How to "glue" broken PCB pads

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi,
I really enjoy as hobby to restore old amplifiers. As electronic engineer, I can deal with most component as well as mechanical issues. However, I still have not found solution for broken PCB. So, hope to find some advice here.
Issue is, some of the old PCBs use such poor "glue" of the copper layer to board, that when even slightly overheat or slight mechanical damage directly remove pads for component pins go out of the PCB. See this issue for example for old Rotel amplifiers.
I’m trying to be precise as much as I can, but with some components, that has mechanically pins bond over, and the pins need to be straighten before component can be pulled out, on such type of PCB it has become nightmare.
Even be really careful, cause me usually at least one pad goes out of the board.
Many people would just scratch the PCB to bare copper next to pad, and solder the component there. But I would like to fix the PCB as much to original as possible.
I have tried to make new pads of same shape and size from thin copper foil. Which works quite well. And then I try to look for proper glue to make it stick to PCB as much as original or even better. Trying to find the glue, that can be resistant to common chemicals like alcohol, solder flux and such and can take heat up to needed soldering point (300+ degree of Celsius). Glue should be ideally transparent, so when little stays around pad, it’s not optically noticeable much.
I have tried multiple different epoxy glues, looking for specific high temperature ones. Tried different brands, some specific heat resistant or some sort of green non soldering film type. All of them failed to stick well or survived 2 second soldering.


Do you have any suggestions what glue to use? Or what is your way of fixing broken pads? Any advice appreciated.
 
Sometimes the best way to remove a component is to clip the lead close to the board and extract it through the back side. This prevents any bent over leads from causing as much of a problem as you are taking a straight piece through the hole instead.

Also, if you are overheating a board to the point where you are lifting and burning traces, I suspect your iron is not hot enough which will cause you to dwell too long. I extract components at 400-450deg Celcius and rarely lift a pad. The trick is to get in and out in a few seconds. If a component doesn't come in a few seconds, don't be tempted to dwell longer as it'll just burn. Last tip: apply fresh solder to an old joint, let it cool, rehea and suck.
 
Totally agree with avtech - use a hotter iron than for soldering for a short duration and take breaks to let the traces cool if you are having trouble. Take your time.

Also as avtech said, resolder with fresh solder as this definitely makes the solder come away better. I use a suction pump first to remove as much as possible, let it cool, and then use braid, and the 'freshened' solder wicks into the braid better than old solder.
 
1) agree with
get better at pulling old parts
and
There was never any glue there

2) or at least do not think there was some special glue added ; point is, PCB base material is always a composite material: fibrous "filler" + some adhesive to bind it into a rigid sheet.

Old pertinax is paper (cellulose fiber) impregnated with phenolic resin, which hardens by application of simultaneous heat and pressure; modern ones are made out of glass fiber + epoxy binder, again hardened under heat and pressure.
There are intermediate products, such as paper-epoxy, cloth phenolic, etc.

The point being that the copper sheet is applied to raw material, goes through the whole pressure + cooking cycle and sticks to binding material used by the insulating board itself, there is no "extra glue" applied, no need to.

So copper to substrate adhesion is as good as that of binder: phenolic resin or epoxy.

In both cases, any epoxy you use is at least as good as original adhesive, BUT surfaces must be clean; if you are trying to epoxy loose pad or track to a surface contaminated with solder flux, it will NOT stick properly.

I have been successful by previously cleaning the area with a little alcohol and a Q Tip and scratching it if necessary with a scalpel tip knife, then applying a minute quantity of industrial epoxy, the 8-12 hour type, never ever "5 minutes" household/hobby type.
Epoxy cures instantly anyway under molter solder temperature.

FWIW I use Araldit.

3) as a sobering thought, both phenolic *and epoxy* greatly soften and weaken at molten solder temperatures, so in theory neither would survive soldering and even less desoldering and removing parts ..... yet they do, the point being that in the weakened state, they stand *compression* , meaning "being pushed against board material" which happens when pulling component leg straight through the hole, they do not stand side to side wiggling and being pushed away from base material is DEADLY as when solder is improperly molten or it´s a multi pin component being wiggled around because it does not exit hole cleanly or one leg is molten but others are cooling back.

So SAFE component removal, from any kind of board, implies:

** melt solder well, as said above with a hot iron tip, remove as much solder as possible (solder wick or sucker) and pull leg cleanly out, NO side to side movement and even less pushing component back in**

* sometimes you need to cut component leads, specially on multi pin ones, so you can remove each one on its own, cleanly, as described above.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.