Photoetching pcb method
Jun 19, 2005 at 7:54 PM Thread Starter Post #1 of 16

tomphillips

New Head-Fier
Joined
May 4, 2005
Posts
29
Likes
0
I want to have a go at making my own pcbs, rather than spend a large sum of money that I don't have (I'm a student you see
tongue.gif
) on getting them produced professionally.

I've seen the two main methods, toner transfer/press n peel and photoetch. I've seen that the photoetch is considered better, however, I don't have the cash for a UV lamp.

So I was thinking, since the Sun emits UV light, and some gets through to us through the atmosphere, could i simply leave a photoetch board with the transparancy laying over the top of it, held down somehow, in the sun? Anyone tried this? I'm a assuming it would have to be for several hours, if not all day, since the intensity would probably be quite low.

Thanks,

Tom.
 
Jun 19, 2005 at 8:10 PM Post #2 of 16
I dont know...sounds good but...

Diffused features
over-exposure
bleaching
etc.

can be problems. With the lan UV source you have monochromatic controlled intensities. You can accurately calculate exposure, feature depth etc.

I am speaking from a MEMS and IC Processing point of view. Dont know how important these factors are for PCB manufacture.
 
Jun 19, 2005 at 8:51 PM Post #3 of 16
Getting the exposure time right will be tricky, since you're not dealing with a constant source. Otherwise, there should be plenty of UV in the spectrum. Also, the sun is an excellent "point source": the rays are very parallel. You won't have to worry too much about light getting under the mask.

So the biggest problem will be timing and getting a homogeneous exposure. In my experience, a good quality photoresist board with a good dark mask can handle a lot of overexposure. Experiment with small samples, and see how it works.

Good Luck,
Alfred
 
Jun 19, 2005 at 9:30 PM Post #4 of 16
although they are TERRIBLE for your eyes, youc an buy uv led's.

emit pure uv... in a monochromatic (umm to use the vilible light term) spectrum range.

available both as "loose" led's and in flashlight type enclosures.

with a little ingenuity, a good setup can probably be made for cheap. maybee $40 tops. im not sure how much "light" you need...

goto:
click this link damnit...

they require that you sign and fax back a form that says that you know that if you fack with the uv light you will go blind... (well, they used to) but after that, its all good.
 
Jun 19, 2005 at 9:51 PM Post #5 of 16
From Google groups: Quote:

> 5) What UV source are people using to expose their PCB? I've tried to use
> the sun, but with inconsistent results. What period of time would be
> expected for sun exposure, with say 30% cloud?

GC quotes 10 minutes of direct sunlight for their positive
process. But slight
clouds or haze can change the light intensity on the ground by a
factor of 5-10 without you hardly even noticing.

The "negative" processes (i.e. Kepro) require real UV. The direct
positive boards I have from GC seem to do fine under fluorescent
bulbs or even tungsten floodlamps.


I'd at least try flourescent light bulbs before more expensive/exotic alternatives.
 
Jun 19, 2005 at 9:55 PM Post #6 of 16
Quote:

Originally Posted by nikongod
im not sure how much "light" you need...


Quite a bit. I use my mom's old sun lamp. These things work great. Mine dissipates about 250W or thereabouts in the Mercury vapor bulb. The light output is impressive. Can't believe my mom actually used the thing and lived to tell the tale. At half a meter, I need about two to three minutes of exposure. The PCB is behind a 4mm glass plate, so maybe replacing it with something more UV transparant would let me reduce the exposure time. I don't think these things are being sold anymore, but maybe you can find one on eBay.

Cheers,
Alfred
 
Jun 19, 2005 at 9:58 PM Post #7 of 16
you do know that "comon" glass is uv opaque?

VERY little if any will get through a sheet of normal glass.

you *need* either a quartz glass (expensive to make) bulb, or a specially made plastic/acrylic to transmit uv.

there are uv flourescent bulbs, but they are somewhat expensive.
 
Jun 19, 2005 at 10:30 PM Post #8 of 16
Quote:

Originally Posted by nikongod
you do know that "comon" glass is uv opaque?

VERY little if any will get through a sheet of normal glass.

you *need* either a quartz glass (expensive to make) bulb, or a specially made plastic/acrylic to transmit uv.

there are uv flourescent bulbs, but they are somewhat expensive.



Yes I know. That's why UV emitting lamps, fluorescent, mercury vapor, etc, are made from quartz. Normal TL lights aren't, since you don't want the UV in that case. Halogen lights are also made of quartz because of the temperatures they reach. As a consequence of both the quartz and the increased temperature, they emit much more UV than a regular bulb.

Anyway, my UV source is strong enough and my sheet of 4mm glass is translucent enough to give me good exposures in a few minutes. I could replace the glass with regular acrylic, and end up with exposure times in the order of seconds, which is too short to measure accurately.
 
Jun 20, 2005 at 1:32 AM Post #9 of 16
The only real problem with sunlight as people have said is exposure is difficult. Too many variables.

But for a budget providing you aren't making really tiny SMD traces the toner transfer can't be beat. $10 for some ammonium perasulphate which will last you about 10 etches, and about $1 for a sheet of nice paper.

The rest of the parts can be found in your home (unless you never iron shirts
rolleyes.gif
)
 
Jun 20, 2005 at 7:40 AM Post #10 of 16
hmm... I would attempt at building my own UV light using UV LEDs, as Nikongod suggested, but I'm in the UK and the company that sells them is in the US so the shipping would be rediculous for such a small component.

Not that cheap either, $24.50 per LED that emits light of at 375nM... still very near to visible light in the spectrum.

Maybe as Garbz said it would be more worthwhile to use the transfer method. Has anyone used the press-n-peel (http://www.maplin.co.uk/module.aspx?...13464&doy=20m6 ) sheets before? Do they work better than photo paper? I want to have a go at making a double sided board; is the paper or press n peel better?
 
Jun 20, 2005 at 1:02 PM Post #11 of 16
No idea on the press and peel, but the same company makes a iron transfer which works fabulously providing your iron isn't set too hot and melts the plastic. Works like a charm, but according to some sites one of your office stores in house papers work just as well if not better... I forget the name now possibly something with a paper clip.

Plus the iron method really is very simple. Takes about an hour and a half from printing to finished product if you work as slowly as me.

/EDIT: paperclip, bah. It's Staples
smily_headphones1.gif
Staples "Picture Paper" is apparently the best. I say apparently because in australia our staples store is officeworks and their paper is plastic coated and melts when ironed
rolleyes.gif
 
Jun 20, 2005 at 1:14 PM Post #12 of 16
Quote:

I want to have a go at making a double sided board; is the paper or press n peel better?


IMO, the double-sided board issue may be more significant than the other variables. With a single-sided board the precision needed is lower- you can scrape away the mask or use a laquer pen to touch up areas.

I may soon try the iron-transfer method myself, though only single-sided boards. If this thread is still active (no immediate plan yet) I'll post back results.
 
Jun 20, 2005 at 7:12 PM Post #13 of 16
Looks like I'll have a go at the iron transfer first, because its cheaper and my staples store is only a 20 min walk away. £14.99 for a pack of 5 press-n-peel sheets.... hmmm, better be good for that price if i try it out.

I might as well have a go at a double sided board, just cut the transfer to exactly the same size as the board with a knife around the copper board on the paper.

ta to all,

Tom.
 
Jun 20, 2005 at 7:28 PM Post #14 of 16
Quote:

Originally Posted by tomphillips
Looks like I'll have a go at the iron transfer first, because its cheaper and my staples store is only a 20 min walk away. £14.99 for a pack of 5 press-n-peel sheets.... hmmm, better be good for that price if i try it out.

I might as well have a go at a double sided board, just cut the transfer to exactly the same size as the board with a knife around the copper board on the paper.

ta to all,

Tom.



I was going to try Staple's Inkjet paper, not press-n-peel. The author of this article seems to have had good results with it.
 

Users who are viewing this thread

Back
Top