Bongo Fury

Prior to going on a 30mile trip I checked the water level in the header tank but stupidly did not replace the cap leaving it on top of the plastic air intake on the engine.I have the Haydn Combo alarms fitted and I regularly check engine temp on every trip.My normal running temp is about 94 but it can go higher but fans kick in around 97 so thats as high as it goes.After 25 miles or so without the cap on the temp showed no sign of increasing but during heavy breaking down a steep hill the coolant loss alarm sounded momentarily which alerted me to the cap situation .I was lucky to find the cap still where I had left it under the bonnet.Why didn't the engine temp rise?Why didn't I lose more water as the coolant alarm didn't sound on a level road ?The coolant alarm sensor is correctly positioned in the header tank .

All sounds fine to me. As long as header tank didn't empty and introduce an airlock, I'd top up put header cap back on And make note to self not to do it again. All that happened was you drove 25 miles with an unpressured, unsealed system. If you were on open roads, in this cool weather, cooling system probably wasn't under much pressure, as it were

But if you'd not had a Haydn alarm.......

The main purpose of the sealed system is to increase the boiling point of coolant. It will work unpressurised but if it gets to the point of boiling, the coolant will be lost by evaporation and you stand a chance of it slopping out while driving round a corner.

You've had a lucky escape there & proved yet again the benefit of a low coolant alarm. There are plenty of folk who are adamant that their cooling system is in perfect working order so why bother to fit one. Even if you put aside the fact that rubber hoses that are coming up to 20 years old are going to have deteriorated, steel pipework will have rusted, aluminium radiator cores will have corroded, header tanks can split, there is still the weakest chain in the link: the person behind the wheel

Been there, done that......

Mine did start overheating, but it was the Mason alarm that alerted me. Water still in the tank but bubbling, so the LCA didn't go off. Be warned! Fotunately no damage.

Thanx for the reassurance and advice all. One other question you might like to answer please.I always struggle getting the level correct in the header even using the dipstick type method due to the cloudy looking plastic tank I'm wondering whether it would be ok to put a small amount of black printing ink into the tank to enable me to see the level more clearly? I have red coolant in at the moment and it is not easy to see through the plastic.Good idea.?Bad idea ?

I find a bright torch somewhere near the front left corner (as you look at it) illuminates the level sufficiently well. When you next have the header tank drained, you could try removing it and adding some Hyperclean or bleach (?) and a handful of gravel - then shake well. I don't think there is a "best" answer, though perhaps more suggestions...

You could get a clear type drinking straw, put a permanent mark 20/25mm from the end, remove cap insert straw to bottom of the tank put your thumb/finger over the end of the straw, lift straw to get reading, release thumb/finger to allow coolant back into the header tank from the straw, refit header tank cap.


Find a suitable place under the bonnet for the straw for future use.

rita wrote:You could get a clear type drinking straw, put a permanent mark 20/25mm from the end, remove cap insert straw to bottom of the tank put your thumb/finger over the end of the straw, lift straw to get reading, release thumb/finger to allow coolant back into the header tank from the straw, refit header tank cap.


Find a suitable place under the bonnet for the straw for future use.

If you make a few you can put them on Ebay...."Mazda Bongo Header Tank Visi Gauge" £10.00 inc P&P.

Thanx for your ingenious solution Rita.Does the 25mm mark relate to the lower level ie minimum ?

Some say backlighting or shining a torch through the header tank does the job. An alternative to clutching at straws

If you put a DRY lolly stick or simerler vertically down into the tank it should be easy to read....
The alarm has a built in delay before it goes off to prevent false alarms caused by the coolant sloshing about.
Do not put dye/ink in the tank unless it's a leak tracing fluid.

Thanx Hadyn I'll probably come across as being a bit dense but I find the "dipstick" method very awkward.Sticking the lolly stick in as you suggest gives you a reading of the water level but how does that relate to the upper and lower marks on the tank?Could you tell me what would be the measurements of of the two lines to mark on the lolly stick to indicate the level required.

You need to read 2 cms on the stick....that will be near enough the full line......do it when cold as it will expand another cm when hot

Thanx for the info that's exactly what I wanted to know.A follow up question arises.What would be the consequences if any of overfilling?I would like your opinion also of my experience yesterday while travelling at highly illegal speed down the M5 in warm weather.My running temp is usually around 94 but can rise above this but cooling fans kick in at 97/98 to reduce it down again.However on Sunday after leaving the motorway and parking in a lay by to consult a map the temp alarm sounded which I have set at 105.I know that the temp will rise when stationary but I was suprised to see it rise to 110 with the cooling fan on with the ignition switch.I had been travelling at speeds of 80+ but I wondered whether there was of a danger of a stationary boil over?What temp is boiling in a pressurised system?

Here's some of my understanding - I hope it helps and doesn't confuse things...

banjoboy1 wrote:What would be the consequences if any of overfilling?

In a healthy system.... simply put, overfilling with coolant means the header tank would have less air in it. There being slightly more coolant to expand when hot, it would squash the air that is in the header tank a little more and you'd reach the 1.1 bar pressure of the header tank cap a bit sooner, which, if operating correctly, would then vent off a little air to ensure the overall pressure stays at or below 1.1 bar. In the extreme, the coolant may thermally expand enough to leave "no" air in the header tank, at which point you will blow off excess coolant via the header tank cap pressure relief cap and overflow hose, and end up with a soggy mess under the bonnet, not in itself a problem. On cooling, the coolant will contract, and the pressure of however much air is in the header tank will drop, if dropping to -1psi, then the header tank cap will suck in a little more air. In short, "it will find it's own balance", but overfilling is not necessarily a bad thing in itself. That said, you may see something like 10-13psi in a "normal" system (though there are many factors) and overfilling will mean the system is "more likely" to see maximum pressure around 16psi (or 1.1 bar), at which point you may be "stressing" hoses more than you might otherwise. It's complicated again if you've just bled the system and are sealing an unpressurised, fully expanded system by putting the header tank on when hot, so you need a couple of heat/cool cycles for it to "settle". Also, it is possible that some other part of the system is going to allow air "in" before the negative pressure valve of the header tank cap - it is generally thought that small amounts of air can get sucked in variously around thermostat seal, waterpump seal, hose stubs, etc..., "instead" of the header tank cap - in these cases, then you may incrementally build up an air bubble somewhere in the system other than in the header tank cap, which if not removed by turbulence during engine operation, would be "a bad thing". There is no need to overfill, other than perhaps straight after bleeding when you're already at temperature, with thermally expanded coolant, and know that there may be the odd little bit of air to come out under driving conditions, overfilling in that case to ensure that when the level drops due to air expulsion and subsequent cooling/contraction, it doesn't drop out of the header tank completely.

banjoboy1 wrote:What temp is boiling in a pressurised system?

Temperature of what, and measured where? What %age mix of antifreeze to water? The hottest part of the cooling system is going to be the waterways around the combustion chambers of the block and exhaust ports of the head. The coolant should be able to tolerate this maximum heat, which you will probably find difficult to measure. "Localised boiling" around these parts is what is going to cause the issue, as steam will expand violently with respect to the coolant from which it comes. If you're measuring metal temperature some distance away from here through metal thickness(es), you're measuring something slightly different, so the temperature of coolant that may have already reached boiling point is not necessarily going to be consistent with the figure on your temperature read-out, though in practice I'm sure it's good enough to make some judgement - I'm just taking my explanation a little further. Furthermore, if you've got an air lock or boiling in progress, then coolant will not be circulating and the cylinder head temperature will rise significantly higher/faster than the point at which you're measuring the temperature either of coolant or of metal surface temperature, until you lose coolant, the heat reaches the temp sensor, etc... and then stop. This is more or less how the damage can occur. Some parts of the system will always be hotter than others - obviously the rad and heater matrices remove heat for example. If you've got airflow under the van and around the engine, you've got heat exchange going on in the rad/matrices and also a bit more cooling of the metal bulk of the engine itself, versus being stationary where the hotter parts of the engine are slowly soaking the rest in heat which is not being removed very much, etc... If your engine is off and you haven't boiled over yet, you're not going to - as you're not introducing any more heat, even if your temperature readout appears to be rising due to heat soak. If you leave the engine running to keep coolant circulating, then the thermostat will open and rad fan(s) will come on to ensure that the temperature of the coolant does not keep increasing. A long way round to say that a heat rise when stationary is not uncommon, the figure you're seeing and the alarm threshold which you have set, may not be unusual themselves...

When you rose to 110, was the engine running to circulate coolant? Or was it the "scavenger fan" in the engine bay running rather than the radiator fan(s)? If the rad fans were to continually cool the same coolant that is in the rad without the engine running and hence the water pump circulating it and replacing it with hot, then that's not going to reduce engine block/head temperature much. If the scavenger fan is blowing air through the engine bay but the coolant itself is not being actively cooled, then again, temperature won't come down particularly fast.

Others with the same temp gauge installed in the same position, may be able to share their experience regarding specific temperatures, etc... that may be more what you're after...