Making your own boards at home is among the heights of achievement for home tinkerers, and one fraught with frustration. The toner transfer process requires carefully peeling away layers of photo paper, and milling your own circuit boards is an exercise in complexity. One of the best options is using photosensitive copper boards, but this requires exposing the masked-off copper to fairly intense UV light. A UV exposure box is a wonderful project, then, and something [Carlo] just about has wrapped up.
The first portion of [Carlo]’s build involved placing 135 UV LEDs on a piece of protoboard. This UV source eats up a surprising amount of power; [Carlo] is using 12V for the supply, so an old industrial power supply is more than capable of dishing out the 1.5 Amps required for the build.
Next, [Carlo] needed a timer for his exposure box. He settled on a design based on an ATMega8 turning a high voltage transistor on and off with a character LCD for the user interface. A few buttons allow [Carlo] to set the countdown timer, after which the LEDs turn on for a set period of time.
All this was packaged into a small box [Carlo] picked up from Ikea. It’s a very useful build, and judging from the video after the break, extremely easy to use.
Very nice build! I especially think that, down the road, the author will be happy that he chose a wooden box, as a plastic box would eventually get brittle from all the UV (not 100% sure about that, since I don’t know if the LEDs would be strong enough to damage the plastic).
My only cringe point was when he opened up the lid while the LEDs were on and looked directly at them!
But how did he make the circuit board for the timer?
Clicking on the first link takes me to the CHIP-8 article…
Nicely built!
For the record, if you’re using presensitized PCBs, you can expose them with a daylight spectrum CFL in a desk lamp. You don’t really need special UV light or a box. Perhaps it speeds the process or has other advantages that I’m unaware of? The sheet that comes with the PCBs says a regular incandescent bulb works fine as well.
http://quinndunki.com/blondihacks/?p=835
That’s what I do. I got a simple lamp strung up on a copper pipe frame that I made. I throw in a red light while I’m setting up, switch out to a CFL for 10 minutes, and then develop.
Of course, I’m just doing small ones. I imagine if you were doing a big board, you’d want something that throws a more uniform UV exposure.
Makes sense, I can see uniformity being a concern for a very large board. I’ve done many boards at the maximum free-Eagle size with a single CFL bulb, about 8″ away. Works great!
Transistor-n00b–question:
Isn’t this a false assumption?
“We have a load of 0.9 A to drive, so
Ic = 0.9 A (Below max constrain of 2A, so, all ok!)
Ic = Hfe * Ib, where:
Hfe = 500 (I chose to use the minimum value from the data-sheet)
Ib = Ic / Hfe = 0.9/500 = 0.0018 A, 1.8 mA
To drive our transistor into saturation we need to source 1.8 mA to the base”
Isn’t there another formula for calculating the saturation/switching mode I_be? One that would give like a 10th of H_fe (no, I’m not meaning the “use H_fe=10”-method)? Or am I wrong on this one? =)
Oops, H_fe(min). That should be OK :) Now I understand… The formula i meant was H_fe(min) > 5 * I_c/I_c(max) but that was something else.
Not I_c(max); I_e…
The thing I meant was that to use a bjt transistor as a switch, it doesn’t work like in amplification mode. Some sources suggest you should use 5x more base current than what you get from “100% on H_fe when calculating amplification mode”, apparently that appears to be H_fe(min) (?). Never seen that one before, I don’t think all datasheets have it, they only have H_fe…
Wait, wat, I think this is STILL wrong. You should use 1/5 of H_fe(min) (or just H_fe if only that is present in the data-sheet). Is that right? Confused :S
How long do people typically expose the pcbs for? I usually only expose mine for about 60 seconds, and I can get 4mil clearances no problem, not sure why you would go longer as I found it just results in more issues.
My exposure setup is similar to this, however i just run 12V to my array of leds, then clamp my pcb and transparency with the circuit between some acrylic sheets. I suspend this about 4-5″ above the leds and time it with my watch. The biggest thing I’ve found is to tightly clamp the transparency to the pcb (which i do with the acrylic).
I’ve also build UV light, except I’ve used lm317 for driving each row with 20mA current. I like the timer and this makes me finish my build, put into box. I’m thinking of using micro to drive leds, timer and perhaps pull-up smps. Having row with 12leds means almost 40v to drive them… R
I like having a dedicated exposure box. I built mine in the early 90’s. 3 foot high box painted white inside with a 300watt photoflood bulb. Exposure time around 45sec depending on what I am using for the positive. I am considering building something smaller maybe I’ll give this a try. 6-7 minutes not to bad for one of boards.
What he needs to do next is use 135 phototransistors or photoresistors to sense the footprint of the board he’s trying to expose and turn off some of the LEDs that are unneeded, reducing power consumption.
It would be cool I guess but ultimately pointless.
I mean, lets say he uses this build to do 100 boards.
At 10 minutes a board, that’s 1000 minutes, or ~17 hours of uptime.
With 12V of power linearly regulated to 1.5A, that’s 18W of power.
In all, for every 100 boards he does, he uses .306kWh of electric energy.
Let’s say that the intelligent controller saves him 50 percent of the power.
That means that for every 100 boards he makes, he is saved .153kWh of electricity.
That means he can save about 3 cents in power, per 100 boards he makes.
Simple timer, cheap as dirt. Digital kitchen timer from a dollar store, a pair of relays and a momentary switch.
Connect the coil of one relay in place of the timer’s beeper. What you build is a circuit so that pressing the button energizes the relay that switches on the exposure lamp, and the timer controlled relay interrupts power to the other relay in such a way that only pressing the button can restore power to its coil.
For a timer controlled turn on, use another timer in place of the momentary button.
I’ll leave it as an exercise to draw the schematic. Far easier to draw it than describe it.
If you think electromechanical relays are too old skool, look up solid state relays.
why do u use a timer just watch the clock turn off the switch then open the box duuu