and if I plan to mass produce I outsource to a fab house.

Mike

 

[HarveyH42]

Something to remember with the photo resist method, is the chemical life, and storage considerations. Time and heat, will ultimately deteriorate the quality and reliability. If you only occasionally make boards, you obviously wouldn't want a huge stock on hand. Not a huge problem, if you order the materials you need, with the components for your project. You also need to learn the process, and be aware of the timing, so you don't over-expose, or under-expose the board, and same goes for developing stage. The equipment and chemicals are also sort of dedicated to the process.

My laser printer, laminator, and paper cutter, all get used as much or more, for other things. I'm sure the Pulsar paper will breakdown eventually, but not for years. I only do maybe two boards a month. I don't see much of need to get for a higher volume solution, it's just a hobby for me.