ReX said:
Try this!
I think I would try to take them all out remembering where each one went of course. Take them in the house place them on a heating vent and dry them out completly. They must have some condensation in them to get sticky only in cold weather. Once I was satisfied they were dried out try applying a layer of silicone where there is any kind of seam to keep them dry. If there is no seam they must have been built in a high humidity enviroment like the island of japan and when cold it condenses and sticks. Worth a try if there is some kind of seam. If they were built with no seam one could drill a very small hole in the casing dry them out completly and seal while still hot and dry with silicone.
ReX
TY 4 Stroke God
Sled Dog said:
Unfortunately silicone doesn't stop moisture in the form of vapor from travelling through it. It only slows the rate. It does stop liquid water from traveling through it.
I have spent many years designing and testing sealed enclosures, many fully hermetic, many using high quality seals and replaceable moisture getters (things that "suck" the water out of the air), as well as designing heated enclosures to provide the right temperature cycles to keep the moisture levels below the required dew point and using sacrificial cold surfaces to "suck" the moisture out of electronics.
If you take a small metal package (about the size of the relay made of a small flat plate and a metal cover) and silicone the two together while monitoring the moisture inside. It will take about 2 days in a humid environment at slightly above room temperature for there to be enough moisture inside to condense when it is brought down to about -10°C. Silicone and all other plastic materials simply allow water molecules to travel through them far too much to let you seal them up and expect them to stay dry inside. Most of the time, deliberately venting the enclosure using a Gortex or Tyvek filter/vent works much better than trying to seal plastic. These work by only allowing water vapour to travel through the vent and keeping the interior moisture level just low enough to be free of condensation.
On the other hand, if you soldered the cover to the base and made a fully hermetic metal seal the inside would stay dry essentially forever. You can pass wires through a hermetic package using ceramic seals. Fully hermetic relays are available, but they cost a lot more.
In the commercial world, usually a vented enclosure is typically the way engineers end up going due to the cost.
I'm going to see if I can pick up a spare relay to carry with me - just in case.
Rex epoxy it then but it would slow or stop the condensation enough so it would not be a problem. If I had one of these sleds I would love to try it.
rlbock
Expert
fuel pump relay position
Double checked which relay controls the fuel pump on my 06 attak, it is the one position closest to the tunnel. As for the discussion of moisture affecting the relay, the unit appears to be well sealed and I would think if it was moisture related then it would happen more often at temperatures below freezing, my experience was it had to be 10F or colder for a failure to occur.
Double checked which relay controls the fuel pump on my 06 attak, it is the one position closest to the tunnel. As for the discussion of moisture affecting the relay, the unit appears to be well sealed and I would think if it was moisture related then it would happen more often at temperatures below freezing, my experience was it had to be 10F or colder for a failure to occur.
Ok was this relay made in a very moist place? Like the island of Japan where the humidity is very high. These relays may have trapped moisture in them from the day they were made and sealed in Japan trapping the moisture. Now go to a place where temp drops to -20C and the relay sticks. Warm it up and it works. I still think if someone drills a small hole in the casing and dries it out completly and reseals the unit will not give you any more trouble. Worth a try anyhow.
ReX
TY 4 Stroke God
Sled Dog said:
Unfortunately epoxy is just as bad at letting water vapor/molecules pass through it.
Basically everything that isn't metal, glass or ceramic lets water vapor through it. Its just a matter of time and unfortunately the time it takes for things to get wet on the other side of a plastic/epoxy/silicone/etc. seal is typically anywhere from several hours to a few weeks. Also if the item is placed in a dry, heated environment the dry out time is similar.
Do a web search on hermeticity or hermetic seals and you'll find tons of information on keeping moisture out of things like relays. There is a tremendous amount of effort put into hermeticity by the electronics industries and there are a huge number of people who devote their carriers to hermetic packaging.
Basically if you want any relay to operate reliably long term in the outside world where temperatures drop below freezing and moisture content of the air varies you need either a vented enclosure or a hermetically sealed enclosure. The vented route will still have some situations where water can condense and freeze though, but the better the venting, the less likely.
Here are a couple of links I found:
http://mechanical-components.globalspec.com/LearnMore/Mechanical_Components/Seals/Hermetic_Seals
http://en.wikipedia.org/wiki/Hermetic_seal
Its possible the relay Yamaha is using is intended to be hermetically sealed. If so, the plastic has likely been metalized on the inside and during the final packaging the metal has been fused or solder together somehow. Also ceramic or glass feedthroughs would be used around the electrical pins.
The 2nd approach they could be using is to use a commercial relay that is made from non-corroding materials and the package is vented. They would then be relying on the relay drying out enough to not freeze in one position or another (self heating, helping to dry it out a little or melt any frost build up internally and the frost evaporating and escaping when the sled is parked). If this is what is being done, sealing it up would almost certainly make it worse because once it gets water inside it, it would take a longer time to dry out (every sudden drop in temperature would make the relay stick for days or until it warmed up again).
The 3rd approach is to have it vented but to also have a heating element inside that quickly heats it up to melt/evaporate any frost when it is turned on. The better the venting the better this will work also.
If you want to try an experiment I would suggest taking a bad relay and carefully cutting one end of it off (the very top). Then glue a piece of Tyvek or Gortex over the end to still let it "breath" while keeping contaminants out and stopping liquid water going directly into the package and then re-install the relay. This should allow enough venting that every morning there's a very good chance the moisture should have been able to escape. There will still be some situations where it can get frost inside, but it won't be as likely as with a typical "sealed" enclosure where moisture slowly goes through the plastic and non-hermetic seals and gets trapped (taking too much time to escape) inside.
Rex you keep saying no without trying it, this would seem to me to be a very easy thing to try out and I would had I had the machine both of mine are RX-1's. Sorry but I cant see that much vapour getting in there once it is dried out and sealed correctly. A very minute amount will not cause it to stick. Why is it only the fuel pump relay sticking or it it the only which will cause the sled not to start?
ReX
TY 4 Stroke God
Sled Dog said:
I don't know what I can say to convince you but I've worked on many mechanical designs that were required to survive natural environmental conditions over the past 20 years and I have tried epoxy to seal mechanical electrical switches (such as a relay). 10 years ago in particular I was working for General Dynamics and one of my projects was trying to get a bunch of commercial switches to work as part of an upgrade to the AWACS aircraft. This was around the time when there was a big push towards using "COTS" (Commercial Off The Shelf) parts in military equipment instead of having everything custom.
My team and I spent months trying different things to make these particular components work and some of the upper management dictated we had to try everything, including epoxy encapsulation (forming a block of epoxy around the switch) and epoxy sealing of the plastic package. This was essentially the same scenario and same environmental conditions our sleds operate in.
I can absolutely guarantee you that epoxy will not seal the moisture out. Within about a week (depending on how thick the epoxy path is) the humidity inside the relay will be almost the same as outside. When you drop the temperature down below the dew point and below freezing, the water trapped inside will form frost on the inside of the package walls of the relay and over time and through additional minor changes in temperature, the frost will transfer onto the relay contacts. It will take at least a twice as long in a cold dry environment to dry out enough to start working properly again.
We have tried exactly what you are suggesting (even though the engineers, myself included said it wouldn't work - it was dictated that we would try it). We even went so far as to disassemble parts throughout the entire temperature cycle and see exactly when and where the frost developed (we would start the test with 100+ identical items and pull them out, test them and destructively open them up to see where the moisture was).
The closest we got to making the components pass the environmental spec was to vent them using a large area gortex filter. The only time the switches would fail then was when the parts were extremely cold (like -40C) and warm, humid air was allowed to get to the components. We almost got a waiver to accept this, except there was concern after flying at high elevations in very cold temperatures and then landing in a very high humidity environment, such as a rain forest, that things could frost up too much when the doors were opened up. At the time I also spent many weeks building detailed finite element models to simulate the temperature profiles and the amount of frost that would form with different ventilation areas and with sacrificial cold plates inside (chunks of copper or other metal that would remain colder than the relays so the frost would be more likely to form on it instead of the contacts).
With the components well vented, the frost would only form for a very short period of time on the colder metal parts (like the relay). With a big enough vent, after 15 minutes or so they would dry up again and start to work (this is compareable to coming inside while wearing glasses). In our sledding environment a well vented package would probably work fine, but we couldn't quite get them to pass the military spec. Do a search on MIL-STD-810 if you want to find out more on one of the damp heat profiles we were trying to meet (the damp heat spec includes cold).
One last thing I'll point out. If you take air at room temperature (70F) at only 30% humidity and seal it into an enclosure that is perfectly (hermetically) sealed and then bring it down in temperature, condensation will start to form inside once the temperature drops slightly below 40F. If you take it down to 0F there will be lots of frost inside the enclosure.
Have a look at the psychometric chart and at this dew point table:
http://www.coolerado.com/CoolTools/Psyc ... 7US_SI.pdf
http://www.benjaminmoore.ca/professiona ... 0Chart.pdf
Basically you would have to hermetically seal the enclosure (metal, ceramic or glass everywhere) with extremely low moisture content in order to ensure frost wouldn't form inside (it doesn't take much frost on the electrical contact to stop electricity flowing).
If you really want to try it you obviously can. It won't work. It will just change the time constants. The relay needs to be hermetically sealed or vented.
Well then how about take the relay out and use a toggle switch to power on the pump.
