Cold joints?

[littlejohn]

We had a flaky failure on a set of TWR lights wherein the red strobe alarm was intermittent. Eventually chased it to the timing board, which I could flex and make relays clatter. Once it was out and on the bench, you could see that all the solder joints to the header pins on the connectors looked like cold joints - gray and crystallized. When they were resoldered, they remained gray and crystallized. So, out with the solder schlurper and replaced all the solder with good ol' Ersin eutectic, nice and shiny. And, no further problems with it.
In the same plant we have a Z10CD which has been field modified to a Z16 low level HD + analog box. Suddenly, it begines to give flaky remote readings, and a check shows the problem is in the remote samples, >not< anything wrong with the radio. One part of the conversion is to remove the TB which connects the remotes to the outside world and replace it with a panel with a slightlyly larger TB, there being a couple more things on the HD remotes than on the analog only radios. Removal of that panel showed the same problem - 'crystallized' solder joints. Same drill as before, replace all the solder, and now (so far anyway) the samples are relaible.
Has anyone else run into this? I can't sell bad soldering technique to two manufacturers on two unrelated products, particularly when the stuff is wave soldered. I remember Altronics had a problem with metallurgy on some of their loads when HD RF was applied, and they had to change the plating on the resistors they use. I'm now wondering - is the HD signal responsible for screwing up solder joints? Is this a consequence of ROH compliant solder? Or do I have a poltergeist who is getting even with me after all these years?

 

[K6JHU]

I'l place my vote on lead free solder. Takes a lot more heat to get a decent joint.

However, if there are gold plated leads, it can also mess up a solder joint.

 

[Tom Wells]

About 5 years ago, I was asked by co-workers going off to some mega home suply retailer for sundries needed for the printing press installation if I needed anything. I thought a moment and told them to get me some more solder. I'd used up quite a bit, and I was not on a job near home, where I could just get more. When they came back, I just went ahead and tried some on the next set of cables I was shortening or lengthening. I hadn't noticed it was some kind of "safer" solder, and it said it was for electrical work. It gobbed, blobbed and if overworked or reheated too many times it became crystalline, and did not re-flow as easily as lead solder.
Blech! I think it was tin/antimony, and it ( insert outrageous negative connotation).
It doesn't seem to wet the surface of the substrate, and even when molten seems like it wants to crystallize rather than "freeze" in the
shiny wet look of lead/tin or lead/tin/silver.

I wouldn't be the least surprised if two modern hifalutin technologies were incompatible.

Hard to think that HD could be any issue, much as I dislike it (on AM).
It would seem to inject enough RF to heat and weaken solder joints would be rendering the equipment non-functional already.
If the equipment is outdoors and subject to sunlight, I would expect heat/cooling cycling from sun exposure would be able to
crystallize/oxidize RoHS solder within a year or two.

Will you get a chance to peek inside any other gear there?
What is the nature of the shielding in both devices mentioned?

 

[littlejohn]

I suspect the boxes are reasonably well sjielded... the TWR stuff won't work with too much RF in it, and the Z has RF bypassing on the cables headed to the outside world. I only brought that possibility up because Altronics changed a bunch of loads, and the plating metal on their load resistors as a result of field problems with HD RF. However, in that service, the load is seeing a LOT of HD RF with very little analog. I suspect I've been had by the EPA. Yet again. Sigh.

 

[oldiesstation]

Yep, you're right Little John.Wonder how long it will be before similar problems show up in other gear as well with that good ole ROH?

 

[vxo]

I do run into quite a few cold solder joints in various equipment. The killer is usually through-hole components with thick leads on a board that's been wave soldered. It seems that the PCB doesn't usually stay in contact with the solder wave long enough to give a good bond.

This problem is notoriously bad on the flyback pins of CRT monitors. The pins are thick and set into a dense glass-reinforced poly resin that's fairly thermally conductive. It's also really bad on pushbutton switches, the four-legged kind with the little snap-dome under a round button. The leads inside that get bridged by the snap-dome dissipate heat very easily during wave soldering operations. Creeeeeeak.

If you're curious as to what this process looks like, some photos of Gigabyte motherboards rolling through a modern wave soldering unit are shown here:
http://www.pcstats.com/articleview.cfm?articleid=1722&page=15

I've used lead-free solder that's worked alright. The stuff Radio Shack sells is tin/silver, and it behaves generally well, aside from having very little wetting action.

Unfortunately, I cannot say I've liked every leadfree solder I've met. I've seen some varieties grow massive whiskers (unknown brand, found on some RoHS compliant no-name pc components). I've also seen another variety turn yellow, looking like someone spilled Mountain Dew over the board and let it dry there. The yellow was tin oxide which made it completely impossible to reflow the joint, even with plenty of flux and 60/40 Sn/Pb solder! The oxide had to be mechanically scraped off to get in there and reflow the joint. This was found on the motherboard of a rather expensive Dell notebook computer!

I'd say... well... screw RoHS. It's a device that's going to have a long service life, hopefully. Suck out the leadfree on the problem joints and resolder it with some 60/40. Just be sure to dispose of the board in an ecologically sound manner at its end of life.

My grandfather, a master electrician of about 65 years, used to call bad solder joints "bird turd".

 

[LA_Guy]

I prefer lit joints to cold ones! :

 

[littlejohn]

I did heal them with good ol' MultiCore. And I'll dispose of the boards in an ecologically friendly manner just as we did with PCB filled capacitors back in The Day: the dumpster behind the Kroger.