Hot water light randomly blinking

[Dwight]

EDIT: PROBLEM FIXED, ANSWER ON PAGE 2

Hello fellow Jazz owners,

First of all, what a car. I love it. Got it from my stepfather who after lending it to my brother, had it back with a knocking engine (brother didn't know the oil light turning on wasn't a fashion statement by the car), decided to put it on the scrap, and I said hell no, give it to me, I'll make it shine bright again.

Changed the engine about 1 and half year ago, car had 240.000km and now about 255.000km. The new L13A1 engine is about 190.000km as of today. In the process I took advantage of the occasion of changing the engine to fully service the car, so have done about anything I could think of at the time, like engine mounts, new belts, oil and fluids services, filters, spark plugs, the clutch and even the gearbox oil etc.

Ran like a charm eversince. It's a car which runs on e85 most of the time. About 70% E85 30% SP95 in winter for smoother start up, and when the weather gets warm mid spring, it's 100% e85 until the last days of autumn.

Maybe this is irrelevant to some, but maybe to some it would be to precise all of those things, like changing the engine and using E85 as fuel daily.


So the thing is, about a week or so ago, I was cruising on the ringroad, and suddenly, in a slope, going about 90 km/h, the red hot water light flashed a few times and kept doing it for few seconds. Hopefully I was about to take the next exit, so I did. And a few minutes later it did it again while cruising at about 50 km/h in the city, on a boulevard.

Kind of freaked out, went home, checked the water level, and doesn't seem like there is any leak from anywhere, it doesn't even seem like water has dropped since the engine and the new fluids have been put on. But it's kind of hard to see the level considering the position of the vase.

Checked it again in the next morning, and the car was slightly tilted on the passenger side, but not enough to give me an approximate reading in case it was critical. And I saw the level of water seemed way above the "max" mark. I don't know if it's supposed to be like that when the engine is cold, I always struggled to know if those min and max markings are supposed to be meaningful when the engine is hot or cold, because when the circuit is open and so is the thermostat, is open, I suppose the level would drop quite a bit, so I guess it could be my first question here:
1. How to read the level exactly, in what condition (hot or cold engine), what to check?
2. What about the radiator itself, considering the cap is on the latter, should the radiator ne filled up fully and the surplus of water then drops into the vase and has to be on the max marking?


So those were rookies questions, but I'm more familiar with VAG cars, it's new for me to put my hand in a Japanese car, I apologize for it.

As for the overheating, it wasn't over. Yesterday, did about 15 km, no problem, big slope and all, made it where I wanted to go. Turn the engine off, 5 mins after I drive back home, on a rather flat road, overheat light starts to flash again for about 5 or 10 secs, then disappear, then it does it again, and it was like that all the way home, even at 90 km/h, where the wind is strong. Weirdly, seems like it does it less when I'm stationnary, so maybe it has something to do with the engine rev.

At some point I stopped for about 10 minutes, when I started the car again, fan kicked on (oh yea forgot to mention, the fan does kick on, seems to work properly at full speed, and so does the heater inside the car) and immediately the heat light started to flash again. Went home and didn't use the car until this morning.

Took the kids to school, short trip, not particularely high rev, no flashing light. However this time I had my ELM 327 plugged in, connected to my phone with some OBDII app allowing me to monitor the car informations, such as the water temp.

Monitored it all through the trip this morning, seemed to heat up quite fast, in about 5 minutes I was at 84°C and then it went up and down, sometimes nearing the 100°C mark, but weirdly going back to like 84/85°C in a matter of second. No overheat light came up at all.

When I arrived home, I let the car engine on, and started reving it gently at about constant 3000 rpms and naturally the temp rose slowly but surely, at about 94/95°C the thermostat must have opened because I heard the fan kicking, got out immediately and checked if it was leaking on the ground, if the fan was spinning quick, if the airflow was in the good direction, even touched the pipes (hot), and didn't see anything in particular...

I went back in the car and waited, and saw the temp keeping on rising while the fan was still running, it went at about 105°C while the fan was turning and I decided to stop the car because I was afraid to damage the engine. Still has not seen the overheating light in the process though, but I'm sure if I kept waiting, eventually it could have reached the temp where the light decides to flash. Don't know at what temp it is though; above 105°C I suppose.

Checked the oil level, fine, no trace of headgasket blown either, considering the car doesn't smoke, and there isn't any white mayo on the caps yet.

Put my hand on the first radiator you can feel when you reach below the front bumper, thinking it would be my water radiator, and found it cold, so I was excited because I thought it was a thermostat problem and could fix it then, but then I thought about it and it might have been the AC radiator which I touched, I don't know. I think there's a smaller one behind it which I touched too and it was not cold. So thermostat opens I guess.

So now I don't really know what to do to diagnose what is wrong with the car. Could buy a new thermostat, but there are several available on Autodoc, with different opening temp and I don't know which one is the good one (My thermostat OEM is 19301PZA305). Could also change the temp sensor while I'm at it, and even the radiator switch despite it seemingly working or at least kicking it. Those 3 parts would cost me a few, and thought I could maybe grab the opportunity to like flush the radiator good since I will drain the water to change the parts.

However I don't know how to bleed the circuit properly once everything is put in place.

I don't know what to do guys, what's your opinion on this one? I'll try to answer any question and will take any suggestion. Thanks

P.S Apologies for english, tried my best to be clear, but my mind is racing, and unfortunately it's racing in my native tongue. Feel free to tell me if some part seems confusing.

[CRC]

I wouldn't panic too much if I was you ........

I've never known a thermostat that failed in the closed position. Almost every car ( and certainly the Jazz) uses a wax type thermostat and when they do fail, you'll find that the engine never gets up to temperature properly because the thermostat is always open and the water is always going to the radiator.

The wiring diagram shows a fan temperature switch which closes at 93C to cause the fan to come on, and apparently it's mounted at the transmission end of the engine, but it sounds like yours is ok because the fan comes on.

So effectively, the engine cooling is controlled by "old school" methods - a wax thermostat and a digital switch.

The ECU needs to know the water temperature to determine how much fuel to inject and the ignition timing, so it has an analogue temperature sensor which tells it the actual temperature in degrees C, and the ECU also controls the "low temperature - blue" and "high temperature - red" LEDS, with neither being turned on when it is a normal temperature range. This temperature sensor is the one that you are reading with your OBD II app.

The general feeling is that the Jazz radiator is easily able to cope with the excess heat from the engine, and as long as the fan(s) run you shouldn't have a problem.

I think I would be investigating how sensible the readings from the analogue temperature sensor are  ..... which you can do with your OBDII tool. Also looking for potential air locks perhaps?

[Dwight]

Hello, and thanks for your answer.

I am not really panicking as I know my way arround cars, and I changed quite some thermostats before, mostly on VAG engine where, as you said, they were done for and always on open position, preventing the car engine from heating up, which wasn't really good either for a diesel engine (1.9 TDI mostly), it was a "known issue".

What bothers me here, is that everything seems to work properly, as the thermostat opening, the fan kicking, the amount of coolant in the tank, but yet it stills overheated.

And when I say overheated, it's not a small occurence when the car was in a heavy load in a steep or anything, as I said, the overheat light came on at least 20 times during a 30 minutes ride last time, including while cruising at 90 km/h, which shouldn't happen at all.

What worries me is that for now the weather is quite smooth, but I live somewhere where in a about a couple weeks it's going to be between 35-40°C everyday until fall, and I'm not sure it's going to improve the overheat, I'm pretty sure it's going to be quite impossible to take the car for a long ride, and that is not natural.

Last summer I had it fixed already and didn't encounter this problem at all, so something as changed and something is not working right.


Today I took it as every morning for an errand in downtown, so it's a 20 minutes journer in and out of city back home. I again plugged the OBD to see how it behaved, and it seemed alright, except for some random coolant spiking at 100°C then droping to 85 without really hearing the fan kicking in.

So when I got back, I did the same as last time and let the car iddle, temp seemed normal nor going up too fast, so I put my feet on the gaz and let it iddle at arround 2000 rpm. The temp rose as it should, but I noticed some odd things,

1. Weird "whirling water" noise coming from behind the dashboard, so I suppose from the heater circuitt, not the first time I hear it
2. The fan kicked in at a bout 98°C
3. Despite the fan kicking in the water temp kept rising at arround 105°C and suddenly the fan stopped for about 20 seconds, then kicked in again and the temp in the meanwhile kept reaching about 110°C and the fan action did not seem to have an effect on cooling it down at all.


I'm starting to have the feeling there's something wrong with the radiator and fan "action" on the coolant, as if it's not working as it should, and that would explain the overheat.

What do you think?

[Marco1979]

You would probably need an infrared camera to see temperatures of the radiator and the tubes connecting it. It might just be an air bubble or a blockage inside the water circuit, as all sensors and mechanics seem okay.
Good luck!

[Dwight]

Before puting any serious money into the car and changing parts which seems to be working fine for now, I'd like to flush the coolant circuit and put some fresh one.

My problem is how should I proceed to really get the whole circuit drained?

Also, I have absolutely no idea how to bleed it. Saw a video on Youtube of a guy bleeding it by letting it run with the radiator cap open and toping it up, but I really don't know if it's that simple.

Any well documented procedure would be really appreciated 

[CRC]

But you're assuming that the temperature sensor that you're reading is correct...... if it's giving spurious readings then your high temperature light flashing on and off is explained as the ECU reads the temperature and decides when to light the light.
The rapid temperature changes point to a dodgy temperature probe, as it's hard to cool and heat water that quickly.

At least check the probe first, as it seems to have a different opinion to the temperature switch that's controlling the fan.

[Dwight]

Yes, you are right. I understand what you mean now.

But how exactly should I test the probe? I don't really see how I can do that beside putting a new one in, which isn't that expensive.

But don't I have to drain the coolant circuit to install the new one anyway? Or is it the kind you have to put in quickly before the coolant drains and top up what you lost in the process?

[CRC]

The Haynes workshop manual says that the probe is at the rear of the cylinder head, but doesn't show a picture unfortunately.

It's almost certainly a probe that changes resistance with temperature and will almost certainly have a plug and socket to connect the wiring.

The first thing to do is pull off the connector and clean it with contact cleaner, then put it on and off a few times to clean up the contacts.

A small amount of corrosion here can lead to altered resistance values that will give odd readings.

Because it's at the top of the cooling system, it's unlikely that you would lose a lot of coolant when you change it if it needs changing.

The internet may have resistance values for the probe at various temperatures, but there's nothing in the Haynes manual unfortunately.

[Dwight]

I've got you for the picture, or more like the exploded view

Looks like from the parts descriptions not shown on the screen that the water temp sensor would be the number 13, and the OEM number is 37870PLC004. Quite tricky to access, but I think I can do it. Doesn't look worse than on a bloody VAG engine.

Problem is, what should I do, because I have no idea how to test it. I have a multimeter but that's about as far as my eletrical knowledges goes. I can work arround 12v cables and rewires or test and stuff, but I don't have the genius to understand how this probe would work and how to properly test it unfortunately.

I have the feeling it the only thing I can do right now is order a new one on Amazon and see with my OBD reading if it behaves more natural and thus prevent the light from coming on again.

However I'm still kind of worried about that bubbly water noise coming from behind the dashboard, I have the feeling it's proof there are bubble in the coolant circuit anyway, which isn't good.

[CRC]

I think that should link to the way to test the probe.

I noticed that Febi Bilstein make a replacement that sells for less than £10 and I think I'd just buy one and replace it for peace of mind. Febi equipment is very good in my experience.

However, in the meantime, this is what I'd do first .....

When the engine is cold, remove the radiator filler cap and check that the radiator was full to the brim.
Turn off the air conditioning ( the refrigerant liquid can gurgle as it enters the evaporator) and you just want to listen for water noise.

Start the engine with the radiator cap off and look for any air bubbles coming out of the liquid. Squeeze the top hose while still cold.

The thermostat should open at around 86C and the top hose should now get hot. See if air bubbles are coming out now.

The water may now start to overflow from the radiator as it has now expanded a bit, so either stop the engine, or carefully put the cap on.

Monitor the temperature with your obd scanner while doing this.

Check that the fan kicks in at around 94C (because the air conditioning is off, the fan will only be controlled by the water temperature switch)

If you can find a surface temperature probe it can be useful, though it will always read a bit less less on the outside of a tube than the liquid inside

See what you find ...

[Dwight]

Wait, is the water supposed to reach the top of the radiator when you undo the cap? Because as far as I remember, when I did unscrew the cap, the water did not seem nowhere to the top, I only checked the overflow tank and saw that the coolant was way above the "max" line which I found strange, but the car was a bit tilted on the passenger side.


Let's say you remove your coolant both from the radiator and the overflow tank, put some fresh one, I was wondering how the filling up works, because in my mind only the overflow tank shows level lines, but I was guessing there's some kind of way in the radiator hole that you have to fill it up until it reaches a certain level and the surplus then goes in the overflow tank, and you have to keep filling it up until the surplus reaches the max line in the overflow tank, and now you're sure you have your radiator topped up, and your overflow tank filled to the mark too.

I thought about it, but wasn't sure until you mention that, and I thought about the possibility, that the radiator may not have enough water, since I didn't saw the coolant anywhere near the top when I removed the cap, thus I suppose would explain maybe some overheat.

Would be quite funny that the cause of my problem is such a stupid thing, me being unaware that the radiator has to be filled to the top, and not only the overflow tank up the max line. Waiting for you answer on that one, but somehow it could make sense.

[CRC]

Some cars only have a coolant reservoir which is how you have to top up, but the Jazz has a radiator cap.

Mine probably needs replacing as it doesn't really want to undo, but in the past, when I have opened it, the water was always right up to the brim.

The radiator cap works as a pressure relief valve, so as the radiator pressure rises due to temperature expansion of the water, it opens up to allow water into the expansion tank. Presumably the radiator cap also allows flow the other way as the radiator cools.

Definitely top the radiator up to the top, then run the engine as I described and see if air comes out.

[Dwight]

Thanks for the info, I'll keep you posted

[CRC]

Also, the Hayne's manual states that the thermostat should open at between 80 and 84C and that the radiator cap relief valve opens at between 14 to 18 psi, and as 1 bar is 14.7 psi, that seems quite high.

It also means that water at that pressure won't boil at 100 C, but opening the top when hot could release water hotter than 100 C which would partially flash off as steam, so be careful.

[Dwight]

Appreciate the advices. Unfortunately learned the hard way that despite being as precautious and slow as you can when unscrewing a pressurized cap linked to some hot coolant circuit, there's always a chance the pressure blows off at once and that your face is sprayed with very hot water.

What can I say, we were all young and stupid at some point, and this kind of things motivates you to understand how an engine works, and what are the hazards when working on it.

So yea, I'll make sure to uncap it when it's cold...

Cheers