How to Make a Printed Circuit Board (PCB) Using the UV Light LED Method.

Intro: How to Make a Printed Circuit Board (PCB) Using the UV Light LED Method.

This instructable shows how to create well done PCB's using Ultraviolet light. It took us about 40 boards before we perfected our PCB, so we will share what worked and what didn't work. All the supplies can be bought from Fry's Electronics, Ace Hardware, Goodwill, Radio Shack, and Ebay . Our PCB tested the Atmel 208 pin PQFP Integrated Circuit (it may not look like it from the poor image quality but there were no shorts between all 208 pins! ).

The materials needed are as follows.

PCB:
-Transparencies (MG chemicals brand http://www.mgchemicals.com/products/416t.html )
-UV Presensitized Copper Clad Boards (MG chemicals brand http://www.mgchemicals.com/products/600.html )
-Laser printer (Brother HL-2070N)
-200 UV LEDs (Bought from Ebay - Asia Engineer, seller giorgio11185. 5mm size, 3.4~3.8 forward voltage, forward current 20 mA, wavelength (nm) 395-400-405, view angle about 25 degrees.)
-200 470 ohm resistors (included with Asia Engineer LEDs bought on ebay. I used 12V for each LED).
-4 Breadboards for LEDs (http://www.radioshack.com/product/index.jsp?productId=2102843 )
-PCB Standoffs (http://www.radioshack.com/product/index.jsp?productId=2102848 )
-Chest/box to house LEDs (Bought at Goodwill. 20 inch length x 12 inch width x 11 inch height)
-Picture frame transparent plastic (Bought at Goodwill)

Chemicals:
-Muriatic Acid (http://www.acehardwaresuperstore.com/transchem-muratic-acid-for-use-in-etching-concrete-p-5243.html )
-Hydrogen Peroxide (http://www.walgreens.com/store/product.jsp?CATID=302248&navAction=jump&navCount=0&nug=VPD&skuid=sku1375525&id=prod1375535 )
-Photoresist Developer (MG chemicals brand http://www.mgchemicals.com/products/418.html )
-Baking Soda (local grocery store. Use if any acid is spilled on your skin)
-Acetone (Ace Hardware)

Tools:
-Soldering Iron (any soldering iron will do, I used http://www.amazon.com/Weller-WESD51-Digital-Soldering-Station/dp/B000ARU9PO/ref=sr_1_1?ie=UTF8&s=hi&qid=1246827601&sr=1-1 )
-Solder (We originally bought Lead free solder. This type of solder did not work. Instead we used Sn63/Pb37, 2.2% Flux, 23 gauge MG Chemical brand)
-Screwdriver
-26 and 16 gauge wire
-Wire stripper
-Gloves (bought at Ace Hardware)
-Goggles (bought at Ace Hardware)
-Power Supply (wall wart or benchtop power supply will do. Make sure your power supply can handle the current the LEDs consume. As you can see from my photo, my voltage was 11.9V and 3.47 Amps were consumed).
-Q-Tips (local grocery store)
-Buckets for chemicals (Ace Hardware)
-Timer

STEP 1: Build Your UV LED Light Box.

We purchased our chest at Goodwill for $6. We suggest finding a box that is about 1 foot in height.

Elevated transparent plastic/glass is used to hold the transparency paper and presensitized copper clad about 10 inches above the LEDs. The transparent plastic was taken off a picture frame purchased at Goodwill. Keep in mind that some plastics/glass do not allow UV light to pass. Some experimenting is needed to find one that works well.

The picture below shows how Monnie arranged the LEDs on the breadboards purchased at RadioShack. We spaced the LEDs out by 6 holes from all sides.

Banana jacks were inserted in the backside of our box to be powered by a benchtop power supply.

STEP 2: Draw Your Circuit Schematic.

One circuit schematic/layout program that is free is EAGLE. The program can be downloaded from http://www.cadsoftusa.com/

Once the schematic and layout are drawn, it is time to print the layout onto a transparency.

To print only the desired circuit, turn on only the Top Layer, Pads, Vias, and Dimension. This is done by selecting View -> Display/Hide Layers in the Layout editor and selecting the correct settings.

The Layout is now ready to be printed, make sure that the image is mirrored so that text shows up correctly after being transferred onto the PCB. We printed the circuit with our Brother HL-2070N (on highest quality) laser printer and MG Chemicals transparencies.

Note: We originally bought transparencies for inkjet printers and laser printers (one side was textured for use with inkjet printers, the other side was smooth for laser printers). We had read a PCB tutorial where the guy used the textured transparency paper without problem. From our experience, the textured side did NOT allow UV light to pass through. Only use transparency paper with smooth surfaces.)

STEP 3: Expose the Presensitized Circuit Board to UV Light.

Place the printed transparency on the elevated plastic about 10 inches above the LEDs. The ink should be facing up towards the sky. While the room is somewhat dark, peel the white film off the presensitized circuit board to expose the photoresist. Place the PCB on top of the transparency, photoresist facing down towards the LEDs. Stack about 8-10 textbooks on top of the PCB to make sure the photoresist is tightly pressed against the ink of the transparency. No scotch tape is necessary to hold the PCB down to the transparency.

Turn the UV light box on. We found that exactly 1 minute exposure time worked well. You may have to experiment with different times depending on the height above the LEDs. Turn off after 1 minute.

The photoresist will look exactly the same after it has been exposed to UV light. We could not tell the difference between an exposed board and a non-exposed board.

Note: The first time you attempt to make a PCB, you will most likely fail (we failed literally about 40 times. Darn textured transparency paper!). We suggest buying a large PCB and cutting it into small squares with a Dremel (I also read a paper cutter works fine). This will save you a lot of money.

STEP 4: Prepare the Developer Solution.

Prepare the MG Chemical developer solution with 1 part developer, 10 parts tap water. I used 1/4 cup developer, 2 1/2 cups of tap water. Be sure to mix the solution with tap water thoroughly before placing the PCB in the solution, otherwise the photoresist will be eaten away in undesired spots.

Place the PCB in the mixed solution and shake it around face up. The parts where the photoresist was exposed to UV light will be washed away in about 5-10 seconds. Do not leave the board in the developer solution for too long, otherwise all the photoresist will be washed away. Place the PCB in cold water immediately to stop the reaction once the board is finished.

The developer solution can be re-used for multiple PCBs.

Note: The solution is dangerous to skin and eyes. We suggest wearing gloves and goggles.

Note: We tried all sorts of chemicals to eat away at the photoresist. We found that the MG Chemical brand developer worked the best. You could also try using Sodium Hydroxide (Lye from Ace Hardware) but you will have to experiment with what % to put in tap water. The percentage of Sodium Hydroxide to tap water is important because too strong of a solution will wash away your non-exposed UV photoresist.

STEP 5: Etch the Copper Away.

Prepare a solution of 1 part Muriatic Acid, 2 parts hydrogen peroxide. I used 8 oz. Muriatic acid, 16 oz. hydrogen peroxide. Be very careful with this acid, it's a lot stronger than the photoresist developer. I suggest using a fan over the bucket so you don't breathe in any of the fumes. Wear gloves and goggles.

Add the PCB to the solution and shake the bucket carefully. The solution will start to turn green as the copper is being etched away. As soon as all the copper is etched away, remove it from the bucket. The copper will take about 2 minutes to be etched away completely.

Wash the PCB off in a bucket or use a hose.

STEP 6: Remove the Photoresist

Remove the photoresist with Acetone and a Q-tip. The photoresist will come off very easily with a Q-tip dipped in Acetone. Keep using new Q-tips until the Q-tips are white after you have swabbed the board.

STEP 7: Populate Your PCB With Parts

At this point, you have a PCB to populate with integrated circuits and parts. Enjoy your PCB and let me know if you have any questions.

STEP 8: Mistakes

-Having the LEDs too close to the printed circuit board in the light box. The LEDs need to be sufficiently far enough for light to spread evenly throughout the circuit board. 10 inches away from the LEDs worked for us.

-Using textured transparency paper. The UV light did not pass through the transparency because of the texture. The solution is to use non-textured transparency paper. The photoresist will have little streaks of un-exposed UV. The MG Chemical brand transparencies worked well.

-Using sodium carbonate and sodium bicarbonate as a photoresist developer. Neither of these worked for us.

-Too much sodium hydroxide when used as a photoresist developer. All of the photoresist (including the non-exposed UV) was washed away in a few seconds. We suggest measuring the amount of sodium hydroxide and finding a mixture that works. If you don't want to spend the time doing that, use the MG Chemical photoresist developer.

-Adding sodium hydroxide while the PCB is already in the photoresist developer solution. This will instantly remove undesired parts of your photoresist. We suggest stirring the solution first, then dipping the PCB in the solution. If you need to add more sodium hydroxide, take the board out, add sodium hydroxide, stir the solution, then put the board in when the solution is mixed up well.

-Using "party" blacklights. We originally used three 4W party blacklights in an old scanner. The blacklights did work but not as well as the LEDs.

-Using a 500W halogen bulb. This bulb was too intense. It was difficult to find the correct exposure time.

-Using full size PCB's for testing. Don't waste your money using full size PCB's when testing, expecting the first board to work. Use a Dremel or paper cutter to cut a large PCB to save money.

Comments

[roybloor1]

I just stumbled accross this site as I was having problems with making a PCB

I wanted to keep the cost low, I don't want to buy special laminator or UV box, and lets face it by the time you've bought that lot you'll probably think 'stuff this for a lark' might aswell get it etched by a profesional PCB service, and so....

I needed to make a PCB (double sided to add to my woes) and the board I made has a complicated layout with some very fine tracks, I started out with the laser toner transfer method, and it was immediately clear from the outset that this was not going to work, so I opted for the 'photoresist' method, after several attempts and an A4 sheet of dodgy paper from an eBay seller (the film was almost fully exposed when I received it, I bought some more on a roll and properly packaged), and eventually I finally nailed it, so here's a few things I'd like to recommend to anyone using this method,

hopefully after reading this it'll only take you a few attempts to succeed, rather than the ten attempts I made before I got it right..

To start with I don't have a laminator that's suitable for adhering the film to the board, so I used a clothes iron, this does work, but temperature is key, after the first few attempts the film was wrinkling (and too cold and it won't stick), (should have realised it from the get go really) any way after getting the temperature right the film stuck to the board and was nice and smooth, the light source which ideally should be a UV lamp was not available so I lashed up a bulb holder with a CCFL light bulb a washing up bowl and a couple of rods across to hold the lamp about 8 inches away from the board, this does work but you'll need to allow half hour to an hour for the film to expose properly, you can tell when its done because the film turns a deep indigo colour.

Another mistake I made was using a solution of developer that was too strong, It was removing film resist from areas that needed to remain covered, as a rough guide I used half a teaspoon of sodium carbonate (washing soda) to 2 to 3 cups or more of water, always err on making the solution weaker rather than stronger, you can add more washing soda if you need to but go careful here or you'll ruin the whole thing and have to start again (been there, done that), then etch the board (when you've etched the board things are a bit simpler thereon) and then remove unwanted resist with sodium hydroxide

Job Done

and if your wondering about the other side of the double sided board (that has a copper surface) I paint it with acrylic (must be Acrylic) water based paint before the etch process (Acrylic paint is impervious to ferric chloride, not sure about other etchant's tho, so look out for that) then that copper doesn't get etched away,

also the author of the article was right in saying try a few small samples first, it's good advice

Conclusion:

WATCH the temperature of the iron (vitally important to ensure a smooth flat surface)

Ensure developer is at the RIGHT concentration keep it weak (to start with) (also vitally important)

hope that helps someone

Thanks

[ArifSae]

dont buy uv lights, use the natural UV i.e. sun

nope despite after all still not close to professional service. for hobbyists, its much cheaper to DIY plus we dont have to order 5 copies or expensive board.

also its not also about solution, but about CONTRAST between the lit and unlit area, the more it is, the better you can get away without ruining the whole exercise (after trying 6 times :P )

[purpulhaze]

My boards are not far from the leds and they' ve came out ok. Though I am using straw hat leds and I also have a diffused piece of plexi over the top of them. What are you using to cut your boards? I can't get a clean straight edge to save my life.

I also used the acetone a few times but have switched to using a solution of water and 3-5% lye drain cleaner. I just let my boards sit in the solution for a few seconds and they come out shinny and resist free.

[JohnathanW15]

pst...miter or tabel saw with 10inch 80 tooth $25 buck blade from harbor freight. my 5400rpm compound miter blasts through 1/4 in plexi w/o a chip... only issue i have is ryobi cm saw doesnt get or stay square very well so for my current project i took the blae to a friends house cause he had a compadible dewalt table saw. saw was like 3xxx rpms and still very little a few chips to none at all

[haze78]

Thanks for the project. Just one advice; LED powering system you're using is rather inefficient, only 29% percent efficiency, you're using 41W to power up 12W worth of LED's. You might consider some circuitry to drive LED's more efficiently, or maybe chain 2 LED's in series at least, this way you could increase efficiency to about 63%. Other more efficient solutions require more complicated and expensive circuitry (like max LED drivers, or constant current circuitry).

[JoeyJunior]

Yup, you're right. The reason I didn't put a couple LEDs in series was because the LEDs came with 470 ohm resistors and I didn't want to buy more resistors. Something like a 5V buck DC/DC converter would be pretty cool to add in the box (something like an LTC3835), maybe I'll make a PCB for one.

[apburner]

LED's run just fine on 3 volts without a resister. So you could wire then as 4 in series and then just hook them up to a 12v supply. Then there would be no resisters to waste power as heat.

[haze78]

No my friend. LEDs do NOT run fine just on 3v without resistor. You do need to limit current, and this is a MUST! Secondly, not all leds run on 3v, that is usualy correct for white LEDs, however red LEDs take something like 1.6v, so it is not a true statement for every LED. Thirdly, you cannot just connect bunch of LEDs in series and multiply nominal voltage with number of LEDs, due to voltage balancing issues. What you need is constant current source or a resistor to limit current, otherwise your LEDs are not going to glow for a long time...

[JohnathanW15]

uh in practice it would seem connecting a bunch in series and multiplying nominal FV is done all the time. esp if hooked to a cc/cv dc-dc converter. Pull the light strip off a dead lcd once (make sure its not old ccfl style )

[mitchells365]

Nope, very bad idea!!! The internal resistance of an L.E.D. is very low which is why it needs a resistor. Semiconductors behave differently than passive circuit components. without at least one resistor the L.E.D.'s will have a very short lifespan. Just because the voltage drop of a single LED can be 3 volts(Not always), doesn't mean you can just hook up 4 L.E.D.'s in series and expect everything to be alright. An L.E.D. usually can only be driven with about 20 milliamps of current, and without a carefully chosen resistor, the LED's will receive much more than their rated current; which of course can burn them out.

[rajkumar181]

i want to buy hole setup . how i can get it.

[puvanv]

1. which chemical should i have to use for photoresistive coating

2. I am using CFL(28 w) lamp to transfer my circuit through ohp sheets, please guide me the wattage required for cfl

[kunstlich]

What was the power rating of your LED's? I've just bought some on eBay but they don't seem very bright. I appreciate that they're UV though so the 'brightness' I see might not be relevant to the brightness that it's actually emitting though.

[agr00m]

Thanks for the writeup! Question about the developer solution. About how many uses did you get from 2-3/4 C mixture you made? Is it like etching solution where it loses potency after several uses?

FYI - I know your UV board is already built, but for those that have not it would probably be better to run 3 LEDs in series with a 30-90 ohm resistor (depending on actual LED voltage). It would save a lot of power loss and heat dissipation, not to mention less soldering :)

[agr00m]

One more question. Do you need to take any precautions when transporting the UV exposed PCB to the developer solution? Should the PCB be put immediately back in a light-protected sleeve, or did you just immediately move it from the UV box to the solution?

[adryanroman]

Hy, can I use just acid muriatic for etch ?!

[Azeunkn0wn]

hmmm... i don't think so.

[jedike]

Hi, I tried this but the problem is that my UV leds don't give the nice glow like I se on this picture. It is more like just a small red point in the middle of each led ?

Its like they don't want to light up completely ?

I got them on ebay like you did from the same guy actually.

Any suggestions what went wrong ?

I use also 12 Volt and same resistors as in your example.

I tried with other normal leds in this configuration and they do work, only the UV leds don't seem to work normally.

[Triakdiak]

Hi.I have bought this UV leds from the same seller on ebay and they are ok. I think that maybe seller has sent you wrong LED diodes. They also sell infrared LED diodes and human eye can not see that light , maybe you can see only small red dot in the middle of LED diode. You can make one test , point your camera on this LED , and you will se light which this diode is producing on your camera if that is the case then you have wrong diode.

[baecker03]

you can buy uv smd leds on ebay fairly cheap (guessing 10-20 bucks) just use a collimating lens to create parallel light waves. you would of course want to keep it a distance away.