Posted by Cees Willig on December 31, 1999 at 09:06:46:

In Reply to: 94 defender plus H2O=unhappy defender posted by shaned90 on December 30, 1999 at 22:04:21:

Hello,
Most electronic problems are locatable at connections (80% chance).
Even gold plated computer boards in auxiliary room in due time if they develop problems: it is the connections or it is exposed beyond specifications (thunderstorm inductions, over voltage, severe spikes, etc.).
Mechanism: we are dealing with DC current, not AC, hence possible surface area property change (oxide, anode, kathode effects) (At AC this effect is very much less!). The currents & voltages involved are very low and are, hence, not able to "clean" or "break through" this layer. As the Motor Management System makes its decisions based on low Voltage/ low current based signals, these signal levels are easily disturbed beyond tolerance.

What to do: I learned that loosen & connecting the connectors mechanically cleans/ scratches the connector surfaces enough to correct effects sufficiently. Also keep relays in exercise!, same story. A particular manufacturer of PABX telephone switches (I still have some trophies of the 1972 relay based machine) mounted its relays such that dust would fall through (vertically!).

Background info:
Also: it is best to use electronics not let it idle. Did you know that computer related electronics have their best mission life expactancy if environment is in the human ideal envelope: say 20 to 24 degCelcius 40 to 60% relative humidity. Large Distributed Control Systems are recommended to be operated in this envelope. Whilst they can perform under much more adverse circumstances (e.g. 5 degC to 50 degC, 10% RH to 90% RH, to give indication), but minimal replacements are required if within recommended envelope.
The mission life of an IC when component temperature is increased from 40degC to, say, 100 degC is reduced by a factor 5 to 10 at least,
hence Motor Management Systems are today taken out of engine bay (e.g. typical Mean Time Between Failure say 10 to 20 years, if under bonnet then too much guarantee claims).

Of a world leading manufacturer (USA, with one of longest experiences) of Distributed Control Systems used in petrochemical plants: about 50% of returned boards were 100% OK after cleaning (repair procedure: Control systems' self diagnostic locates faulty board/part, this is subsequently replaced, mostly on the run (can't stop a Very High Pressure steam boiler for such a swap!), some of these graceful degradation techniques appear in motor cars, eg. BMW featuring a crippled mode in its M.M.System to bring you home when faults develop. Actually these experiences are more near to my daily bread & butter: Process Control & Instrumentation, Instrumented Safety Systems, Safety Audits & Classifications, Reliability Assessments, System Design to match Availability Targets & (economical) consequences in terms of estimated manhours of labour (call out & planned)to maintain such a level, Risk Assessment for ICT, Computerised Systems in Petrochemical & Pharmaceutical Industries, etc.
Most practical aspects were learned from my maintenance technicians/ engineers at Shell Chemical Moerdijk (I was Secion Head overthere), always amazed how quickly they sniffed out the real problem (important! plants/ plantowners hate interrupts).
Happy New Year
Cees Willig
(ps. indeed I do some teaching as well, invited to write some books, etc. if answer too long: sorry).