I had a strange failure Monday morning. My project searchlight has been mounted on the bow for the better part of the last year (http://www.ustream.tv/channel/bowcam) and has, for the most part, worked well.
A serious squall blew through at about 7:00AM Monday with 45MPH official sustained winds, but I suspect gusts were over 60. At 7:18, according to the light's log, it went nuts. Since I was already watching the storm, I saw that the light was misbehaving so I rebooted it remotely - but it, again, wildly threw light. I shut it down.
That afternoon I brought the light in and put it on the bench. I found that two unrelated signals had 250Ohms between them. Suspecting salt corrosion, I dragged a sharp point around the pins and traces, washed the board with solvent and soaked it in dielectric contact cleaner, brushed and again washed it. None of that helped; the resistance remained unchanged. I pulled out the heat gun; under heat the resistance climbed to 300Ohms and then fell back to 250Ohms as it cooled. I heated again and, poof!, it went to infinity. I chilled it with a cooling spray until it was sweating but it remained an open circuit. The light is now back in service.
Although the light enclosure is water-resistant and has seen many rains and salt-water splash in testing, I suspect that the driving winds got inside and moistened a board, although it appeared to be dry when I opened it. I assume that there was corrosion under the processor socket - which was wetted and became conductive. Why heat freed it is mystery, but I've left it coated in mineral oil as a crude conformal coating.
An odd one.
Corrosion, I suppose
Nine months later the same symptoms appeared in my searchlight project. I'd concluded that a conductive corrosive path had formed under one of the IC sockets, which bridged two signal lines.
In March I cleared the path by alternating heat and cold. This time I just sprayed a protective product "Corrosion Block" in the suspect area and let it soak for a few days. Today I cycled the light and it works fine once again.
So, FWIW, mild conductive corrosion can apparently be reversed, not just prevented, with this mineral-oil-based compound. http://www.learchem.com/products/corrosion-block.html
In March I cleared the path by alternating heat and cold. This time I just sprayed a protective product "Corrosion Block" in the suspect area and let it soak for a few days. Today I cycled the light and it works fine once again.
So, FWIW, mild conductive corrosion can apparently be reversed, not just prevented, with this mineral-oil-based compound. http://www.learchem.com/products/corrosion-block.html
Tom
Hi Tom,
I'm not sure if one can upload photos (jpg's) here like on the AVRFreaks Forum, or not. It would be interesting to see the case and the through-case connectors you are using. Do any of your external connections show corrosion?
When I test vehicle equipment in a non-sealed case, mounted in the engine compartment, I know it will be a short-lived prototype, maybe 3 - 8 months, depending upon the seasons of the year.
JC
I'm not sure if one can upload photos (jpg's) here like on the AVRFreaks Forum, or not. It would be interesting to see the case and the through-case connectors you are using. Do any of your external connections show corrosion?
When I test vehicle equipment in a non-sealed case, mounted in the engine compartment, I know it will be a short-lived prototype, maybe 3 - 8 months, depending upon the seasons of the year.
JC
The case is non-metallic so it's fine. The electronics are in a compartment that is normally relatively sealed - but I've added a small blower that continuously evacuates the compartment and thus draws atmosphere - from about eight feet over salt water - through it. Sounds bad, but it's been doing that non-stop for several years. I've expected corrosion problems but the thing has held up pretty well.
The only obvious corrosion inside the light is benign, on the blower's aluminum frame. A DB9F on the prototype light's exterior, though, has not fared particularly well in that environment. While still functional, its body is severely rusted.
Aside from that DB9F for emergency access to the processor the only cabling is power - hard-wired on a pigtail; no connector. All signalling is via Bluetooth.
The only obvious corrosion inside the light is benign, on the blower's aluminum frame. A DB9F on the prototype light's exterior, though, has not fared particularly well in that environment. While still functional, its body is severely rusted.
Aside from that DB9F for emergency access to the processor the only cabling is power - hard-wired on a pigtail; no connector. All signalling is via Bluetooth.
Tom
Corrosion, I suppose
I have had excellent results with these products. They really do repair corroded connections and protect connections from future corrosion.
They have a lot of products, including paste forms. Check out every item on the red horizontal menu bar on their page.
http://caig.com/
On 12/13/2011 9:51 PM, General wrote:
They have a lot of products, including paste forms. Check out every item on the red horizontal menu bar on their page.
http://caig.com/
On 12/13/2011 9:51 PM, General wrote:
The case is non-metallic so it's fine. The electronics are in a compartment that is normally relatively sealed - but I've added a small blower that continuously evacuates the compartment and thus draws atmosphere - from about eight feet over salt water - through it. Sounds bad, but it's been doing that non-stop for several years. I've expected corrosion problems but the thing has held up pretty well.
The only obvious corrosion inside the light is benign, on the blower's aluminum frame. An RS232 female on the prototype light's exterior, though, has not fared particularly well in that environment. While still functional, its body is severely rusted.
Tom
http://www.ustream.tv/channel/bowcam
http://www.ustream.tv/channel/cape-coral-marine-radio VHF
http://67.207.143.181/vlf9.m3u Lightning, spherics
FYI, another method of removing light corrosion is a mineral spirits bath in a small ultrasonic cleaner.
Just for the hell of it, I submerged an operating board (initially with a signal short due to corrosion) and ran the cleaner through a three-minute cycle. The short, apparently between two adjacent 0.010" traces, opened in about 20 seconds.
To see if cavitation would ultimately do damage I ran the thing for 30 minutes. I got a super-clean circuit board (including some component markings that were erased) - but no apparent other harm. Still, I suspect that, eventually, cavitation would break something down; it's pretty aggressive.
Just for the hell of it, I submerged an operating board (initially with a signal short due to corrosion) and ran the cleaner through a three-minute cycle. The short, apparently between two adjacent 0.010" traces, opened in about 20 seconds.
To see if cavitation would ultimately do damage I ran the thing for 30 minutes. I got a super-clean circuit board (including some component markings that were erased) - but no apparent other harm. Still, I suspect that, eventually, cavitation would break something down; it's pretty aggressive.
Tom