- 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.
Even simple Sodium Hydroxide will do fine, but like PerOxide etching it's taken a back seat, because it's more sensitive to correct dosing to get the right balance between effective and not too aggressive (on the product and/or your hands).
Yes you can make MetaSilicates yourself, yes you can order them. Ordering them will be much more pure.
As for the whole pentahydrate thing, as you should know if you're strong in Chemistry, is just an indication of the amount of water molecules the thing hangs off itself. If you have dry "Sodium MetaSilicate" it'll be a fine-ish powder. If you then leave it open long enough in a normal household you will eventually get clumps of PentaHydrate, because it absorbs water to go to its preferred state of with water molecules.
The only difference between them is the weight of the powder per amount of active ions, so find the right mixing guide for the stuff you order, or you'll be off by quite some. (5 water molecules are quite noticeable in weight against a simple MetaSilicate)
Best Answer
Back when I did board design more actively, typical process was purported to be:
Actually, it may be that they only drilled the registration holes (which may or may not end up on the final board) before etching and bonding, and drilled all the component holes AFTER etching and bonding.) I do distinctly remember a probelmatic layout (by a fancypants outside contractor with CAD!) where the registration holes for one layer were off by 0.050 inches so the holes were not always catching full copper on that layer.
Never visited a fab house, but I designed a few boards and checked a lot more (including good old 4x photo-reduction art - who-hoo!) as well as making my own single-layer stuff. Multi-layer we sent out.
So, missing from your list would be stacking and bonding layers, and plating the through holes. On your list but I don't think it happens is plating on the copper in the first place - as far as I know it was and is a copper sheet/foil that is bonded to the substrate. In some cases it might be built up by additional plating for a hefty upcharge, but mostly it's not. Cheaper and faster to use 2 oz foil than to plate 1 oz up to 2 oz.
Solder mask was, as far as I know, silkscreened on just like the silkscreen layer - but this is also pre-SMT days so that might well be different now that things have to be more precise. Wave soldering was state of the art and all pins were at least 2.54 mm apart.