Coolant leak - otter switch

[rbgosling]

My project over the New Year was to try and keep all the coolant inside the coolant system. Well, the project was originally to pull the dash and replace the heater matrix, since the continuing unexplained loss off coolant and large puddle in the passenger footwell pointed that way. But I was wrong, the heater box was undoubtedly the driest place in the car, so that was a waste of time (well, almost, I found a couple of other details to fettle while I was in there...).

Anyway, with the use of a coolant pressure testing kit I chased down a number of minor leaks at different hose ends, all improved by tightening (or replacing) the hose clips. Each time I could pressurise the system a bit more, until I heard a new hiss from somewhere else. Eventually the remaining hiss came from the otter switch - see this video.

For the testing I had drained all the coolant and pressurised the system with air, on the basis it's much easier to detect the hiss of escaping air than try and spot the very slight oozing of a tiny drop of water. Dribbling a mixture of washing up liquid and water helps prove the location with many bubbles.

So, I recall that I replaced the otter switch 4 years ago. On that occasion I tried using Loctite 518 flange sealant ('cos I happened to have some), which seemed at the time to seal OK, but I didn't have the pressure tester back then. This time I first tried Hylomar Blue (useless), then this great stuff called Fernox LS-X which my father always has some of as it seems to seal anything. It got to about 12 psi before I heard it hissing (cap pressure is 16 psi).

So, before I start just throwing random goos at the problem until something sticks, has successfully managed to seal this with something?

Also, the original was held in with twisted wire, which I would imagine is capable of putting a bit of compression into the joint if twisted enough. I used a cable tie, but it's possible this isn't compressing the joint so well, so could this make a difference?

[Lotus-e-Clan]

I seem to recall that my pressure cap setting was only 10 psi (before converting to Evans @ 0 psi). Not sure what's OE these days (is it in the service notes?), but 16 psi seems high. My other wedge (Clan) used only 4psi before converting that, too, to Evans @ 0 psi.
I know some modern cars running hot lean-burn engines use around 20 psi, but is 16psi really necessary for the SE 912 , or is the potential for local boiling at the wet liners the reason for 16 psi?

Regarding the otter flange seal. Tie wraps won't work for long cos they get hot and stretch. I'd go back to twisted wire and a new grommet with no sealant. If it can't seal with the grommet alone it's time to inspect the mating surfaces and renew as necessary. Or reduce cap pressure?

[rbgosling]

The cap came from SJ Sportscars, attached to the new Aluminium header tank I fitted last year. I therefore assumed that they know what they are doing and it is the right pressure.

[Alan_M]

On mine the otter switch is a tight fit in the rubber bung and expands it to make the seal when inserted. This rubber goes hard over time and will stop sealing. So ideally get a new rubber bung and your problem should, hopefully, be solved.

[rbgosling]

On mine the otter switch is a tight fit in the rubber bung and expands it to make the seal when inserted. This rubber goes hard over time and will stop sealing. So ideally get a new rubber bung and your problem should, hopefully, be solved.

Yeah - mine should be the same, but the fit isn't particularly tight. The seal was replaced alongside the switch 4 years ago, and the rubber still feels supple.

With no sealant it didn't come close to sealing, I found that both when I installed the new switch and this time. Could be the housing is a bit corroded (I did try cleaning it up with a wire brush on a Dremel-a-like), making it harder for the rubber to make a perfect seal.

[AndrewWebber]

I seem to recall that my pressure cap setting was only 10 psi (before converting to Evans @ 0 psi). Not sure what's OE these days (is it in the service notes?), but 16 psi seems high. My other wedge (Clan) used only 4psi before converting that, too, to Evans @ 0 psi.
I know some modern cars running hot lean-burn engines use around 20 psi, but is 16psi really necessary for the SE 912 , or is the potential for local boiling at the wet liners the reason for 16 psi?

Was out in the Elite today, it is 10+ years older of course so may not be relevant. Checked the cap it says 10 psi. Checked the otter, wire in place, no sign of any leak stains around it, but do remember one blowing out (that had no wire in place) in my +2 30 years or so ago. Once wired up no problems after, 7 psi cap.

Andy

[Alan_M]

On mine the otter switch is a tight fit in the rubber bung and expands it to make the seal when inserted. This rubber goes hard over time and will stop sealing. So ideally get a new rubber bung and your problem should, hopefully, be solved.

Yeah - mine should be the same, but the fit isn't particularly tight. The seal was replaced alongside the switch 4 years ago, and the rubber still feels supple.

With no sealant it didn't come close to sealing, I found that both when I installed the new switch and this time. Could be the housing is a bit corroded (I did try cleaning it up with a wire brush on a Dremel-a-like), making it harder for the rubber to make a perfect seal.

If the bung is a bit of a loose fit in the orifice, could you wrap the bung in 1 or 1.5mm thick strip of rubber sheet. Would need to possibly chamfer ends and overlap them for best seal?

[AndrewWebber]

Perhaps cut the finger off a 'rubber' glove to make a tube, hence avoid the join issue?

Andy

[rbgosling]

Hmm... it's an interference fit, for sure, just not that tight an interference. I appreciate the idea, but I'm skeptical that adding another component just gives more routes to leak through.

I've got some lockwire on order (7.5m of the stuff, when I only need 7.5 cm or so!), I'll give that a try getting it nice and tight in combination with some fresh sealant, and see how it goes...

[Pete Boole]

Or just replace the whole thing with a solid state version

Only involves the tiniest bit of project creep!!

Pete

[ianhateswork]

Would the later version with the screw in sensor not give what you need, or something like these,

ian.

https://www.ebay.co.uk/sch/i.html?_from ... _PrefLoc=1

https://www.ebay.co.uk/sch/i.html?_from ... _PrefLoc=1

[rbgosling]

Would the later version with the screw in sensor not give what you need, or something like these,

ian.

https://www.ebay.co.uk/sch/i.html?_from ... _PrefLoc=1

https://www.ebay.co.uk/sch/i.html?_from ... _PrefLoc=1

Mine is a 1990, so I don’t think there were any changes after that, I don’t know about a screw in one? But those parts you sent links to look like a great solution, I’m tempted.

[Lotus-e-Clan]

Would the later version with the screw in sensor not give what you need, or something like these,

ian.

https://www.ebay.co.uk/sch/i.html?_from ... _PrefLoc=1

https://www.ebay.co.uk/sch/i.html?_from ... _PrefLoc=1

Mine is a 1990, so I don’t think there were any changes after that, I don’t know about a screw in one? But those parts you sent links to look like a great solution, I’m tempted.

My Christmas trip fan fail issues will have to be sorted before mine is taken off a SORN in spring. Apart from the fan relay issue I had, the wider picture includes the fact that the otter switch replacement more than 10 years ago (a 32 mm Revotec solid state switch) became flakey 2 or 3 years ago, and has now failed.... and I don't want to buy another (for fear of repeat falure).

I'm intending to replace it with the said 32 mm inline fan switch adapter and a standard temperature switch (if I can find a 90 C version). But some of the std fan switches sink to earth, unlike the original otter switch which switches 12v, IIRC. So some fundamental rewiring may be necessary if going down that route.

[MetBlue]

Would the later version with the screw in sensor not give what you need, or something like these,

ian.

https://www.ebay.co.uk/sch/i.html?_from ... _PrefLoc=1

My very late car has this conversion. PO purchased the Ali in line housing and a switch from Car Builder Solutions,.

The fan switch is an M22 x 1.5 thread, readily available on Ebay.
FWIW, Car Builder part Numbers were MOIN32 and FANSW5, but will be cheaper on Ebay.

I prefer this solution, No need for tie wraps or wire to hold it in - just a good old thread

Tony

Tony

[Lotus-e-Clan]

The fan switch is an M22 x 1.5 thread, readily available on Ebay.
FWIW, Car Builder part Numbers were MOIN32 and FANSW5, but will be cheaper on Ebay.

Tony, thanks for that Car Builder heads-up . However, a quick browse through the range of Car Builder fan switches comparing to Service Notes values, has given me exactly what I want to fit to mine, but it's not the M22 FANSW5 item.

The M22 FANSW5 makes/breaks @95/86C.

I'm going for the M14 FANSW which makes/breaks @87/82C.

Please challenge my thinking on this in case I'm missing something , but the logic behind my thinking is that in the Excel Service Notes, the latest Otter switch values are 92/82C for temperate climates.

The break value of 82C makes sense because the fans will bring the system temperature down to the opening-closing thermostat value of 82C. At this point (approaching stat closure) the slower coolant flow through the rad means the fans will drop the radiator temperature more efficiently. And the lower overall system temperature will better keep a lid on the oil temperature and the resultant oil pressure drop during stop-start traffic.

For me, the potential issue with 95/86C is that the break point of 86C means the 82C stat will still be wide open, so the rad fans will have a drawn-out fight against the incoming faster-flowing hot water passing through the rad.

Also the knock-on resultant higher oil temperature will drop the oil pressure even lower during stop-go traffic. You get a seemingly disproportionate drop in oil pressure at tick-over with +4C or more additional oil temp when you are already in the 80 - 90C range.