Problem with creating my PCB's.

Thread Starter

jdraughn

Joined Jan 30, 2009
33
A couple of years ago I tried making a couple of PCB's using the light exposure method on positive presensitized pcbs. They came out sorta ok, my main problem is that any place there was a hole for the lead of a component to go through, the very fine copper traces that surround the hole would be missing.

I just tried making another board for the first time in two years, and this time instead of exposing with UV bug zapper bulb (I made a pretty nice custom enclosure), I just used a little desk lamp with a small flourescent bulb. But the light enclosure made it hard to keep it even with the top of the glass, so the light was tilted and part of it was closer to my pcb then the other half. One end of the board was maybe 3/4" away, the other end 1.5" away.

So when I was exposing, after a few minutes I turned the lamp around so it would ultimately expose the board evenly. Well this time, half my board turned out perfect, and the other half turned out the same as from two years ago, where the fine details were gone. I didn't time it perfectly so each side was exposed an identical amount though, one side did get more exposure then the other side.

My question is, is the loss of details from over or under exposing? It would seem to me to be caused by overexposing, since the light makes it so the developer breaks down the coating. If you expose it too much, then there could be more bleed through causing the loss of fine details. And two years ago I was using much more powerful light.

The problem is I don't have much PCB left and it's pretty expensive so I didn't take the time to do any test exposures. If I had taken the time to do this 2 years ago, I would have ultimately had a lot less pcb go to waste.

I bent my transparencys so I had to print new ones (I use 2 or 3 at the same time overlapped to make sure the blacks are very black), and I am ready to try it again. I have enough pcb for two more attempts, then I am out. I am trying to make the board for a pickit2 clone. I am wondering if I should use half of what I got left to try for more test exposures, exposing small pieces to different amounts of light and timing it. If I do that it would leave me with 1 more try for the real board. I was really hoping to get two good pcb's though, I wanted one for myself and one for my brother who was wanting to learn how to get started with programming Microcontrollers.
 

jpanhalt

Joined Jan 18, 2008
11,087
A couple of years ago I tried making a couple of PCB's using the light exposure method on positive presensitized pcbs. They came out sorta ok, my main problem is that any place there was a hole for the lead of a component to go through, the very fine copper traces that surround the hole would be missing.
That problem is easily fixed. Are you using Eagle? There is a simple routine called drill-aid.ulp. That fills in the holes and leaves a very small dot to help center the drill. You can set the dot size.

I just tried making another board for the first time in two years, and this time instead of exposing with UV bug zapper bulb (I made a pretty nice custom enclosure), I just used a little desk lamp with a small flourescent bulb. But the light enclosure made it hard to keep it even with the top of the glass, so the light was tilted and part of it was closer to my pcb then the other half. One end of the board was maybe 3/4" away, the other end 1.5" away.

So when I was exposing, after a few minutes I turned the lamp around so it would ultimately expose the board evenly. Well this time, half my board turned out perfect, and the other half turned out the same as from two years ago, where the fine details were gone. I didn't time it perfectly so each side was exposed an identical amount though, one side did get more exposure then the other side.

My question is, is the loss of details from over or under exposing?
It's a combination of over exposing, exposing on a slant, and using multiple layers of mask. Using a slant and multiple layers allow the light to get under the mask.

Here is what I do: I print my transparencies very dark (if laser, I add dry-erase black and wipe off excess). I usually use ink jet, as the transparencies have fewer pin-holes with that method, because the ink flows together. I place the printed side ("ink" side) against the board, and use only one layer. I clamp between sheets of glass to keep everything close together.

The board needs only UVA (about 360 to 390 nm). Shorter wavelengths provide no advantage and create a few disadvantages.

John
 

marshallf3

Joined Jul 26, 2010
2,358
UVA is 400 nm–315 nm and doesn't go through common glass very easily but it will through quartz.

If you can make your transparencies lie flat enough a bright sunlit day can do the same thing in a very, very short time.
 

jpanhalt

Joined Jan 18, 2008
11,087
UVA is 400 nm–315 nm and doesn't go through common glass very easily but it will through quartz.
You misunderstood my statement. Perhaps, I should have said that 360 to 390 nm is in the UVA region.

As for whether regular, single strength/thickness window glass passes those wavelengths, you are mistaken in your statement that it does not.

Source = Wikipedia
Ordinary glass is partially transparent to UVA but is opaque to shorter wavelengths, whereas Silica or quartz glass, depending on quality, can be transparent even to vacuum UV wavelengths. Ordinary window glass passes about 90% of the light above 350 nm, but blocks over 90% of the light below 300 nm.
Finding an actual spectrum for window glass is a bit harder. Certainly, impurities in the glass and various surface treatments can affect transmission. However, my comment referenced ordinary glass. Here is a representative spectrum of similar (but purer) glass (Source: http://www.schott.com/advanced_opti...hott_tie-35_transmittance_october_2005_en.pdf ). The thickness is 5 mm, which is considerably thicker than single thickness window glass.



As for your recommendation to use unfiltered sunlight without a plate to hold the transparencies down, I would strongly recommend against doing it that way.

Just curious, marshallf3, have you ever used the photo process for PCB's? Have you exposed them to sunlight without a hold-down plate? If so, what was the transparency made of?

John
 

jpanhalt

Joined Jan 18, 2008
11,087
@#12: Do you use anything to hold the transparency against the PCB? If so, is it glass, fused quartz, or something else? That is the issue I raised in my response to marchallf3. Glass, of course, acts as a good filter and blocks shorter wavelengths as shown in graph attached to my post.

John
 

marshallf3

Joined Jul 26, 2010
2,358
I don't know what it's made from but when I'd photoresist a board I'd use some glass I got from when I worked for the Physics department, that or I'd just use a mist of water to hold the transparency to the board and take it outside. A fine mist of water + a squeegee would hold common film down just fine for the short exposure required.

I've also done it in my old darkroom but that involved a lot of setup and exposure time. I've also experimented with a germicidal lamp (mainly UVB) but you barely need to fire it up then back off to get a good copy.

Might make for a good box if you can find one of those tubes, I've got a simple 15W tube that works in any 18" fluorescent fixture.

No argument that most thin glass will transmit UVA but it isn't very efficient at it and the shorter the exposure time necessary the better.

[EDIT:] I just don't bother with it anymore. The toner transfer method works fairly well for single sided boards which was a great improvement over using the tape method I started out with. Nowadays it's just easier for this old man to get prototype boards made from a house that specializes in small runs or is willing to do them for students and such at a reasonable price. Most everything I design has to be two-sided with properly made feed-thrus so it really saves time and money. If you search on the internet you'll find several places that are more than willing to make you up 1 - 5 boards at a really reasonable cost provided you aren't in a hurry for them.

You will need a good PCB routing program and although I've got the full version I've tried them all and have the least complaints about DIP Trace for all but the most complex designs, mainly because their internal library is somewhat limited. You can download a copy of it for free for simple designs, I think they give you a bit better free version if you can prove a student status.

Hey - it's free, no nagging ads and I absolutely love all the schematic and other features it has. Try it out, if you don't like it you can always delete it. The autorouting feature will take some time on a slower PC and in all cases you need to verify what it came up with, I've always had to manually adjust a trace or two to my liking after it was done but it's a piece of cake once you learn it. It will produce exact size prints for transferring to toner paper as well if that's your route or generate all the files a PC board house needs to make the board.

http://www.diptrace.com/
 
Last edited:

#12

Joined Nov 30, 2010
18,224
Being unaware of the picky part, I just grabbed a piece of broken window and clipped it on with clothes pins. (Sometimes ignorance gets lucky.)
 

jpanhalt

Joined Jan 18, 2008
11,087
My point here is that some people feel since it a photochemical reaction, the "more UV, the better." So, they set up something with a germicidal lamp. In fact, glass plates, acrylic plates (Plexiglas,
http://www.rplastics.com/plexiglass-transmittance.html), and polyester (PET) transparency films all block short wavelength UV. Thus, anything that uses those materials is actually just exposing the plates to light greater than 350 to 360 nm or so. As mentioned previously, the PCB plates also contains a UV sensitizer that would negate any potential advantage from use of short wavelength UV, assuming the user even had an appropriate set-up for doing that.

Use of germicidal lamps by some hobbyists in the mistaken belief that they are better is my greatest concern. Such lamps put out virtually all of their light energy at 254 nm, which is damaging to eyes and skin, and you do not perceive how really bright the source is or sense damage until it is too late. The fluorescent tubes with phosphors, such as the BL tubes, produce more usable light (i.e., light with wavelengths >360 nm; some are pretty narrow banded at about 390 nm) for the same wattage as the germicidal lamps. They are cheaper and much safer.

John
 

marshallf3

Joined Jul 26, 2010
2,358
No doubt they're safer, I've still got that tube somewhere and it will give you a sunburn in rather short order if you expose yourself to it. No, it won't go through common glass, that's why the tube itself is made out of quartz. It will however still work on the photosynthesis boards probably because the sheer output of energy. It's been ages since I've done it.

I will say something for it though. Probably not many (if any) here are familiar with the old EPROMs and CPUs that included same but you sure could erase one in 15 minutes with that lamp.
 

thatoneguy

Joined Feb 19, 2009
6,359
I will say something for it though. Probably not many (if any) here are familiar with the old EPROMs and CPUs that included same but you sure could erase one in 15 minutes with that lamp.
I've still got one from the 80's, used it for erasing EPROM ICs on motherboards to update the BIOS to overclock a 486 to a blazing 36Mhz.
 

Thread Starter

jdraughn

Joined Jan 30, 2009
33
Thanks for all the input guys. I have been sandwiching my pcb and film between two pieces of glass, both salvaged from scanners. Maybe I would have better luck if I just used a little water so my film stuck to my pcb and skipped using the glass?

I have always made sure that the printed side is down, and I use a laser printer that I have access too while at work. I don't have any kind of printer at home.

I have also tried the laser transfer method, but I did not have much luck with that. I tried about 4 kinds of paper, different heat settings on my iron, different printers even.

I seemed to have the best luck with using my stovetop which is solid glass and preheating my pcb, then placing my paper, print side down on top of it, and then using paper towel pinched between my fingers I held the paper against the pcb with one hand while rubbing it with the tissue in my other hand. Then I would turn the stove top off and let it cool for 5 minutes before I removed the paper.

I also tried two different laminators but they didn't get hot enough.

I never actually went out and bought the paper that people said worked best (some kind of staples house brand I think), I did try glossy magazine paper, gloss photo paper, and a couple of other kinds of paper too.
 
I use 1/4" glass from our entertainment center. I attach the transparencies (doubled up) to the glass with painter's tape, then put the photosensitive pcb against the transparency and then a block of 3/4" mdf clamped onto the glass. I always have perfect pcbs using either 3 minutes of direct sunlight or 18 minutes using 2 - 20 watt 24" blacklights spaced about 6 inches away from the glass.
 

jpanhalt

Joined Jan 18, 2008
11,087
Thanks for all the input guys. I have been sandwiching my pcb and film between two pieces of glass, both salvaged from scanners. Maybe I would have better luck if I just used a little water so my film stuck to my pcb and skipped using the glass?
It is hard to say something is not worth a try, but blocking by the glass is insignificant to what the transparency does. If you look at the graph I posted, 90% transmission is almost no attenuation. Reflection losses are probably greater.

While water can act to hold the transparency in place, it also can keep it away from the photosensitized plate (note, they are hydrophobic). Bubbles can be an issue too. If you use inkjet printers, the ink is not water proof. Also, water's high refractive index relative to air may make non-perpendicular rays bend under the image more.

Glass plates are very flat, so in air, they squeeze the transparency pretty tightly against the PCB.

John
 

Thread Starter

jdraughn

Joined Jan 30, 2009
33
I forgot to mention, I have a bottom piece of glass, then a layer of paper towel, then my PCB, then my transparency, then my top layer of glass. I would use some tape on opposite ends of the glass to clamp the tissue, PCB and transparency tight so my transparency could not slide on my PCB.

I use 1/4" glass from our entertainment center. I attach the transparencies (doubled up) to the glass with painter's tape, then put the photosensitive pcb against the transparency and then a block of 3/4" mdf clamped onto the glass. I always have perfect pcbs using either 3 minutes of direct sunlight or 18 minutes using 2 - 20 watt 24" blacklights spaced about 6 inches away from the glass.
What brand of PCB did you use? From what I have read different brands require different exposure times. I think I exposed my last one a total of around 10 minutes with a that little desktop florescent lamp. 2 years ago when I was using my bug zapper light I was waiting maybe 3 minutes and still (apparently) overexposing. I would use my bug zapper light again, but I borrowed it's ballast for some other lights in my garage and it's too damn cold to go unhook it and bring it in.

Tonight while i'm at work I will make a mock pattern using eagle with different line thicknesses and cut a few 1" square pieces of PCB, get myself a box setup so I know exactly how far away the lights are from my PCB's so once I figure out my timing with my particular lights and distance, I will always be able to get repeatable results.
 
What brand of PCB did you use? From what I have read different brands require different exposure times.
I actually applied the photosensitive material myself. I use SS copper sheet and apply riston to it. You need a laminating machine to apply it properly. My wife has one she uses for crafts that takes the 1/16" pcbs just fine.

This is an older test project board, but you can see how well the lines come out with it.
 

Attachments

marshallf3

Joined Jul 26, 2010
2,358
Yea, at first I tried a laminator 4 passes through and it wouldn't transfer anything. Ended up using an iron but it took about 4 different printers and several different papers to find something that would work properly despite how "sterile" I had the blank PC board. Luckily using this method you can at least clean the board back off and try again, and again, until you come up with a combination that works correctly.

Ferric Chloride and Ammonium Persulfate were disappointing etchants for the process due to the time, I went over to a mix of 1 part Muriatic Acid + 2 parts of Hydrogen Peroxide and have been rather pleased with it despite the precautions you've got to take.
 
I went over to a mix of 1 part Muriatic Acid + 2 parts of Hydrogen Peroxide and have been rather pleased with it despite the precautions you've got to take.
I use this combination of Muriatic acid/H2O2 as well. About 8 minutes and the board is pretty darn nice. Plus, the acid and H202 are very readily available.

FWIW, I only had one board come out decent with the glossy paper / laser printer deal. I wasted so much time trying. Once I went with the riston, I always had the same excellent result.
 
Where do you get the Riston and at what price? It sounds like great stuff, but I have only found a source for like 100 feet or more.

John
Unfortunately, the guy I was buying it from quit selling it. I was buying a 6'x12" foot roll shipped for $20.

All I see now are Hong Kong dealers on eBay.

I haven't had the need lately, so I haven't been looking. I used the MM100 series Riston. I'm sure someone is selling it by the foot somewhere. I'll see what I can find.
 
Top