Sunday, January 16, 2011

I washed the upper PCB of the ART DR-X effects unit today. This process consisted of the following tools:

One used toothbrush (largely defunkified through prior use with isopropyl alcohol), one 8x10 photo developing tray, one gallon of distilled water (of which approximately one pint was used here), an air compressor and a heat gun.

I pulled the following shot from the previous write-up:

Since chemistry isn't my strong point, I can only guess that the white chemical growth at the legs of the pins is caustic potash. Therefore, I reason that a mild detergent is probably not required, since the caustic nature of the build up is going to provide more "cleaning" power than I probably need to begin with.

I poured enough distilled water to just cover the top of the ICs, and before agitation a considerable amount of the white build up dissolved. After 30-40 seconds with the brush, I hit the PCB with compressed air to disperse most of the water, then followed up with the heat gun. In order to avoid overheating the PCB, I held it by hand while distributing heat (which was a huge motivator to keep that gun moving). The idea behind the heat gun is to encourage evaporation of moisture remaining after the compressed air.

Something had been bothering me about the source of the saponification; in the past when I've seen this build up it has been localized (or at least centered around) a decomposing battery or electrolytic capacitor. This was distributed more or less evenly across the entire PCB, moreover the back-up battery and capacitors did not exhibit any of the build up.

An aspect of metallurgy is that dissimilar materials will often facilitate corrosion at contact points. In comparing before and after wash pictures, I can't help but notice that a good portion of the build up is centered on joints that were not completely filled in the soldering process.

A closer inspection of the lower, largely digital board reinforces the theory that the poor infill is, if not the root source of the failure, at least largely contributing to the problem. By the mid to late 1990s I recall that ART effects from the earlier part of the decade had gained a reputation of unreliability, which leads me to wonder if the chemical build-up around the legs, which certainly may have been allowing functionally disruptive crosstalk, wasn't isolated to this unit alone.

Perhaps I'll trip over some similar era non-functional ART units upon which to test this theory.

I'll need to grab a larger tray to wash that board though, so this process remains in a state of limbo.


the road dogg said...

I've seen this build-up on hand assembled boards with lots of surface mount components. I think it's a actually a flux residue, and I've seen it worst *after* cleaning processes. Some chemicals will lift crap but not dissolve it, so agitated the board just means that the stuff floats around until it sticks to an IC leg.

I like to get rid of it using these (you can get a few for like $1) with the bristles trimmed down to make the stiffer. Dip the brush in acetone, scrub the board clean and then rinse with pure water and dry with a nitrogen feed (or can of compresses air if that's all you have)

crochambeau said...

That would certainly explain the wide distribution across the board, any idea how long it takes to crystalize? My rinse may have just redistributed it into nooks and crannies to form up again, unless the universal solvent did its work.

the road dogg said...

Wow, that was some serious red-eyed spelling in that first comment.

It depends on the board - if I use way too much solder or too much flux the board can have a "cloudy" residue after a day or so. When I tried cleaning with diluted ethanol the residue just got washed into components and built up in little white piles.

It should be fairly easy to tell if you've removed it, the crystals are visible soon after the board is dry. Brushing seems to work much better than just allowing the board to sit in a solvent (maybe this is what ART did?)

crochambeau said...

This residue has not returned, thankfully. It wasn't like the melted sugar crust of a rosin flux, more of a whispy, dry snowflake like build up.

I think the majority of it dissolved immediately in the water. I'm opting not to reflow the joints, to see if it starts building again.

I guess I've been reading youtube comments too much, nothing in your first comment really jumped out at me as needing the mental correction that some internet spelling demands.