List of things I learned:
DO use no-clean solder OR expect to clean the boards of whatever flux comes inside your solder.
DO NOT use separate solder flux if you can avoid it, as it means another thing to clean. (Not using flux goes against typical soldering technique.)
DO clean the tip of the soldering iron frequently, especially just before use.
DO burn off solder residue by allowing the joints to heat sufficiently (except for the polystyrene caps--you will need a heat sink for these, and even then you have to be careful). Joints will look wet in shape of the bead but not be shiny when cool. Be aware that burning off residue will probably leave residue on the surrounding PCB that will need to be cleaned off if using non-no-clean solder or flux.
DO NOT bother with isopropyl alcohol for cleaning. It is dangerous to your polystyrene caps but doesn't clean well enough to be worth the danger.
Acetone turned out to be the best cleaner for me, but my solder and solder flux responded well to it. Yours may not.
DO NOT bother with lacquer thinner, even though I mentioned it. It did not evaporate cleanly and left behind residue.
Remove all excess solder, including from underneath the board, where it may be easy to overlook.
On the U87 v16 schematic, basically everything to the left of R8 needs to be super-clean, and those are the solder joints to check the most.
Testing requires listening to the mic for at least 15min between making changes. I found that it was easiest to listen to one mic while working on the other one and then switching back and forth.
Building up from just the populated boards without a capsule and without any switches or switch wires is easier than working backwards from a completed mic. It was this method that allowed me to track down all the many sources of noise. Working backwards was more tempting but ended up being a waste of time.
I probably would not buy polystyrene caps for any future builds and would use NP0/C0G caps instead just to avoid having to tiptoe around the polystyrene caps.
The 25W iron I used was actually too weak for the job. I chose it because it made it much safer to solder the polystyrene caps, but it also meant that heating joints up enough to burn off any residue took a lot longer. I probably would not have had any noise problems had I used a hotter iron (still with a fine tip). It certainly would have made it easier to deal with the few ground plane connections.
Final listening tests should go on for many hours, as some noise problems do not manifest for a very, very long time.
DO use no-clean solder OR expect to clean the boards of whatever flux comes inside your solder.
DO NOT use separate solder flux if you can avoid it, as it means another thing to clean. (Not using flux goes against typical soldering technique.)
DO clean the tip of the soldering iron frequently, especially just before use.
DO burn off solder residue by allowing the joints to heat sufficiently (except for the polystyrene caps--you will need a heat sink for these, and even then you have to be careful). Joints will look wet in shape of the bead but not be shiny when cool. Be aware that burning off residue will probably leave residue on the surrounding PCB that will need to be cleaned off if using non-no-clean solder or flux.
DO NOT bother with isopropyl alcohol for cleaning. It is dangerous to your polystyrene caps but doesn't clean well enough to be worth the danger.
Acetone turned out to be the best cleaner for me, but my solder and solder flux responded well to it. Yours may not.
DO NOT bother with lacquer thinner, even though I mentioned it. It did not evaporate cleanly and left behind residue.
Remove all excess solder, including from underneath the board, where it may be easy to overlook.
On the U87 v16 schematic, basically everything to the left of R8 needs to be super-clean, and those are the solder joints to check the most.
Testing requires listening to the mic for at least 15min between making changes. I found that it was easiest to listen to one mic while working on the other one and then switching back and forth.
Building up from just the populated boards without a capsule and without any switches or switch wires is easier than working backwards from a completed mic. It was this method that allowed me to track down all the many sources of noise. Working backwards was more tempting but ended up being a waste of time.
I probably would not buy polystyrene caps for any future builds and would use NP0/C0G caps instead just to avoid having to tiptoe around the polystyrene caps.
The 25W iron I used was actually too weak for the job. I chose it because it made it much safer to solder the polystyrene caps, but it also meant that heating joints up enough to burn off any residue took a lot longer. I probably would not have had any noise problems had I used a hotter iron (still with a fine tip). It certainly would have made it easier to deal with the few ground plane connections.
Final listening tests should go on for many hours, as some noise problems do not manifest for a very, very long time.
