MG-Cars.info

Welcome to our Site for MG, Triumph and Austin-Healey Car Information.

Parts

MG parts spares and accessories are available for MG T Series (TA, MG TB, MG TC, MG TD, MG TF), Magnette, MGA, Twin cam, MGB, MGBGT, MGC, MGC GT, MG Midget, Sprite and other MG models from British car spares company LBCarCo.

MG MGB Technical - Coolant loss

I'm losing coolant but there is no sign of oil in the coolant or coolant in the oil. I fitted a coolant overflow tank (well a plastic bottle actually) but there's nothing in that. No obvious leaks from hoses, which are all fairly new. Does that sound like a head gasket problem do you think? Performance seems ok, a little bit stutter under load but might have been the very cold weather. Are there any simple tests I could do to help isolate the cause.
P K Newbery

It would be useful to know the age of your car and the type of cooling system it has. Does the radiator have a cap? Or is there an expansion tank with a pressure cap? Over the 18 years the MGB was made there were many changes and without knowing where your car fits into the time line any diagnosis is difficult.
Mike Howlett

More details would help but another head gasket variant is combustion gasses in to the coolant. An (older type) garage will have a device that can be stuffed in place of the cap and a dye changes colour is combustion gasses are present. Might also pay to peer down the air intake in front of the screen with a torch to see if there is evidence of the heater maxtrix leaking. Only works if the car has not been out in the rain for a while!! How good is the rad cap seal?
Michael Beswick

As well as what year what rate of loss?

It would be very unlikely to have a head gasket leak that let coolant into the cylinders but not combustion gasses into the water jacket, the latter is more likely.

My Halfords has the combustion leak detector Michael mentions and was the final diagnosis of my headgasket leak. Based on other symptoms it had probably been there the 26 years I had owned the car, until it got bad enough to show significant oscillation of the temp gauge during warm-up, and loss from the overflow.

Incidentally under normal circumstances a catch-bottle on the overflow i.e. unpressurised, if the overflow tube goes right to the bottom, during cool-down will suck back up any coolant that was expelled during warm-up or running. The catch bottle would need to be checked when the engine is fully hot.

You can also pressurise the cooling system from an external pump, or monitor the pressure with a gauge, by putting a Tee in one of the heater hoses of pre-77 cars i.e. those without a remote expansion tank, or in the expansion tank hose of 77 and later cars.

I don't think you can see the matrix from the air inlet at the base of the screen, leaks in the matrix usually appear as a wet floor either or both sides of the tunnel in the cabin.
Paul Hunt

<< leaks in the matrix usually appear as a wet floor either or both sides of the tunnel in the cabin.>>
That's right - been there, got the T-shirt. A further test is to poke a finger into the warm air outlet down in the footwell. If the heater is leaking there will be a small pool of coolant in there.
Mike Howlett

Another possibility, if the car is an early one, that has the filler neck out the back. On mine I kept having coolant loss, same as you. What I eventually discovered is that normal expansion blows excess coolant out the overflow, then when the engine cools down the coolant level drops to a level where you can't see it. in the neck. Not really a coolant loss, as such, but a real PITA.

I fitted a catch bottle and fully sealed cap, problem gone.

See this article on what I did.

http://www.mgb-stuff.org.uk/saga10.htm

Herb
H J Adler

Water loss into the combustion chamber will be noticed by the sweet smell of the exhaust and the clouds of water vapour. More difficult to spot the steam at this time of year. If you remove the plugs you will also notice a very clean plug.
But how much are you losing?
Allan Reeling

Sorry, didn't realise the car details don't show in the sign off. It's a 71 Roadster. Rad cap is a couple of years old, but I'll check the fit. No expansion tank as standard so I made one up but catch nothing. I'll do some more scientific loss measurement but it's enough after 10/15 miles to disappear from the header tank. I've had the car since 94 and it's never been possible to keep a full header. It always spits out a pint or so, but then holds. Just lately it seems to have got worse. Thanks for the interest and help though.
P K Newbery

I had a '71 too. Mine never had any coolant in the header tank - you could just see it on the top of the fins. Topping it up was a waste of time as it just expelled it again. I came to view the header tank as necessary expansion space. The engine always ran well with no overheating and once the header tank was emptied, no more coolant was lost.
Mike Howlett

As ever these cars do vary hugely. My 73 can easily keep an inch or more above the tubes, that's when cold. Air trapped in the system could well cause the level to rise more when hot, even though it should be under pressure.

But if it were ejecting it from overfilling during warm-up, it would be in a catch bottle on the overflow tube, sealed or not, as long as you check when hot or the tube doesn't go to the bottom of the bottle.

A pal's 78 regularly spat some out during heat-soak at switch off, a pressure test (Tee in expansion tank hose) showed the cap was only holding 2 or 3 psi. With a new cap and retested to confirm it was OK it was fine.
Paul Hunt

Thanks each. I just came in from a (chilly) 15 mile run, and after moving the car from drive to garage saw coolant on the drive. Quite a lot. Looked underneath and could see evidence of dirty wet stuff on the rear of the front cross member. I can find no wetness at any of the hose clamp points and the overflow pipe and catch tank are bone dry. The heater connections and heater control valve are all dry. No damp carpets. The bottom of the rad is dry and the rad itself is unstained anywhere. So now I'm thinking water pump or hardly dare say it.....cracked block. The 15 mile run was lovely. The car went really well and is very responsive. No hesitancy at all. At the moment my 4 ramps are in use raising up the good old M Type so I can't get under the B properly.
P K Newbery

Have you checked that coolant is not weeping from the water pump / water pump gasket area? Last time I changed my water pump , I had a mysterious leak that occurred when the car was cooling down which I traced to one of the water pump bolt areas,notwithstanding the fact used new bolts,new gasket & gasket paste. It was difficult to trace as partly covered by the pulley (I know the set up is different on a 1980 car compared to yours but worth drawing your attention to a possible source of your problem ).
CMD FARRAN

If you have antifreeze in that is usually quite good at showing itself by staining.

The other thing is to top it up when cold and warm it up on the drive having a good look round the water pump area, minding the fan and belt of course.

The pump can also leak from the spindle seal, which can drip from the back of the pulley and be less easy to spot. If by waggling the fan blades the pulley moves, then the bearings have had it anyway.
Paul Hunt

OK, thanks Paul, I'll do that once hurricane horrible blows over. I'm hoping it's a simple as the water pump, although on the B it's a bit of a pain having to remove the rad and battle with the bottom hose. Could be worse.
P K Newbery

When you installed the overflow tank, did you replace the standard radiator cap with one that incorporates a double seal? The second seal prevents the overflow coolant, coming out of the radiator, from bypassing the the line going to your overflow tank. Without the second seal on the radiator cap, the coolant will not be captured when it exits the radiator and will not be drawn back into the radiator when it cools down. RAY
rjm RAY

UK rad caps have always had the double-seal as far as I'm aware, although I understand that some American caps do not.

With standard UK caps I've used coolant recovery with nothing more than a 'open to atmosphere' plastic bottle with the overflow tube going into it - and reaching the bottom - on both my cars when they have had cooling system problems.
Paul Hunt

Water pump is leaking at the bottom. Not sure yet if it's seeping through the joint faces or through a bearing seal. Either way, I'll replace it.
P K Newbery

Paul, until overflow recovery became standard fitment on new cars in the late '60s, all radiator caps only sealed the radiator neck internally. None had the dual seal that is standard today. RAY
rjm RAY

Overflow recovery? I've not known a car with that as standard, any overflow went onto the ground. When I said 'always' I was referring to MGBs, and the Parts Catalogue indicates the MGB always had the double-seal cap.
Paul Hunt

New water pump fitted, radiator back in place and now have to tighten up hose clamps and test. I read that you should run the system dry for a minute or two to bed in the pump carbon bearing. Brown and Gammons thought that unnecessary. Just fit it and go. Any thoughts? And while I'm at it, what's the recommended way to ensure the heater isn't air locked. Just to run it with the return pipe left open?
P K Newbery

Running the system dry is recommended for the newer carbon seals. This mates them to one another, when new, and insures a longer lasting seal. Paul, I've never seen a B fitted with a double sealed radiator cap that didn't have an overflow container fitted. It would serve no purpose. RAY
rjm RAY

Thanks rjm.
For what it's worth, there is no overflow recovery shown in Clausager photos. Even the 1977 US spec car with separate header tank has an overflow to earth.
P K Newbery

The double-seal cap is there for safety, the fact it also allows for coolant recovery is just serendipity. If you were leaning over the rad when the cooling system suddenly decided to evacuate itself, and the cap didn't have the upper seal, then scalding water would spray all over the place. With the upper seal it can only escape through the overflow and down to the ground in relative safety.
Paul Hunt

The earlier radiator caps, without the double seal, simply used a piece of round spring steel to prevent the hot water from escaping out of the neck of the radiator. It worked very well until the introduction of the overflow recovery systems that utilized the double sealed caps. RAY
rjm RAY

A "normal" cap has the bottom seal, which opens at set pressure. It also has a top seal to prevent water from escaping past the cap. Recovery caps have these same two seals, but they also have a valve that allows the water to be sucked back in. Normal caps don't have this valve.

Herb
H J Adler

OK I was somewhat wrong about the seals and valves. Both caps have the return valve, but the normal one doesn't have a seal at the top, just a brass disk, that directs any splash downwards, but doesn't stop air being sucked in.

The recovery cap has a rubber seal here so that air cannot be sucked in. See photos.

Herb

H J Adler

Interesting stuff. Meanwhile, re my post of 19:43 on Jan 28, will I reduce the chance of airlocks in the heater if I run the engine for a mo with the heater return disconnected. My heater currently blows slightly less than warm.
P K Newbery

You need everything connected or you cannot fully purge the system for normal use. It helps during the initial warm-up if the nose of the car is slightly higher than the rear, certainly not lower. Top-up when cool. A couple of heat/cool cycles after that should be all that is required to fully purge it.

There can be 10 or more other causes of poor heater output. Some of these are lack of coolant flow from the head, through the valve, the matrix, and the return pipe back to the bottom hose. This should be obvious by feeling the heater hoses and the return pipe.

The air/side of the matrix could be blocked with rubbish, and the foam seal between the matrix and the box could have decomposed which allows air to bypass the matrix.

The heater vent at the bottom could be stuck in the closed position.

The cold-air flap could be open nullifying the effect of the heater.

The thermostat could e stuck open meaning the system never gets up to temperature, which should be obvious from the temp gauge, and also from feeling the rad header tank during warm-up from cold. This latter should remain cool while the temp gauge rises, then suddenly should get very hot so take care.

This rad cap thing could run and run. I've only noticed the additional rubber seal against the spring steel seal on a recent replacement cap. It made the cap a beggar to fit and remove as it was thicker than the available space, so I've removed it.

Previous caps, and the one that has been on my V8 for years, do not have the rubber seal, just the spring steel. That spring steel seal is shaped to sit away from the top of the cap, so it is pressed down on the upper neck of the rad when fitted. You can clearly see the shiny witness marks on that seal (attached) and on the neck of the rad where they have been rubbed together as the cap is fitted. YMMV. It is true that if that seal, or the neck of the rad, are not flat and smooth, it would impact on vapour recovery. But as I say I've had coolant recovery working on this cap and others without the rubber seal on more than one occasion.

Paul Hunt

This thread was discussed between 23/01/2016 and 30/01/2016

MG MGB Technical index

This thread is from the archives. Join the live MG MGB Technical BBS now