BTW, have you tried scraping as much off as possible with dental picks, flat blade screwdrivers, ... before you grabbed the IPA and the toothbrush?
No, but scalpel helped.. But that was difficult.
I'd suggest getting some other tools that would be easier to use IME, and you don't need to spend a lot of money either.
Dental or electronic probe/pick/scribers in particular, but a wax carving set, nylon spudger, a soldering aid set, and toothpicks can be useful too. And not just for cleaning.
In my case, I've the following:
- Moody 55-0296 25mil 5-Piece Precision Probe Set
- Moody 55-1784 10mil Precision Probe Set, 4-Pieces (useful for getting between pins on small pitch IC's; say .65mm & smaller)
- Beau Tech SH-121 Soldering Aid set (includes a SH-20C, SH-20D, SH-20G, and a nylon spudger, SH-80)
- A couple of dental picks with slight spatula ends (got these used)
- Wax Carving set*
- Orange sticks (wood versions of a nylon spudger, and are disposable)
- Various brushes (goat, horse, and hog hair for natural fiber based versions, and ESD compliant synthetic bristles as well)
Ebay may have better prices for ^, so worth a look IMHO.
* The spatula ends on the carving set are good for mixing epoxy, the knife ends for applying it, and the pointy ends for flux removal.
Space constraints, it's a tiny PSU. Although, there is nothing around that bridge that would be sensitive to leakage or temperature. Yet, one year later uncleaned flux started to corrode wires (it's a proto-board). Fortunately, I noticed this and tried to clean with IPA and toothbrush, but not all places are accessible due to relatively dense mounting.
Sounds like you didn't fully remove the solvent/s allowing the resin to properly contain the activators.
You can run into the same issues with improperly cleaned flux as well (activators still remain, and nothing to contain/seal them from corroding traces, pins, and even bonding wire on a package that's not fully sealed).
Regarding the diodes, are they SMD or thru-hole?
I ask, as if they're thru-hole you can raise them off of the board by using a U or V shape bend in the legs between the diode body and PCB. They also help by reducing the mechanical stress to the PCB that results from thermal cycling. In fact, there are even specialty pliers that do this in a single operation. Otherwise, get a pair of needle nose or flat nose pliers and something round, say a #1 or #2 Phillips screwdriver to make them on occasion.
BTW, the U or V shape is aimed toward the diode body for horizontal mounting (or any other PTH passive). In tight spaces, you can mount them in a vertical position, with the U or V shape on one leg only, and is aimed toward the other hole. Do note that the horizontal method offers better thermal and strain relief. PnP machines don't typically do vertical mounting well (or at all on the lower cost machines IME).
I really would love to use only non-conductive non-corrosive flux, but it's not that easy to get data on this. So I conducted my own one-month experiment. I left six different fluxes on a copper clad. Please see the picture. There are two strips, both with same fluxes, but the bottom was heated with a heatgun. Not for long because fluxes became liquid and tried to mix with each other. They also started to dry out, which is not how it works when I do repairs or prototyping.
The interesting part is NC-559-AS-TF is only corrosive before heating. While FL-22 vice versa, safe until you heat it.
PS do you think colophony is a good coating for PCBs? I'd like to coat to prevent oxidation. That's what this flux suggest to do:
PPS suggestions for cheap and good fluxes are welcome. But I'm going to test them before putting into use.
Colophony is rosin.
I've not tested for conductivity not had issues thus far with the stuff I use, but I'd bet money they're all conductive when in a liquid state. Specifically I tend to use Kester 186 (RMA that borders on RA performance), Kester 1544 (RA), and MG Chemicals 835 (RA) for liquid flux.
For paste/gel flux, I use Kester RF741 (reasonably priced in the US, and easy to obtain). For solder paste (63/37), I use Kester Easy Paste 256 (usually referred to as EP256), which is simply EP256 + 63/37 solder balls.
BTW, all of these can be left on the boards for most cases (I like to clean mine though).
Regarding your experiment, I wonder if the solvent was fully boiled off while in the liquid/melted state. The reason is, once the resin is cooled (all solvent gone + remaining deposit hardens), it fully contains any activators that are part of the flux content making them inert under normal conditions (max PCB temp never reaches the activation temperature).
P.S. It's a bit long, but hopefully the detail will be of help to someone.