Bouncing temperature gauge

[Lebro]

Since last summer my '3 has been suffering from erratic readings on the temperature gauge.

It is a capillary type, which I calibrated before fitting some 3 years ago, & I have no reason to disbelieve its accuracy.

 

What happens is, from cold the temperature slowly rises, & goes well beyond the normal spot (183°F approx.) to around the next mark up, then it will suddenly dive down to the mark below the normal temp.

It then slowly moves up again, doing the same, but usually dropping down at an earlier point.

Eventually it settles down at the correct place, & will stay there for a while, then occasionally it will start oscillating again,

once again eventually settling down.

I originally thought this must be a thermostat problem (bellows type with shield), but I have tried two other (new) thermostats of the same type, & the symptom remains. The coolant level always settles to be just covering the base of the filler neck (when cold), & does not seem to loose any more from that point.

I have checked inside the thermostat housing to make sure the thermostat is not catching on anything, it all seems clear.

 

Any ideas?

 

Bob.

[RogerH]

Hi Bob,

can you take the gauge and cable off the car and then test it in water being heated from cold to see/eliminate the gauge/cable.

 

Roger

[Lebro]

Well I could, but it's not easy being a capillary type. I can't see how that type of gauge can behave in that way, it's just a pressure gauge.

Also it's not sticking in one place, 'cos it moves quite smoothly from one reading to the other.

 

Bob.

[RogerH]

Hi Bob,

so it could be reading accurately but what it is reading is unstable !!

 

Roger

[Don H.]

I'm with Roger -- one doesn't need to remove the gauge to test it in a different orientation than the engine. Maybe remove the bulb and test it with the tea kettle in the shop?

 

Is there any way it could be mechanical? Perhaps something sticking in the gauge itself?

 

Ya gotta question everything until it's resolved.

[RobH]

I would agree that a capilliary temperature gauge is unlikely to exhibit instability because it is inherently damped by the thermal inertia of the bulb and ether. This does not sound like an accuracy issue - but more like the 'transient' response of the cooling system itself. The first sudden dive in temperature is obviously the thermostat opening initally. I would guess that the gauge is telling the truth and that what you are seeing is the cooling system 'hunting ' before reaching equilibrium. Since you are using a shield-type thermostat is the bypass tube clear? Have you always used that type ot thermostat or did you change anything before the effect started ? Is the fan-belt slipping sometimes so that the the water pump performance changes?

[Lebro]

Agree on your "hunting" theory

I believe the bypass pipe is clear, but will check ASAP

Have always used shielded bellows type

Nothing changed which would affect cooling system

Fan belt tension is OK

 

My bypass hose is a little the worse for wear, so I will order a new one, & will check out the various water passages by flushing when I change it.

My biggest fear is that maybe I am getting combustion gasses passing into the cooling system, & as they rise to the top they upset the temp reading ?

 

Bob.

[RobH]

Hmmm... that's possible I guess Bob, but would the gas bubble not tend to linger at the highest point ? Your coolant level sounds OK - mine is happiest with just a trace of water in the neck too. There is a tester for combustion products in the coolant:

 

https://www.amazon.co.uk/COMBUSTION-LEAK-BLOCK-TESTER-KIT/dp/B018PM3ZFQ/ref=sr_1_1?s=automotive&ie=UTF8&qid=1461530959&sr=1-1&keywords=block+tester

[Ian Vincent]

I suspect your cooling system is clogged up and you gauge is accurately reflecting the temperature in the head as it tries to circulate the coolant.

 

Rgds Ian

Edited April 24, 2016 by Ian Vincent

[Mart O]

Mine does exactly the same, moving between 85 and 80C. This started after I cleaned out the cooling system (2 stage Holts system), new rad core, thermostat (non bellows), put in an air duct behind the grill (for the 1st time) and put in a restrictor on the bypass hose. Im not worried about it, I think it is just the thermostat opening and closing being more apparent now the cooling system is working cleanly with a good airflow.

 

Martin

Edited April 24, 2016 by Mart O

[Lebro]

I suspect your cooling system is clogged up and you gauge is accurately reflecting the temperature in the head as it tries to circulate the coolant.

 

Rgds Ian

Possible, but IMO unlikely. Engine was flushed very thoroughly when it was out of the car, & I have flushed it several times since (every time I have for whatever reason had to drain the coolant) water comes out clear from every orifice (the cars not mine). Rad core is new (2013), & heater was thouroghly flushed when out of car. On one of the flushes I used one of the chemical flushing agents, but it did not get much more gunge out. The drain tap at the rear of the block still works, & after all the flushing now allows water out whenever I open it. So I don't think I have a blockage problem. I will try running engine with rad cap off next time I get it warmed up, & look for bubbles, I could also try running with no thermostat (as a test only).

Any more ideas welcome.

 

Bob.

Edited April 25, 2016 by Lebro

[ianc]

It may be that the shield type thermostat is sticking, then releasing, and that this performance is being repeated.

I would fit a modern Waxstat and restrict the flow through the bypass hose, as many of us have done.

Ian Cornish

[Lebro]

OK Chaps, It's like this:

 

Went for a short run this afternoon - about 10 miles, enough to get up to normal temp.

The gauge was doing it's usual thing - oscillating around the correct point.

Got home, & decided to see if I could detect any bubbles coming from the open rad.

1st thing I noticed was a fair amount of coolant splashed around the front of engine compartment, & I know that the system was only filled to around 1/3rd of the way up the header arm of the rad.

Took cap off carefully using cloth etc., no drama, just a hiss as the pressure was released, level was down to just covering the base of the rad neck.

Topped the level up to about 2/3rds of the neck height with engine running, no visible bubbles, even when revs increased to 3000.

Put cap back on, dried everything up, & waited for pressure to build up again - looking for leaks in rad, hoses etc - none seen.

I then took cap off again, & this time gave it full throttle (just for a moment - not over revving) A gush of coolant shot out of the open filler neck.

I then put the cap back on, & repeated. No drama, but I could see coolant passing into the overflow bottle, & a lot of steam rushing out of the overflow bottle vent holes in it's lid. This happened every time I gave it full throttle.

I'm now thinking partially blown head gasket, which only leaks when under heavy throttle.

I will try torqueing the head bolts tomorrow (this is a factory recon engine not taken apart since it was made by S.T. ! so I am not expecting that to work)

After that I guess it's head off, & who knows where that will end - new liners - pistons - crank bearing etc etc

 

Anyone got any better ideas ??

 

Bob.

[Don H.]

Welcome to the over 83mm club, Bob. I'm an 87mm member...

 

is there any sort of bypass hose/hyperactive water pump scenario that would work? Heat going up and down doesn"t exactly sound like head gasket. and there's no disappearance of coolant?

 

I'd try to confirm with one of those pH-based combustion-gasses-in-the-coolant testers, either as a purchase for your kit or at a friendly garage.

Edited April 30, 2016 by Don H.

[Ian Vincent]

I had similar symptoms to you Bob and when I removed the head there was a lot of crud around the liners but no clear evidence of a blown gasket. The liner protrusions were not correct though.

 

Rgds Ian

 

PS lifting the head only took a couple of hours or so, the hassle is removing the liners.

[Guest ntc]

OK Chaps, It's like this:

 

Went for a short run this afternoon - about 10 miles, enough to get up to normal temp.

The gauge was doing it's usual thing - oscillating around the correct point.

Got home, & decided to see if I could detect any bubbles coming from the open rad.

1st thing I noticed was a fair amount of coolant splashed around the front of engine compartment, & I know that the system was only filled to around 1/3rd of the way up the header arm of the rad.

Took cap off carefully using cloth etc., no drama, just a hiss as the pressure was released, level was down to just covering the base of the rad neck.

Topped the level up to about 2/3rds of the neck height with engine running, no visible bubbles, even when revs increased to 3000.

Put cap back on, dried everything up, & waited for pressure to build up again - looking for leaks in rad, hoses etc - none seen.

I then took cap off again, & this time gave it full throttle (just for a moment - not over revving) A gush of coolant shot out of the open filler neck.

I then put the cap back on, & repeated. No drama, but I could see coolant passing into the overflow bottle, & a lot of steam rushing out of the overflow bottle vent holes in it's lid. This happened every time I gave it full throttle.

I'm now thinking partially blown head gasket, which only leaks when under heavy throttle.

I will try torqueing the head bolts tomorrow (this is a factory recon engine not taken apart since it was made by S.T. ! so I am not expecting that to work)

After that I guess it's head off, & who knows where that will end - new liners - pistons - crank bearing etc etc

 

Anyone got any better ideas ??

 

Bob.

Why a overflow bottle? and you are over filling the expansion neck

[Lebro]

Neil.

I normally fill to about 1/3rd up the expansion neck, I over filled today to try to see if any bubbles were evident.

Overflow bottle was belt & braces to try not to lose any coolant, & to allow filling neck to a little higher than normal level.

 

Looking at the engine / rad levels, by the time the coolant has dropped to being level with bottom of the neck, it is below the centreline of the thermostat housing which may explain the varying readings.

 

Don.

But I do lose coolant, it gets sprayed all over the front of the engine after filling to normal level. It only stops spraying when the level drops to the point where it only pushes out gas.

 

Shame this is happening now, just when the weather will be warming up. Will check head bolts tomorrow, & if all tight already I guess it's head off time.

Must get it sorted before the IWE

 

Bob.

[Guest ntc]

Bob

It should only just be visable in the bottom of the neck, the bottle will just draw the water back in. Can you post a pic of your setup? You are jacking the front of the car up ?

Edited April 30, 2016 by ntc

[Don H.]

I musta misread your earlier post, Bob -- sorry 'bout that. I thought it was only puking with the cap off and engine revving.

 

If you have combustion gasses (CO2) in the coolant, it'll quickly show up in those tests and is confirmation of blow by the head gasket.

 

When my car was back on the road right after the engine rebuilt, it liked to run hot. As in coolant spraying all over the engine compartment no matter what the revs or road speed. Very frustrating. I looked at all kinds of things, from water pump, thermostat, radiator cap, radiator, coolant passages, mixture, timing, airflow... I'm not sure one thing alone fixed it, except for the loosening up over a thousand or two miles. Fingers crossed it's reasonably well-behaved now. I put the overflow bottle back on the car -- it had one when I got it thirty five years ago that had been fitted many years before that. Folks advised me to let the system find its equilibrium, that the coolant would drop to the level where it balanced expansion and everything would be fine. The steady increase in the temperature gauge argued against that logic. Perhaps it would work with a normal, coolish engine, but not one inclined to run hotternhell.

 

The overflow bottle is about 1/4 filled when cool and will rise to about 2/3 when hot. The radiator cap allows overflow and sucking back in perfectly. The normal level in my filler neck is right up to the rim. Guidance for coolant partly visible in the filler neck is appropriate for no overflow bottle, as far as I can tell, but not with an overflow bottle.

 

I might have expected at least some hiss when you removed the cap from a hot system. Are you confident your radiator cap is holding the right pressure? Could it be boiling over as the system doesn't allow even the modest 4 psi specified?

 

If you're losing enough coolant to expose the temperature bulb, or partially expose it so that splashing coolant alternates between coverage and air exposure, I guess that could explain your oscillating needle.

Edited April 30, 2016 by Don H.

[Mart O]

If you have clear waterways and an efficient radiator, why wouldn't the temperature steadily rise until the thermostat opens. At which point freshly cooled water now gets pumped round. If the water has sufficient cooling effect to take the exit water to below the temp at which the thermostat closes, it will close. The cycle then repeats. Try a cooler thermostat, to see if it remains open and the temp gauge no longer oscillates?

 

Martin

[Lebro]

I suspect the coolant loss has been going on for some time, thinking about it that is probably why I fitted the overflow bottle, to catch the "blown out" coolant, & let suction pull it back in as the engine cools.

I have a 4PSI cap, with the longer spring / seal assy to match the deeper rim in the "long neck" type of rad filler. I modified the cap by removing the steel outer seal, & replacing with a cork one, thus achieving the 4psi seal on the lower rim, & a complete seal at the top rim. this was required to get the vacuum required to pull the coolant back from the overflow bottle. basically same setup as later TR4's.

 

My thinking is that the fluctuating temp as shown by the gauge is due to the thermostat housing being filled with gasses whenever I apply hard throttle.

 

So I am thinking head gasket letting gasses through on heavy throttle, or worse - a crack in the head. I will check head bolts today, & order a combustion gasses test kit as well before pulling the head.

 

By the way, heater works fine soon after starting up, even with lowish coolant level, so pump is OK, bypass circuit is ok, & waterways should be clear.

 

Bob.

 

Bob

It should only just be visable in the bottom of the neck, the bottle will just draw the water back in. Can you post a pic of your setup? You are jacking the front of the car up ?

 

Neil. It was intended to "pull the water back in".

Car not jacked up, but was on an upward slope yesterday when doing the checks.

Nothing special about my setup, just an overflow bottle sitting just above the steering box, connected to the overflow pipe on filler neck, pipe going the the base of the bottle, & bottle cap has three ¼" holes in to allow breathing, & the rad cap modified to allow suction as detailed above.

Edited May 1, 2016 by Lebro

[ianc]

To overcome this blow/suck problem once and for all, fit a separate, pressurised, header tank - the system used on more modern vehicles. See my article in Section B19 of the Technicalities CD.

This will ensure that the coolant fills completely the thermostat housing and the whole of the radiator (including the neck, if it has one), as expansion/contraction takes place within the header tank.

Ian Cornish

[Lebro]

Appreciate all the above suggestions, but this time last year I did not have bouncing temperature,& nor did I have a coolant loss problem, so something must have gone wrong.

I have tried three different thermostats (at two different temperature ratings) - no difference except in the "norm" temperature reached.

I have searched for leaks while engine hot, running & under pressure (4PSI) none found

It only ejects coolant if I open the throttle fairly wide.

Anyway more tests today, will report back !

 

Bob.

[dpb]

Hi Bob

 

I had coolant expulsion and loss, and high temperatures last year, and also sometimes normal temps. Mine turned out to be a blown head gasket. Finally identified by one of those gas tester kits.

 

Dave

====

[Lebro]

Just ordered one off ebay !

 

Bob.