Want to start making my own circuit board where to start

[EPTech]

Do not get me wrong. This is a friendly advice from a guy who (looking back and in his own opinion) wasted his time doing what you are about to do.

Find yourself an online PCB prototype manufacturer and invest the time you would need to get the process fine tuned into your electronics knowledge and design. I have been there and for the money you spend on descent equipment and consumables you can easily get 5 different PCB's professionally manufactured for you.

Happy prototyping.

[sasa]

Beamin,

With laser printer instead (minimum 600DPI, better 1200DPI), results can be quite decent even with plain iron. After "mastering" the process on several boards, with Toner Transfer method, it is possible to make even 0.3mm tick trace (yes, 3/10mm, it is not the error) for single side boards (probably double-sided as well, personally never tried to made any, even after decade I do homemade PCBs). I have attached some latest boards for TQFP32 and TQFP100 (trace is around 0.3mm, space only 0.2mm).

[Beamin]

Beamin,

With laser printer instead (minimum 600DPI, better 1200DPI), results can be quite decent even with plain iron. After "mastering" the process on several boards, with Toner Transfer method, it is possible to make even 0.3mm tick trace (yes, 3/10mm, it is not the error) for single side boards (probably double-sided as well, personally never tried to made any, even after decade I do homemade PCBs). I have attached some latest boards for TQFP32 and TQFP100 (trace is around 0.3mm, space only 0.2mm).

So you have a laser printer at 1200 dpi then what do you print the pattern on? Then do you take an iron to it or do you put it through a laminator?

I like HCl because there's no toxic waste. But are there some methods this won't work with?

This thread has left me with more questions then answers. I still don't want to order them due to the mismatch of long lead time to a short attention span and my constant need to tinker with RF stuff I will never fully understand. I could see myself trying out 10+ board for a simple FM transmitter each one spawned off the last trial. That could take a day of diy or a few months of mail order. 

[Beamin]

What do you guys think of this deal?
https://www.jameco.com/webapp/wcs/stores/servlet/ProductDisplay?langId=-1&productId=2113244&catalogId=10001&storeId=10001&krypto=9x3mj8umRTpahTsnAiwyN18R8EJaby4S0PCFBoQWKMBzDrMPYTRmLsRcOFW13lVZjJOu6I2ftaNp%0D%0A438add3HkOFtx141x%2F4K&ddkey=https:StoreCatalogDrillDownView

What I like is no need for laminator or laser printer. The article say inkjet works just as well:
http://makezine.com/2010/02/16/circuit-skills-circuit-board-etchin/

Is $50 a rip off or is that fair?

EDIT: That only comes with two boards. Not worth it. Probably best to buy the chemicals separately.

[Beamin]

Another idea with ink jet:

[sasa]

So you have a laser printer at 1200 dpi then what do you print the pattern on? Then do you take an iron to it or do you put it through a laminator?

I like HCl because there's no toxic waste. But are there some methods this won't work with?

Yes, 1200 dpi laser printer, even I almost always print in 600dpi, except for upper 0.3mm tick traces boards.

I use glossy paper from any quality paper magazine with 70 or 80g paper (some old computer magazines). Specially designed Press&Peel paper for toner transfer method is extremely expensive. Using laminator with temp. control is much better, however I use plain iron over cut-off old cotton shirt... No special equipment is needed to homemade quality board.

All in all, the whole process after printing will last 45 minutes or so (25 for etching, 10 for soaking and rest for pealing off and inspection before and after etching). In the same time may be etched multiple small boards, without any specific effort, except inspection.

I may write exact process step by step with exact timings in order to make down to 0.3mm traces, if that can be beneficial for you.

[KL27x]

Find yourself an online PCB prototype manufacturer and invest the time you would need to get the process fine tuned into your electronics knowledge and design. I have been there and for the money you spend on descent equipment and consumables you can easily get 5 different PCB's professionally manufactured for you.

I've had many boards made by overseas manufacturers. Thousands of dollars. This is obviously the way to go for any kind of volume order or even for one-off of a finished product of any complexity.

The reason I do toner transfer is because for every one of those boards, I have several times as many one-off simple boards for prototyping or whatnot. Oftentimes the board may be just a part of the whole circuit which I am working on and perhaps for testing of an IC, some components, and/or a microcontroller I am not familiar with. It is often helpful for me to do actual testing with the device before I am ready to commit to a complete schematic/pcb/firmware, only to find out I made false assumptions or other stupid mistake when reading the datasheet, or that the theory was easier than the practice. And it's not the cost of the PCB I care about, it's all the time and effort I spent potentially going in a bad direction. All those traces I will be ripping up and rerouting. All the code I will be deleting.

Specially designed Press&Peel paper for toner transfer is extremely expensive.

This is the part I don't get. Why anyone care about the cost of the paper? I use Pulsar, and it costs about $12.00 per 10x A4 sheets. That is 1.20 per sheet, and you can get 6x 4"x3" boards from a single sheet. That is 20 cents per 4"x3" board. I don't care if magazine paper or even photoresist is FREE. If I'm doing a large number of a board, I will be submitting my Gerber to a manufacturer, which costs even MORE. 

It's crazy adding up pennies for process which takes perhaps 30min to several hours to design the schematic/gerber/firmware, and another 30 minutes to make the PCB, and another 30 min to populate it. And after all that, it might just go in the garbage can when you find out how bad you messed it all up. And you're talking about saving 20 cents. I spent $30.00 just in shipping charges from Mouser on my most recent project. Think of it that way; if you make 30 of your toner transfer boards with free paper, and you can save ONE 7.50 shipping charge from Mouser, lol. Step aside China. Americans are ready to work for 1/3 of your minimum wage.

If you want to "save" anything significant, you have to be making a lot of boards. You might think, hey, one day I might need to make a lot of boards. And if it takes me 45 minutes to make 1 board, I can make 100 of them in 46 minutes. You will find out why the Chinese manufacturers charge you more than 20 cent a board, very quickly.

10 minutes for peeling off the paper and 10 minutes for inspection. This doesn't include the extra time for when your inspection reveals something wrong, lol. Then hello, you get to start all over. Pay 20 cents and you can eliminate this 20 minutes AND all the time for errors, because you can actually do toner transfer without any errors.  And the soak/peel takes 5 seconds.

[sasa]

This is the part I don't get. Why anyone care about the cost of the paper? I use Pulsar, and it costs about $12.00 per 10x A4 sheets. That is 1.20 per sheet, and you can get 6x 4"x3" boards from a single sheet. That is 20 cents per 4"x3" board.
...

The whole point is ability to make decent quality prototype PCBs almost immediately without any special tools and parts.

Press&Peel foil would cost me personally 5 Euro per single A4 sheet. And only if local store have it in stock, usually does not, as no one buy it. If I would have possibility to get it for reasonable price as $1.20 I would certainly buy it, however I do not. Then use it what is the most suitable for quite decent results...

Even with Press&Peel foil, garbage boards can be done as result if all other steps are not performed well. Hopefully, I have spent whole A4 copper board during initial experimentations 10 years ago, now garbage is excluded. I do not produce it much, in average once per month, at most.

[KL27x]

If you're talking about http://www.techniks.com/, I agree it's not always perfect, either. It is not even close to ease of use and margin for error you get with Pulsar. iME, it takes more heat on the low end with PnP vs even regular printer/mag paper. Whereas Pulsar takes less. The PnP plastic is an insulator to the heat, and it requires a higher minimal adhesion of toner for it to be able to rip out the plastic film, rather than the other way around. On the high end of temp, the plastic will start to distort/melt; compared to Pulsar where I can shoot 1200W heat gun at full power right on the paper, an inch away, with no worry). And unless you buy 100 sheets at a time (which I have, twice, because despite its cost and drawbacks, it is still a bit better than regular paper for the lack of paper fibers and distortion and lack of soak time, when you get it right), it is more expensive than Pulsar in the US. It's not incredibly difficult to get a good transfer with PnP. But it's not incredibly difficult to get a 99% transfer with 1% problem (in a corner or edge... it's always near the damn edge) that makes it unusable.

Pulsar is the good stuff. If you can't get it in Europe, that's too bad, because it really is that good. I have 50 sheets of PnP collecting dust. In the US you can buy it from Digikey and Mouser. While you're spending 7.50 on shipping for that last part you need, anyway.

https://www.digikey.com/product-detail/en/pulsar/50-1101/182-1003-ND/3386

They (Pulsar) also sell a foil which may help reduce porosity in ground planes. I purchased 3 packs of this film the first time I bought 3 packs of the transfer paper, and I have only opened one package just to look at it. It's a long roll of wafery thin plasticky material (not unlike dry film photoresist, I imagine), and after seeing results without it, I never even bothered to try fooling with this hassle. Results with just the paper, using my printer and process, are way better than I even hoped.

Speaking of which, one of the bonus niceties which makes Pulsar >>> PnP is simply the handling. It handles like thin cardstock. It's a dream to handle, cut, apply. PnP is slippery curly/staticy plastic and more annoying and time-consuming to handle. 

Someone in Europe needs to start importing this stuff.

[anishkgt]

Out of experience i can say Toner transfer needs some experience and if the paper is not good enough like not glossy enough it would ruin during etching process (dipping in Ferric Chloride).

I then moved to Photo resist which is much better and easy to do. You can use normal CFL also but i'd recommend UV or even better get a photo resist PCB which already comes with the photo resist material.

[KL27x]

Out of experience i can say Toner transfer needs some experience and if the paper is not good enough like not glossy enough it would ruin during etching process (dipping in Ferric Chloride).

I then moved to Photo resist which is much better and easy to do. You can use normal CFL also but i'd recommend UV or even better get a photo resist PCB which already comes with the photo resist material.

I'm beyond pedantic, at this point. Probably borderline mentally ill. I can't stop myself.

Unless you can achieve resolution finer than 8/8, there is nothing significantly important which is better or easier about photoresist for me, personally. At this level of resolution, it is my printer which fails, so perhaps with better printer we could even remove this limitation. I'm not sure if inkjet has better resolution, giving photoresist the edge, or if, perhaps, a better laser printer can do so. But trying to print 6 mil trace, I can see occasional break in the print, even on the paper. Perhaps with photoresist you can stack two very thin papers to get finer resolution around limitation of print, but...

Personally, I don't want to make a DIY board with traces any finer than 8/8. 30 AWG wrap wire is 10 mil in diameter. If you make prototype board which you can't easily modify, you may waste a lot of time with multiple iteration. Personally, I save the really tight layout and trace for final production board, and even then only where/when absolutely necessary. Most of my production boards will be 10 mil traces for same reason I do them at home. I like a board that is moddable/repairable, not disposable. And if you order your board panelized, you will see even the manufacturer has significantly lower than 100% yield with finer pitches. (Even at 10/10).

If you do not have access to the tools I use, that's a shame. But there is nothing hard or unreliable or unrepeatable about toner transfer the way I do it.

I can see potential benefit in photoresist for mass production, esp for finer trace. You can make multiple attempts at making a good template. And after physically screening them under the microscope, you can use the best one and then crank out relatively high yield. Also, with transparent/tracing papers, you can potentially get good enough registration for double sided boards with less time/effort.

[ptricks]

Toner transfer method:
If you go with this method look for an older printer, something like the hp laser jet 4. Old but prints very dark on the paper.
Iron or laminator , buy a laminator for about $20 , the iron can work but it really isn't ideal .
Paper, magazine paper worked best for me , but it had to be attached to another regular paper sheet , feeding it through the printer wrinkled magazine paper, but taping another sheet to it worked.

Photoresist:
I now use the photo method because it is really easy and low cost. I put the film on the board, run it through laminator twice and done, board is ready to expose.
I print the image onto transparency sheets using a laser printer.  inkjet printers use dye based inks and can allow the uv light to pass through.
Developer can be purchased on ebay usually from the same people selling the film.
If you want to save cost you can buy from a chinese vendor, it is usually the same stuff manufacturers use and cheap because it came off the unused rolls from production runs.
Exposure lamp I use is a regular fluorescent lamp , takes about 15 minutes.
Etchant - I use the peroxide method, but try to find somewhere that sells hair and beauty products, they have the stronger solutions that contain less water.

If you want to save time you can laminate a bunch of boards with the film and store them in a dark place until needed, the film will last months when applied, then when needed you just print your design, expose , develop , and etch. You can do a board in 30 minutes , and most of that is just waiting for the exposure time and etching.

You may read about needing a dark room with yellow light bulbs or some such thing, but I have learned that the resist film isn't that sensitive to light and I now just do it in a  room with low light but not anything dark like you would need for camera type setups.

If you use the positive photo resist you get an added benefit in that after etching the resist can be left on the board and acts as a mask for the traces that protects them from oxidation .

[KL27x]

If you have Mouser service in EU, it's essentially same price as here.
http://eu.mouser.com/ProductDetail/Pulsar/TONER-TRANFER-PAPER-50-1102/?qs=%2fha2pyFadujK5gk6R5%252bg4RvuMs6ieUwGkHlMXYmNwftaMGnYTnrRwA%3d%3d

If you don't live in EU, don't clink on that link. If you're curious of other locations you can "change location" to the right of the cart icon on the Mouser webpage.

So if this is correct, 10 pack of Pulsar is 12.74 Euros anywhere in the EU. 17.37 funny dollars in Australia. 58,11 lei in Romania. But Mouser's shipping charge is much higher in EU, unless you meet minimum order of 50 euros (then it's free), which I think many people could manage a few times a year.




A pretty hefty markup if you had to buy it from Amazon:
https://www.amazon.co.uk/TONER-TRANFER-PAPER-50-1102-Pulsar/dp/B017V3SSDO/ref=sr_1_2?ie=UTF8&qid=1502825771&sr=8-2&keywords=pulsar+paper


In US, Digikey even sells in 3500 sheet discount, lol.
https://www.digikey.com/products/en?keywords=pulsar+paper&pkeyword=pulsar+paper&v=

On the website, Pulsar claims 5 mil minimum trace width.  I would not be surprised if this is technically possible, but I suspect very few people own a laser printer capable of achieving this. At this point, you may also need to customize the driver and/or do some intermediate processing tricks. And even then, you would probably need a relatively new drum and quickly lose this resolution if you used the printer for much of anything else. Practically, 8 mil is what I call it, with any common equipment and low maintenance cost.

IMO, UV method will be relegated to specialized users - the kind who do thru hole plating and diy soldermask and who modify laser printers to try to get finer traces. It may take time, but one day the bulk of hobbyists will look back at photoresist with nostalgia.

[tooki]

I print the image onto transparency sheets using a laser printer.  inkjet printers use dye based inks and can allow the uv light to pass through.

Though I still agree that a laser is usually better, just FYI, most inkjets now use pigment black ink, even if the other colors are dye. My inkjet, for example, actually has pigment black for plain paper and dye black for photo paper, since pigment black causes metallic sheen on glossy paper. (Color pigment inkjets also exist, since the pigment inks are far more lightfast, but at the expense of somewhat reduced color accuracy.) It's eminently possible to use a pigment black inkjet for photoresist exposure.

[janoc]

I like HCl because there's no toxic waste. But are there some methods this won't work with?

Umm, what do you mean by "no toxic waste"? Aka copper compounds, very acidic and corrosive etchant, bad corrosive (and very nasty, possibly lethal if inhaled!) fumes are not toxic enough for you? Compare that with the other etchants - actually out of the common stuff the acid+peroxide combo comes out as the worst one, mainly because of the high acidity necessitating neutralization before disposal and the fumes (the copper compounds are about the same with all the etchants). I would say you  should do your research on this first before you make such claim.

The disadvantages of using HCl as etchant are well documented, even on this forum. I am not going to rehash them. Do a research on trace undercutting and you will see right away. It is a very poor quality etchant for a reason.

[Zero999]

IMO, UV method will be relegated to specialized users - the kind who do thru hole plating and diy soldermask and who modify laser printers to try to get finer traces. It may take time, but one day the bulk of hobbyists will look back at photoresist with nostalgia.

I doubt that. What's more likely to happen is the bulk of hobbyists will stop making their own PCBs altogether.

As far as I'm aware, there's only one maker of the proprietary paper you like. What happens if they go out of business? The photographic process is actually used by professionals and there are plenty of suppliers of the materials required to do it.

[tooki]

IMO, UV method will be relegated to specialized users - the kind who do thru hole plating and diy soldermask and who modify laser printers to try to get finer traces. It may take time, but one day the bulk of hobbyists will look back at photoresist with nostalgia.

I doubt that. What's more likely to happen is the bulk of hobbyists will stop making their own PCBs altogether.

As far as I'm aware, there's only one maker of the proprietary paper you like. What happens if they go out of business? The photographic process is actually used by professionals and there are plenty of suppliers of the materials required to do it.

Did I miss something? All I've seen mentioned here is toner transfer paper, which is in no way proprietary; many brands exist, including no-name stuff on eBay.

[Zero999]

IMO, UV method will be relegated to specialized users - the kind who do thru hole plating and diy soldermask and who modify laser printers to try to get finer traces. It may take time, but one day the bulk of hobbyists will look back at photoresist with nostalgia.

I doubt that. What's more likely to happen is the bulk of hobbyists will stop making their own PCBs altogether.

As far as I'm aware, there's only one maker of the proprietary paper you like. What happens if they go out of business? The photographic process is actually used by professionals and there are plenty of suppliers of the materials required to do it.

Did I miss something? All I've seen mentioned here is toner transfer paper, which is in no way proprietary; many brands exist, including no-name stuff on eBay.

The type mentioned by KL27x is Pulsar. I'm not surprised there are alternatives available (I haven't done any research because it doesn't interest me) but are they exactly the same as the branded product? I doubt it. The toner transfer process is dependant on many factors. Change one of them and you have to experiment a bit to get it working again.

[janoc]

The type mentioned by KL27x is Pulsar. I'm not surprised there are alternatives available (I haven't done any research because it doesn't interest me) but are they exactly the same as the branded product? I doubt it. The toner transfer process is dependant on many factors. Change one of them and you have to experiment a bit to get it working again.

Actually most people don't use the Pulsar paper (or any special paper) at all - it is expensive and may not be easy to get, depending on locale. Normal printer or magazine paper works fine for most. Yes, you have to experiment, the results may not be as good but that has hardly ever stopped anyone.

[KL27x]

The founder of Pulsar got the idea in 1992, after watching silkscreen graphics product which used similar paper, which was shortly after Press N Peel came onto the scene.

I don't know how I first heard about it, but I'm pretty sure it was through DIY community/maker websites, sometime around 2011ish, when I first learned about dextrin coated paper. And shortly after that, about Pulsar. If you care to search, you can find instructions for making your own dextrin-coated paper, even.

There are some "toner transfer" papers sold on eBay, which are very cheap. I bought 100 sheet pack to see what it was. They were nothing more than printer label backings, and TBH, I'm ashamed to say I threw them away without even giving them a shot. . Maybe they are great?

[sasa]

I like HCl because there's no toxic waste. But are there some methods this won't work with?

Umm, what do you mean by "no toxic waste"? Aka copper compounds, very acidic ...

I have quoted this as well, but missed to comment...

Nothing much to add to janoc comment, except that you may use 1L of Ferric Chloride almost forever, if only add small quantity domestic use HCl when bath start to weakening.  It is not the same, but "the green bath" (Cupric Chloride + HCl) is corrosive almost as fresh Ferric Chloride. I have used that when it was not easy to find fresh Ferric Chloride.

As well be double careful when etch at winter, when bath must be warm up. Ferric Chloride will work best between 25 and 40C, but never exceed that or boil it! That is the reason I put on hold all etching when outside temp is under 25 - always etch outside.

This may sound a bit overwhelmed, but with acid there is no game, especially if there is no specially quite well ventilated room for that. As I produce less then one prototype per month in average, nothing special I really need.

[Zero999]

The type mentioned by KL27x is Pulsar. I'm not surprised there are alternatives available (I haven't done any research because it doesn't interest me) but are they exactly the same as the branded product? I doubt it. The toner transfer process is dependant on many factors. Change one of them and you have to experiment a bit to get it working again.

Actually most people don't use the Pulsar paper (or any special paper) at all - it is expensive and may not be easy to get, depending on locale. Normal printer or magazine paper works fine for most.

Yes, I've done that myself before. My favourite is magazine paper. I've found that the more ink on the page the better.

Yes, you have to experiment, the results may not be as good, but that has hardly ever stopped anyone.

Hence why I prefer the photographic process. No doubt I could get better results with toner transfer if I purchased a laminator and used special paper, but I already have a UV exposure box and some pre-sensitised copper clad board and I'm more comfortable with it because it seems more professional.

Sadly I don't do much hobby electronics or making my own PCBs nowadays.

[M4trix]

Why are you guys so afraid of HCl and peroxide ? I have been using this etchant almost 30 years and I'm still alive and kicking ! 
My lab ( bathroom  ) is clean, no stains, nothing has corroded. Well, recently I switched to sodium persulfate. The main reason was, it is
odorless. The only downside is its short shelf life. 

[tautech]

The founder of Pulsar got the idea in 1992, after watching silkscreen graphics product which used similar paper, which was shortly after Press N Peel came onto the scene.

I don't know how I first heard about it, but I'm pretty sure it was through DIY community/maker websites, sometime around 2011ish, when I first learned about dextrin coated paper. And shortly after that, about Pulsar. If you care to search, you can find instructions for making your own dextrin-coated paper, even.

There are some "toner transfer" papers sold on eBay, which are very cheap. I bought 100 sheet pack to see what it was. They were nothing more than printer label backings, and TBH, I'm ashamed to say I threw them away without even giving them a shot. . Maybe they are great?

Waxy label backing paper works great.
The toner releases easily so the transfer needs to bond it well to the PCB so it stays put.
These papers work very well in my HP 6MP at 600dpi but not so well in my 1200dpi Oki.
I use ordinary printer paper as a backing with a piece of waxy paper stuck to it that's a bit larger than the PCB.

[KL27x]

Waxy label backing paper works great.
The toner releases easily so the transfer needs to bond it well to the PCB so it stays put.
These papers work very well in my HP 6MP at 600dpi but not so well in my 1200dpi Oki.

I wonder if you just peel the paper off? Or do you still soak it in water? Also, I'm curious what happens in the Oki which makes it not so good? Does it looks fine when it comes out of the printer, but the toner falls off/smudges before you can get it on the board? Or does it print fuzzy out of the gate? Perhaps certain kind of toners/printers don't stick well enough to this kind of paper.

It was very unlike me to not try it. I love it when I find cheap tools that work. When I bought it, I was kinda hoping it was going to be dextrin coated paper from china, or that maybe it was some kind of improved PnP, or that maybe it was some different technology altogether. I had a nagging fear that the waxy surface might degrade the accuracy of my printer. And I balked. I have broken 3 laser printers in the last decade trying things which seemed innocent enough at the time.

*The dumbest thing I did was try to print directly to a plastic mailer. The most recent thing was realizing I made an error about 3 seconds after hitting "print" on a 400 page document, and turning off the printer with the power switch to prevent it from printing. Yeah, that completely broke the printer. Next time I'll remove the paper.