- Not sure about exposure time (for my lamp).
- Not sure about UV light position and height required.
With these you just have to experiment, I'd suggest putting the board fairly close to the lamp (20-50mm or so).
I suggest you take one longish strip of photoresist PCB and put something on top of it that blocks UV light, then expose the strip while moving the UV blocker away at predefined intervals (say, every 10 seconds). What you will end up with is a PCB exposed in steps for different exposure times (10s, 20s, 30s, 40s, 50s etc.), then just pick the exposure time that gives off the best result.
- Not sure if my mask pattern is opaque enough (to UV light).
If it's a normal transparency sheet it should be fine. Printing to these with a laser printer works but this depends on your printer. Large areas of black might get overexposed but this depends on your printer. If your quality of black is bad you can print two sheets and align them on top of eachother.
- Not sure about correct proportions of caustic soda to water
I have been using a 3% solution of NaOH (KOH works too) which seems to work well.
- Not sure about time to leave the board in the caustic soda - whether I'm going too far developing.
You just basically move it around in the solution until it seems developed (black stuff comes off the board), usually it develops quite quickly (under 30s) but this depends on your solution. If your solution is too strong or you develop it too long it will strip off all the photoresist if you keep it in the solution too long. A solution not concentrated enough won't successfully develop the board (again, 3% solution seems to work well). Washing the board under a faucet afterwards works well. You can continue developing after taking it away from the solution. You can actually even continue developing after etching the board a bit, just remember to wash the board. Putting the board in the etching solution shows you pretty well where the board has photoresist and where it doesn't, the copper exposed to the etching solution goes to this "dull" color/texture in under a minute in the etching solution.
- I also dont know if a developed pre-sensitized board stays so, and wont fade as its left in daylight over time. How long before normal daylight affects a pre-sensitized developed board?
A board with the plastic shield SHOULDNT be affected by daylight but to be sure I'd keep them in a place shielded from light (like a drawer).
Btw. I'd suggest you try to expose your boards with a 11W fluorescent table lamp (the ones with a "U"-shaped lamp, they should be pretty common and cost like 10e from Ikea). Put the lamp quite close to the board (like 50mm close, and have a thin plate of glass over the board to keep the mask close to the board). You can get suitable plate of glass from picture frames (again, Ikea is a good place to get these), just make sure its real glass and not plastic. Expose for 12-15min (I've used 13,5min for my boards and setup). I have been using this method succesfully for a long time. Won't work for large boards due to the lamp being so narrow but for small boards it works well.
After you make sure power/ground aren't shorted, make sure that any silk screen or pin 1 markings for polarized components are correct - you don't want to go soldering that kind of stuff in backwards. After that it's really kind of random.
I was involved in bringing up a number of multi-layer boards over the years (not my designs) and we had pretty much any kind of screw-up you could make - traces that didn't go where they should, traces that simply weren't there, pads that weren't connected to their traces, etc, etc. I once even saw an issue where a broken trace was caused by the guys who built up the board gripping it wrong with pliers to break off a break-away section.
We were doing high layer-count designs, so we also had a lot of internal layer foolishness that you aren't going to see on a 2 or 4 layer board (registration can be...interesting when you've got 10 or more layers).
Once you get past the power/ground thing, you're probably not going to have too much trouble. Just take your time, test each bit of functionality one by one, and you should be good to go. If you're feeling paranoid, you could try building up the board one bit at a time (first put on and test the power, then the CPU and it's communications.
If you're feeling REALLY paranoid, you can sit down with the schematic and a meter and buzz-out the entire board. But unless the board is really small, that's gonna take a while.
You definitely want to get the main board working before you start on the secondary boards.
Good luck!
Best Answer
(1) In one application I am involved with we use a two adhesive process.
Silicone rubber is used as the long term adhesive and sealant. Tack time is hours and full cure time is 1 day +.
Many silicone rubbers set by absorption of atmospheric water and cure times may be measured in mm's of distance from surfaces in 24 hours.
"Hot melt" adhesive is used to form a rapid bond with tack times dependant on thermal transfer away from the joint site, but typically tens of seconds to a few minutes. In our case the parts concerned are held together with a velcro wrap and tensioning block - these are removed after a few minutes and recycled.
(2) Hot melt glues by themselves are not very strong and will almost invariably debond from most surfaces after a period of months to years. MOST uses of hot melt in commercial products result in joints falling apart in time.
(3) You can buy adhesives which are applied as a hot melt adhesive but which cross link and transform to a silicon rubber in hours. These are relatively rare and cost typically about twice as much as equivalent standard silicon rubbers. Dow Corning USA advertise two such silicon rubbers.
(4) The hotmelt + silicon rubber arrangement would work equally well with hotmelt + epoxy.
(5) You can buy epoxy resin with 5 minute working time and also with one minute working time (Araldite make both). These are less strong when fully cured than standard setting time epoxy resin (24 hour full cure) but MAY be strong enough for your purpose or could be used in a two adhesive system as above.
Any epoxy can have its setting time accelerated substantially by application of elevated temperatures. Be wary of "cooking" 'Araldite' by too high temperature curing.
Be aware that whole batches of "slow" adhesive can "go off" rather rapidly when warmed somewhat due to an exothermic reaction which causes self heating while setting. Ask me how I know :-).
(6) Cyanoacrylate adhesives ("superglue") set when air is excluded and can have setting times of seconds to 10's of seconds. eg Loctite make a wide range with differing characteristics. These may or may not suit depending on your materials. Some materials use adhseiomn after longer periods when used with cyanoacrylates (eg glass) and the manufacturer's recommendations should be followed.
Any of the faster setting adhesives could be used in conjunction with a slower setting one. Mechanical retention in place while an adhesive sets may also be adequate.