Making PCB at home - some help needed

This is for those who make PCBs at home.

I want to make a 12inx8in PCB at home. I've bought the board from Maplin Electronics (for those from the UK) and the developer and echting chemicals. After researching a bit on the net i found that the developer is Sodium Hydroxide and not suitable.

I've also come across a few press and peel laser toner methods. I have a few queries.

1) Can I use press and peel method and then Etch it directly using Sodium Persulphate Echant
2) Most PCB making guides suggest using Ferric Chloride, my question is will using Sodium PErsulphate make any difference other than maybe the time it takes to etch the board.
3) have you got any tips regarding the whole process?

Thanks a lot for all the help esp since this is not directly related to PICAXE.

Is the PCB coated with photo-resist or just the copper-clad board?

Atharvai,

A 12" x 8' pcb is rather large to be made at home. If this is your first pcb, I would suggest making a few smaller ones for practice and to learn the technique.

There are many 'gotchas' in making the pcb, whether it is photo method or toner transfer. Don't believe all the advertisements.

For more info. checkout the Homebrew PCB gproup on Yahoo.

http://tech.groups.yahoo.com/group/Homebrew_PCBs/

Myc

atharvai said:

After researching a bit on the net i found that the developer is Sodium Hydroxide and not suitable.

Click to expand...

I bought a commercially available developer from MG Chemicals, and after using it I found that it was also Sodium Hydroxide. It can be used effectively as a developer and you can get good results, provided your techniques are adjusted to accomodate issues. I found that using a foam brush (as recommended by the manufacturer) is extremely effective at removing the exposed photoresist and leaving the rest in place.... the result was very good, clean etched traces (also using Sodium Persulphate).

Now that said - bgrabowski raised a good question - you mention press & peel which would negate the need for developer, and would require a bare copper-clad board. If you've got one with photoresist on it, it should have come in a sealed foil package, have opaque plastic on one side, and would require developer - but not the press & peel.

bgr has hit nail on head with his question.

Sodium Hydroxide and other 'developers' are to remove UV-exposed photo-resist which looks like a greeny-bluey coating over the copper on photo-resist copper clad boards and protected by a plastic self-adhesive sheet.
This has to be done before etching in the etching solutions you mention.
You would need a UV Exposure Box . And one of these for big boards is not cheap (assuming you are thinking of something good quality and not some anonymous POS from you know where). Some people have used glass sheets and sun exposure. I have never tried it and have always used UV boxes to ensure consistency.

Obviously you will need a good quality transparency print of your tracks. Any errors left un-touched will mess up your final board.

Iron-on stuff is done onto bare copper board then dropped into etcher - no developer.
I've never used it because I'm scared too. I fear that there may be an error on really thin tracks. And with the cost of board, making an error would cause me to explode. A lot of people are, however, very impressed, but I'll stick to photo-resist (not literally).


As to the 2 etching solutions you mention, a couple of comments.
Ferric Chloride is a faster etcher, but will badly stain your clothing, carpets and your Mum's bathroom sink basin. [It can be removed (ish) with the use of very nasty Oxalic Acid. Not for kids, very toxic.]
Sodium persulphate is much cleaner albeit a bit slower to etch. Bit more expensive too. Well used S/P etchant will go a nice shade of blue and can stain, though nowhere near as much as ferric chloride.
At a temp of 35-45oC or so in a baby Bubble Etcher a pcb will etch in <10 to 15 minutes for Ferric and add 5 minutes or so for Sodium persulphate. This obviously varies with how much the chemical has been used.

Etching done in a non-heated tray (e.g. a photographic type tray) will take at least twice the above times.

I've used both and can't see any difference in etch quality, though it is generally considered that the faster the etch the less chance of the etcher burrowing under tracks.

Just remember, those chemicals are unpleasant.

Get a SPARE couple of small sheets of PCB and experiment before the biggun'. If you get it right first time you'll have been lucky!

And how about the toner transfer method, which kind of paper would you guys recommend?

thanks

bgrabowski said:

Is the PCB coated with photo-resist or just the copper-clad board?

Click to expand...

OK after reading the replies my board is copper clad. it doesn't have photo resist i don't think as it did have a sealed foil package.

another point is i realised that i just touched the corner of the board and the colour changed slightly as if there was some coating. I think there is some coating on it but not sure what.

in regards to other questions, i have seen pcbs made in school but don't really recall the details although i do remem that UV light was used.

So now i have a copper clad board (non photo resist). what is the best was of transferring my schematic onto it then? my thinnest track is 1.02mm wide and the smallest pad (hole) is the standard size for an IC socket.

I would like the method to garantee (as much as possible) near zero errors. This board is for a prototype at my workplace (no one has PCB making experience only repair.)

Thanks again for the help

EDIT: sry didn't mention that i have laser printer, acetate sheets, just no CAM and conductive paint. but i'll see what u suggest and see what i can obtain. thanks again

In that case it looks like a toner transfer method, like you can do with T-shirts. Or, the iron-on press&hope method.

Marcos and others can probably give you tips on the materials etc.

I would just stress AGAIN; get a small cheap board or 2 to do a dummy run on.
You have two main possible areas for disaster... the adhesion of the transfer ink or film and getting a mirror image of the desired artwork.

If it is not bang-on , and a lot of this is in your hands, then etcher will get under.


"I would like the method to garantee (as much as possible) near zero errors." - wouldn't we all

Hang on, you'd better get yourself some CAD/CAM stuff or else you'll have to resort to pen and letraset.
There's various free stuff kicking around. Others can advise as i've never used the 'boys' stuff.

marcos.placona said:

And how about the toner transfer method, which kind of paper would you guys recommend?

Click to expand...

Marcos,

Unfortunately it is not just the type of paper used, but also the type of toner. Different brands of laser printers /copiers use different toner chemistries. You will need to find the paper/toner combintaion that works for you. Then you will have to determind by trial and error, the correct heat / pressure needed to fuse the toner to the cleaned copper.

You can go online and fine dozens of "best' methods, papers and toners. There is a paper / toner comparison chart on Yahoos Homebrew PCB group.

Many people swear by the "blue film' as a great roduct, by others swear at it because it does not work.

It is very do-able, but there is an investment in time; fine turnng the technique.

Myc

For that size board, your safest option is to have the board made professionally. The price is very reasonable.

You will need a pcb cad program to create the gerber files needed. Do Check out DIPTRACE. It is very easy to learn.

www/diptrace.com

Myc

transfer methods

Dippy - sry when i meant CAM i just meant milling equipment. I've designed the layout using PCB wizard

Ok i searched the yahoo group but didn't find a chart as such but i've joined it. I have photo paper and normal paper of all sorts here and at home so that shouldn't be a problem. The only thing is trial and error. I'm starting to think that i'll buy a PRB and do the final design there and then use my current copper clad board to test on so i don't have to buy the chemicals and borrow a UV lamp in the future.

Thanks for all your help guys!

I shall keep you posted on my progress not sure when but i'm currently finishing testing on a breadboard.

Thanks again

EDIT:
Mycroft - I use PCB wizard as i've used it since year 7. and i only ever need to design single sided PCBs so it suffices. thanks for the suggestion

I have used the blue film with great success. A couple fo things Ive learnt. Scrub the copper surface with new fine steel wool under running water with some detergent to remove that tarnishing copper gets on it. Dry really thoroughly. I usually use my heat gun to dry it out quickly. Hair dryer would work.

When you iron on the film put plain printer paper over the film to stop the iron sitcking and really press down hard. REALLY put your shoulders into it. I think this is why people have problems with it, lack of pressing with the iron to ensure the toner adheres properly to the copper surface.

Buy a Dremel tool and a stand for it if at all possible to drill out the component holes. Get the chuck accessory for it so you can use small drill bits I have .8mm bits and they work well in it. You'll get the Dremel+ stand cheapest on ebay

I too use Circuit Wizard, if you are drawing complex pcb's back up frequently, Ive had problems with it corrupting files on saving that I then couldn't open. That might be my pc rather than the software.


Mycroft, can you recommend any board makers in the US?

I like Advanced Circuits, also known as 4pcb. I've seen and used a lot of boards that they have made.

http://www.4pcb.com/index.php?load=content&page_id=129

There prices are fair and the quality is excellent. You have a lot of options, as long as you are willing to pay for them.

I'm working with them on my new Stick-AXE 8-14-20 breadboard adapter.

One thing to consider with any pcb house is the shipping costs and the time in transit. You can find cheaper pcb houses, but....

Sparkfun has a decent price on small boards, but they batch up a bunch and send them to china to be made.

Dr. Ac did a good Instructable abouut this. He used Olimex in Bulgaria(?)

Myc

prototype board making

I really can't afford getting a PCB done proffesionally. as the quotes I got online were ~£22 per board with a min order of 5. and all have a min order of 3 or 5 and the prices are around £20 each. This is just a one off board so it doesn't justify. It would be cheaper to make 2 at home for less than half the cost.

if you have any sources (shops or online) who do 1 or 2 prototypes please state the cost as well, do let me know. I know Maplin used to make PCBs a long time ago.

atharvai said:

I really can't afford getting a PCB done proffesionally. as the quotes I got online were ~£22 per board with a min order of 5. and all have a min order of 3 or 5 and the prices are around £20 each. This is just a one off board so it doesn't justify. It would be cheaper to make 2 at home for less than half the cost.

if you have any sources (shops or online) who do 1 or 2 prototypes please state the cost as well, do let me know. I know Maplin used to make PCBs a long time ago.

Click to expand...

That's a very good price for an 12" x 8" pcb!

I think Sparkfun charge $5 per sq inch

Remember a home brewed pcb will not have the solder mask nor the plated holes. You can "fake" the component legends with toner transfer as long as you like black.

Then of course, you have to hand drill all the holes. (I hate drilling holes)

Double sided boards are a challenge to align.

There are other options not discussed, You can buy a photosensitive film that you laminate before exposing. It is as if you are using a presensitozed board. There is also a version for making solder masks.

Another option is to print directly on the copper with an inkjet printer. This does work but is not for the faint of heart and there is a very large learning curve buofre success strikes.

In all cases, your succes rate will depend on the time you have spent learning and fine tuning the process

Just for the record, I am speaking form personal experience, not theory, I have tried all of the methods at one time or another.

Myc

fair enough. but all the chemicals and the board i bought cost in total £5.00.

I understand that your experience tells u that none of the homebrew methods produce professional results. but as a very early prototype i require only 1 PCB with a layout on the copper side.

ah i've been drilling holes in PCBs for years now getting quite good at it shall i say.

Thanks a lot for all the advice, i'll be trying all the methods sometime.

Just to chuck in my thoughts, having used both ferric chloride and sodium persulphate, I would go with the latter.
The fact that it's less 'messy' and the transparency of the solution and makes for easy monitoring of the process. For the most part, any increase in etching time isn't usually an issue for the hobbyist.
A 12" x 8" board is rather large, could it possibly be broken down into smaller boards? It may make production/modification/fault finding a little easier. It could also get expensive if there are any failures with whatever method is used.
I use the photo-resist method with a homemade light box requiring an exposure time of 100 seconds....good luck

Atharvai

At 12" length you are beyond the capabilities of an A4 printer to print out an acetate unless you can find one at legal paper size which will give you a length of 35cm approx.

Looks to me like you will be sticking transfers to form your resist. Good luck and i hope you have plenty of time as it is the most frustrating and hair pulling process going.

Richard

I have the impression that there's two options for this if you wanna stick to homebrew. Either you get those press n peel papers and give it a shot, or you go to photoresist (may work better). Either way it'll be really tricky.

here's a thought, depending on the proposed layout why not just make use of an etch "resist pen" and draw the circuit on the bare copper freehand style,
the only real problem would be if there is any surfacemount devices,other than that it would be just a case of careful measuring with a ruler , it is after all a 1 off board

Yeah, that's a good idea DPG. The technology is so advanced this days that we think about the "old fashioned" way less and less

Drawing would take a hell of a time, but would make sure there's no problem or errors with the tracks (considering extreme care is taken).

I myself still use resist pen sometimes to "fix" tracks that were not well transferred (yeah, I use toner transfer with normal A4 paper, and so far it works like a charm when printed in a laser printed with 1600 dpi)

For a *one-off* board, the white "Contact" or similar adhesive plastic, a felt tip pen as wide as the smallest track, and a sharp Xacto knife can make a workable PCB. Draw the sketch on the plastic, cut along the lines and expose the copper to be etched away.

You find all the crossover points as you draw the board and it's easy to wipe off the ink and redraw as needed.

You can even place components on the sketch to determine layout and routing - just remember to put ICs upside down ("dead bug" style) when verifying pins.

John

I don't see mention of component desnsity or smallest line width - perhaps I missed it...

I've hand drawn a lot of boards for 'old style' through hole type compnents... and used ferric chloride... warmed...

My technique was to draw the layout on graph paper, make 1:1 copies - clean tbe board with a chemical copper pot clearner paste - wash - dry - glue graph paper copy onto copper foil side - drill through the graph paper - peel graph paper copy - clean up rough edges around drilled holes - use the drilled holes to locate leads marking them with pencil - paint on resist tracks - dry - etch - clean off paint resist - done. [or patch any etch errors].

This is obvously only for fairly coarse work and a steady hand... good luck however you do it!

I have not tried it, but apparently expensive toner transfer paper can be substituted by pages from glossy mags. Something about the clay on the surface that makes it shine. You laser print onto the magazine paper > iron onto the copper > soak in soapy water to remove the paper.

Just a few words about Press-n-Peel

It works!
I can go from artwork to finished board, in most cases, in less than an hour.
We do prototypes so many boards need some minor rework before they are released to a board house. Sending them out for each revision is not an effective option.

The secret of PNP is really quite simple.
Make the artwork traces as wide as practical and use a a heavy font for text.

Some printers, like Brothers, are not suitable. We use a copier so we can adjust the density. Run a paper through the tray bypass, cut the PNP to fit the layout and tape it on the sheet and run it again. We have never had a problem with typical office 'scotch' tape. (do so at your risk)
Tip: Need a mirror image from the artwork and your software won't help? Use overhead transparancy sheet in the copier or printer to do the reversal!

Clean the board well. We use a 'scotch brite' scourng pad, under warm water, no soaps or solvents.

Most importantly, don't bother with momma's iron. Use a pouch laminator. We use standard bond paper instead of the pouch carier so it is not as thick and it gets better heat transfer. Run the board through without the PNP to be sure it us dry and warmed. Then run it through with the PNP tapd to the leading e. Flip it over, and maybe make several passes. Some experimentation is necessary but once it is determned the process will be stable. If the image is botched, just use solvent to remove the image, clean the board and go again

Do not remove the PNP yet! Cool the board under cold water to set the toner and the PNP carrier will almost pop off.

If there is a rare blemsh, just touch it up with a 'Sharpe' pen or trim potential shorts with a blade.

We etch with heated (aquarium heater) ammonium persulfate and get repeatable and excellent results with minimal under etch and a low mess factor.

Drill next then clean the PNP so the pad centers are more visible.

We do use a commercial board house for large quantities and high density boards but an amazing number of prototype and low volume boards are done in-house and at very low cost. We use a matrix of smaller boards then cut them apart before assembly.

Questions? Fire them off!

Len

Earlier there was a mention of an Instructable I wrote so here it is http://www.instructables.com/id/Professional-PCBs-almost-cheaper-than-making-them-/

The cost is on the website. Big boards are about $10 each, or about 5 pounds. I had some little 2x1inch boards and they were 20c each. But the setup costs are fixed and are about $50 for setup and $20 for shipping. So the boards represent excellent value if you want 10, but poor value if you want 2.

I spent my whole youth being too poor to have boards made. I have built hundreds of boards using point to point wiring with wirewrap wire. It works well and is very cheap, though you do need the proper stripper tool.

An 8x12 inch board is big. You may well be able to shrink it down if you get professional pcbs made - you can route tracks between IC pins and you can go double sided, and the board might end up 1/4 of the size.

But if you keep the tracks wide and don't try to route between IC pins, it should be quite possible to build boards using press-n-peel and all the other techniques mentioned here. Do a test bit of board - eg 1x1 inch and just see if it works.

the board is 12"x8" but the circuit on it is 10"x8". I have a lot of tracks routing between ICs. The reason the board is big is because i have inductors on it which are fairly big and are using up half the board space. they are used for several filters. I also have 1 surface mount component which measures 1.2mmx0.3mm yes thats in millimeters its an inductor. for that i'm just gonna cut the track and hope i can solder that fine without short circuit and/or melting the wire on the wirewound inductor.

i'm gonna talk to my colleagues and see what they think, but most likely i'll push towards buying a new photoresist board.

thanks for all your help and i will be trying some of the above mentioned methods for other smaller projects

I'm sure DPG will draw the artwork for you with felt pen before he goes to the pub , but seriously, that sounds like quite a job.

If you go towards photoresist have you considered all the expenses and issues?
To guarantee board-wide exposure consistency you will need an expensive UV box.

If you have very fine <0.6mm tracks then both exposure accuracy and consistent etching will be very important. I'm not sure a tray being shaken by hand every few minutes is good enough. It may be. I made many a pcb like that prior to bubble-etcher. And I had many disasters too.

You could go to the hassle of a home-made UV exposure box. My first one was home-made. Fine for small boards but the UV uneven-ness became a a real pain on bigger boards. Obv careful construction would alleviate that - have you the time?

"i'm gonna talk to my colleagues and see what they think, but most likely i'll push towards buying a new photoresist board." - is it a major corporate decision to allocate a tenner for a ph/res pcb? (Just kidding).

i need the credit card from them

Atharvai,

If you are going to use the PNP method, take the advice from BITMOVER&#8217;s post. Sounds like he&#8217;s an expert and the PNP method. But you don&#8217;t have to have a photocopier or laminator either.

I&#8217;ve been using the PNP method for sometime now making backyard DIY boards and yes&#8230;..

1) I use Mama&#8217;s Iron

2) I use a laser printer not a photocopier. (after hours laser printing at my work, don&#8217;t tell anybody!)

3) I don&#8217;t use an aquarium heater to heat the ammonium persulfate solution, but rather I boil a full kettle of water and then use a simple &#8220;double boiler&#8221; system with 2 old square plastic(!) ex-ice cream tubs. Filling up the bottom tub ¾ full of very hot water and then really letting the top tub somewhat float on the bottom tubs hot water. I only use enough hot ammonium persulfate/water solution to just cover the board in the top tub. Maybe 5 to 6 millimetres of liquid solution.

4) I don&#8217;t use an aquarium air pump to agitate the solution as some people do. I just continuously rock the top ice cream tub slowly back and forth. After a short time&#8230;. Presto! A nicely etch DIY board. Yes, I wear my gloves and goggles too!

5) I used to use .8 mm wide copper traces but I&#8217;ve recently started using 1 mm wide traces. (eye sight... old age!)

I would venture to say I have a 98 percent success rate with the PNP method. I&#8217;ve used every board I&#8217;ve etched, but rather they have not all been 100 percent perfect. But that&#8217;s good enough for me for &#8220;one off&#8221; DIY circuit boards. But then the largest board I&#8217;ve etched is 95mm by 130mm.

Cheers,

Arthavai,

By now you can see there are many different ways to make a pcb. Even using the same materials, there are different procedures that are successful for different people.

All of which require a certain amount of time to learn and fine tune for your own use. Even then you are not guaranteed 100% reliablity. You just need to start somewhere and get some practice.

Good Luck,

Myc

Add-on to previous post about board making

I do use an agitated tank, but not with air. I use nitrogen at very low pressure. It retards oxidation of the chemical and, a cylinder lasts almost forever.

Out of my expertise, so I wonder....
Since Ammonium Persulfate is much cleaner but more costly then Ferric Chloride does anyone think the copper (copper sulfate I suppose) can be removed from 'loaded' Ammonium Persulfate using electrolysis to restore the chemical? I have a small photographic silver recovery tank and the next time I have a 'loaded' batch of chemical, If the opinions warrant it, maybe Ill try it!

OK Chemistry buffs, hands up!

No more to say!

Len

Len.
Easily achievable. But crawling with caveats.

Kodak's Silver recovery.
http://www.kodak.com/eknec/documents/01/0900688a800f8101/J212ENG.pdf

But, don't use your tank for Copper.

For Copper recovery, I searched using the four words
copper etch recovery electrolysis
Hundreds of results!

For rough experiment, try a plastic/glass tank and charcoal electrodes.
And IF you can buy it, Sulphuric acid.
And a &#8220;carefully controlled&#8221; power supply.
(Neatly bringing us to a Picaxe controller.)
Electrolysis can be VERY dangerous!

(You obviously understand the electronics. Others may not.)

I can't do the sums, because I don't know the volume/molarity of your solutions,
but, would it be economically viable?
Le Chatelier didn't have to pay modern prices!

And finally,
http://www.megauk.com/ionex_water_treatment.php

e

Making PCB's

I did a tutorial on PCB making a year or so ago. These days if I have a major project I get them made overseas, but still make prototypes at home. http://andrew.meachen.net/pcb.html

The "press and peel" film method is the quickest and easiest way to do them, especially for a beginner.

If you are doing lots of boards the UV photoresist method is cheaper and faster in the long run. Mind you it took a bit of trial and error to get the exposure time right.

Either method you use, you have to etch the board with Ferric Chloride or Ammonium Persulphate. I have heard of somone using Sulphuric Acid for this although I wouldn't recomend it.

A Combination of Hydrochloric acid and hydrogen peroxide is also a very inexpensive and effective etchant. Both of these are readily availible in hardware stores.

Hydrochloric acid is also know as Muriatic acid and is used in swimming pools and to clean brickwork.

Whatever etchant you use, USE SAFETY EQUIPMENT. GOGGLES AT THE MINIUMUM.

Myc

Googles rather than gloves? I always wear both, but if I had to only wear one, it would be gloves.

I have never had etchant splash, while I have touched it (through gloves), when moving boards etc.

Andrew

Developing the board

hi, I've bought the PRB and got the layout printed on two transparancies which will be used togather.

i jsut have a small question: what are the approximate times for developing the board? and do i need to let it cool down before i put it in the developer? What should i see when the developer starts working? does the green/blue layer dissapear from the exposed board?


Thanks so much for the help!

I don't know what board you are using or the developer.

I use Fotoboard 2 and compatible plus SENO 4006.
From memory about 30 seconds at 20oC.
I would say not to use it too warm as it can overdevelop quickly and bye-bye board.

Your print should be beautifully black with no pinholes. Check and touch-in where necessary.
(Any cockups after developing can be touched-in with a photo-resist pen. Some felt-tip ink can resist etchant too by the way e.g. Sharp permanent marker.)

Remember, for best resolution, to print it so that ink side is in contact with board. So for d/sided this means a mirror print for one side.

If just aggitating in a tray you should see the exposed bluey/Green resist come off like ink. If it looks like it is partly peeling then this (usually) means you haven't exposed long enough to UV.

But other products will be different. YOU MUST R-E-A-D the product Data Sheet and ensure the developer is compatible with the photo resist board.

And you must have a couple of dummy runs first. You have been warned !!

I'm using a FR4 board from CPC Farnell. they don't seem to have a datasheet as such. they have just put develop using suitable developer. THe developer is Sodium Hydroxide.

and don't worry i will practice first! the warnings can't be ignored.