Car Overheating Problem

GayanR · May 8, 2014

Dear all last week my mazda 3 car came up with a strange overheating problem...

This is what happen...sorry for the long story

Last week i went to this garage to get my radiator expansion tank replaced due to a small leak,it was leaking near one of the upper small hose that connecting to the expansion tank...so after removing the hose found that the expansion tank plastic line was corroded a bit and cracked...and they also found some plastic pieces like sand inside the hose they cleaned it...the lost plastic part was around 2mm..they replaced the tank with a oem one..

I came home with no problems...then the other day i went on a trip around 200km..anyway i kept looking at the gauge the whole time because of the new tank.had no issues but on the way back when i was near my hometown i noticed my fans were noisy than normal then i checked the gauge it was around 3/4 of the gauge...I suddenly turned off the car and let it cool and opened the bonnet..I didn't feel that much abnormal physical heat and there were no leaks..so again started the car it was back to normal temp then i drove home without any issues....then the other day i went to the garage(gauge was normal) they couldn't identify the problem anyway they replaced the thermostat and bleed the system for any air...they said they checked for 2 hours it didn't heat up...

again when i was coming home the gauge gradually began rising than normal so i revved the engine a bit hard.. to my surprise gauge came back to normal...then i drove around 2 hours without any issue...the issue came again today then i revved the engine and it was fine after that....

Fans working fine,they kick in and working,heater works fine...I have this anxiety could be this due to that small plastic chip getting clogged somewhere randomly...how can i sort this out...will a coolant flush work?tanx

Davy · May 8, 2014

This could be due to a failing water pump. It might not be pumping properly whereas it could pump when you rev hard. Could the "sand" like stuff be corroded blades of the water pump?

On a side note, I hope the correct thermostat was used when it was replaced. The thermostat is temperature specific and using one with a higher temperature rating might cause the car to overheat. If the replacement thermostat had a matching part number, then you do not have to worry about this.

What was the coolant being used so far?

GayanR · May 8, 2014

On 5/8/2014 at 12:27 PM, Davy said:
This could be due to a failing water pump. It might not be pumping properly whereas it could pump when you rev hard. Could the "sand" like stuff be corroded blades of the water pump?
On a side note, I hope the correct thermostat was used when it was replaced. The thermostat is temperature specific and using one with a higher temperature rating might cause the car to overheat. If the replacement thermostat had a matching part number, then you do not have to worry about this.

What was the coolant being used so far?

@davy Garage replaced the thermostat with a genuine mazda one...this whole week the heating issue has been very random,i drove it day before yesterday without any issue temp gauge was right at what it was used to be all the time....but today morning the issue came again,gauge started climbing and i revved it to about 6000rpm again came back to normal ..after that drove about one hour in traffic without an issue....only when you said i start to doubt about the sand like plastic stuff,but the stuff was near the corroded line hose of the overflow tank....I wonder why it can last normal after hours i rev the engine....there is no leaks or water pump noises...can a water pump randomly fail like this

Edited May 8, 2014 by GayanR

Rumesh88 · May 8, 2014

Let's look at your observations above from a different view point assuming you do not observe any coolant loss or air bubbles in the radiator. Now when you say the indicator goes from half-mark to 3/4 mark does it happen within 30 seconds or less? And then when you rev up does the gauge revert back to half-mark within a similar time period? If so the problem may be in the gauge circuit and what you get may be a false indication. In this case it would be difficult to trace the actual issue for the fault is intermittent. But it is worth checking the integrity of the engine block grounding strap. Just check if the connection is firm or not, whether it is corroded etc..Ideal test is to check if there a appreciable voltage drop (like 0.5V) between the engine block and the battery -ve terminal when the problem occurs. Keep in mind however, this is only one of the many possibilities.

GayanR · May 8, 2014

On 5/8/2014 at 2:31 PM, Rumesh88 said:
Let's look at your observations above from a different view point assuming you do not observe any coolant loss or air bubbles in the radiator. Now when you say the indicator goes from half-mark to 3/4 mark does it happen within 30 seconds or less? And then when you rev up does the gauge revert back to half-mark within a similar time period? If so the problem may be in the gauge circuit and what you get may be a false indication. In this case it would be difficult to trace the actual issue for the fault is intermittent. But it is worth checking the integrity of the engine block grounding strap. Just check if the connection is firm or not, whether it is corroded etc..Ideal test is to check if there a appreciable voltage drop (like 0.5V) between the engine block and the battery -ve terminal when the problem occurs. Keep in mind however, this is only one of the many possibilities.

@rumesh there is no coolant loss or air bubbles in the radiator....but i noticed that inside the expansion tank.there are some water condensation like water drops...with the old one i didn't see something like that....um afraid the coolant is actually boiling..once the car up to operating temperature i see coolant shoot into the expansion tank like they are bubbling,not a steady flow like once you start when the engine is cool...but yes the heat up process happen within seconds...heat goes up gradually within 15 seconds then fans runs fast then when i revved pass 5000 rpm it comes back to normal within seconds then after that can drive for hours without an issue...but lately i didn't let it go over 3/4 of the gauge...

Rumesh88 · May 8, 2014

Perhaps you may have to inspect your water pump as Davy pointed out although the chances a pump impeller intermittently disengaging is very slim. Still checking the grounding connections at least visually wouldn't cost you anything.

Davy · May 10, 2014

On 5/8/2014 at 12:46 PM, GayanR said:
@davy Garage replaced the thermostat with a genuine mazda one...this whole week the heating issue has been very random,i drove it day before yesterday without any issue temp gauge was right at what it was used to be all the time....but today morning the issue came again,gauge started climbing and i revved it to about 6000rpm again came back to normal ..after that drove about one hour in traffic without an issue....only when you said i start to doubt about the sand like plastic stuff,but the stuff was near the corroded line hose of the overflow tank....I wonder why it can last normal after hours i rev the engine....there is no leaks or water pump noises...can a water pump randomly fail like this

To be honest, water pumps don't randomly fail as most of them (like yours) are belt driven. So as long as the engine is spinning, the water pump keeps pumping. From what I gathered, your car's water pump is driven by the timing belt. Removing it for inspection is quite a bit of work.

So may I suggest that you get your cooling system cleaned up before you move ahead. The grime in the system might be clogging one or more lines in the radiator core that end up hindering the cooling efficiency. A good radiator repair shop will do a thorough clean up by first removing the radiator and detaching the top and bottom tanks from the core and then by cleaning each core line to make sure that they are not blocked. Following the clean-up, a flush could be done to ensure that the rest of the circuit is also okay.

Since you say that the issue did not resurface again, it actually might be more air trapped inside the system which has now been automatically purged.

GayanR · May 10, 2014

On 5/8/2014 at 5:03 PM, Rumesh88 said:
Perhaps you may have to inspect your water pump as Davy pointed out although the chances a pump impeller intermittently disengaging is very slim. Still checking the grounding connections at least visually wouldn't cost you anything.

HI rumesh does ground connections means the wires that connect to the engine frame with a bolt?

GayanR · May 10, 2014

On 5/10/2014 at 2:24 AM, Davy said:
To be honest, water pumps don't randomly fail as most of them (like yours) are belt driven. So as long as the engine is spinning, the water pump keeps pumping. From what I gathered, your car's water pump is driven by the timing belt. Removing it for inspection is quite a bit of work.
So may I suggest that you get your cooling system cleaned up before you move ahead. The grime in the system might be clogging one or more lines in the radiator core that end up hindering the cooling efficiency. A good radiator repair shop will do a thorough clean up by first removing the radiator and detaching the top and bottom tanks from the core and then by cleaning each core line to make sure that they are not blocked. Following the clean-up, a flush could be done to ensure that the rest of the circuit is also okay.

Since you say that the issue did not resurface again, it actually might be more air trapped inside the system which has now been automatically purged.

@davy I'll do a flush soon and see..car has a timing chain..going through the net it seems mazda 3 has a external pump,I'll do a full flush and see next...tanx for the info

Rumesh88 · May 10, 2014

On 5/10/2014 at 5:24 AM, GayanR said:
HI rumesh does ground connections means the wires that connect to the engine frame with a bolt?

You have several of these between the chassis and the engine and also a few coming from the ECU. All the wire terminals directly ending up on the engine body are grounding connections and they are easily visible.

MkX · May 12, 2014

My first suggestion is to do a complete flush in the coolant system. Sand like residue you mentioned could be minerals gathered inside the water lines. Did you use the water for prolong period instead coolant? I've seens vehicles that used water for long period without much problem.

Other suggestion is to check the condition of the pressure cap. If the spring clogged over the period there will be a delay in coolant coming back to radiator after pressure decrease. During this period you can see sudden increase in temperature. Once coolant start to move from expanstion tank to radiator temperature will come down.

GayanR · May 13, 2014

Update:I did a radiator flush ran 2days fine after that....today the issue came again....I noticed that following is what is exactly happening...

car won't overheat when running,not in traffic but if come to a complete stop after a run and let it idle about 10 min the temperature gauge rises( not all the time)....I was aware this time so i was looking at the gauge every second, the gauge didn't even take a sec to climb to 3/4 of the gauge(needle was exactly where the first time this issue occurred 3/4 of the guage)...so i revved the car again gauge needle came back to exact operating temperature within about a second,at the time needle was climbing i could hear the fan running loud so the fan works...I turned the heater on when temperature was high it blowed hot air so the water pump is working....as rumesh88 said i checked the ground cables and all seems fine.....after that i idle about another 20min without an issue....as i said earlier once the issue got cleared i can idle for hours...

Could be this due to a bad relay somewhere responsible for the gauge or something?I cant find a competent mechanic to sort this out...

Note: previously car had 2 bad relays,the a/c was working time to time,horn was working randomly,all were cured by replacing the relays....

Edited May 13, 2014 by GayanR

Davy · May 13, 2014

Not confirming tha the water pump is at fault here but the heater blowing hot air is actually not an accurate way of determining if the water pump is working. When the engine is hot, the coolant in the entire system is pretty warm, and the heater core will definitely get heated up and as a result, hot air will be thrown out of the vents.

If you think the gauge is showing an incorrect reading, then the coolant temperature sensor might be at fault here. A good mech can get this tested for you.

Rumesh88 · May 13, 2014

The way your cooling system behave (rise and fall of gauge indication within a couple of seconds - I already asked this) your problem must be in the electrical system. It may be the coolant temp sensor as Davy says but the very intermittent nature of the problem makes it really hard to trace to a single component. When the gauge reach 3/4 mark stop the car, just switch off the engine without revving up and switch it back on without starting. Now see if the reading is still at 3/4 mark. If so, then get hold of the wire harness in the engine bay, particularly close to ECT sensor, and shake it and see if the gauge reverts back (well, you need someone to keep eye on the gauge). Look for wires in the engine bay close to metal parts. You may have an intermittent short circuit. There is a way to check the ETC sensor/gauge circuitry but in this case your auto-electrician would have to travel with you until the problem appears.

GayanR · May 13, 2014

Tanx davy and rumesh for the info...yes this is what making me mad it only happened like 3 times this week but once it happened after that no problem through out the whole day again even if i idle for a long time...I'll check with a electrician to find this out.....bdw do i have to remove the battery cable before unplugging and plugging any sensor plugs?

Rumesh88 · May 13, 2014

On 5/13/2014 at 10:38 AM, GayanR said:
Tanx davy and rumesh for the info...yes this is what making me mad it only happened like 3 times this week but once it happened after that no problem through out the whole day again even if i idle for a long time...I'll check with a electrician to find this out.....bdw do i have to remove the battery cable before unplugging and plugging any sensor plugs?

You need to exercise lot of patience until you sort out the problem. Repair manuals would advise you to remove battery negative before you replace or disconnect sensors, but you can work safely with sensors with ignition off as long as you do not use things like test lamps to check continuity between terminals.

H.C · May 13, 2014

One of my friends had similar issue in his swift and it turned out to be fulty coolent temperature sensor.

@Rumesh, can you use an IR thermometer to check whether there's actual temperature elevation? .

Rumesh88 · May 13, 2014

On 5/13/2014 at 3:34 PM, H.C said:
One of my friends had similar issue in his swift and it turned out to be fulty coolent temperature sensor.
@Rumesh, can you use an IR thermometer to check whether there's actual temperature elevation? .

Yes you can use a IR thermometer but the reading may not be that accurate (I mean the absolute accuracy of coolant temp) for you have to aim it at an external part like the engine block or the radiator tank top but not directly at the the coolant. But you can get a general idea and decide on the ECT sensor. For example in this case you can measure the temperature of radiator tank top when the gauge is at 1/2 mark and compare the reading with the temperature you read when it is at 3/4 mark.

Edited May 13, 2014 by Rumesh88

GayanR · May 24, 2014

Update:Guys it turned upto be a bad water pump....still wonders me how quickly the gauge could go up and come down so fast and how rarely and randomly the overheat issue occurred.hope this will be a help for others...

NRX · May 25, 2014

On 5/24/2014 at 1:31 PM, GayanR said:
Update:Guys it turned upto be a bad water pump....still wonders me how quickly the gauge could go up and come down so fast and how rarely and randomly the overheat issue occurred.hope this will be a help for others...

What was the total damage

GayanR · May 25, 2014

On 5/25/2014 at 3:33 AM, NRX said:
What was the total damage

water pump was 15000..repaired it at m**na motors..Total bill was 16000/=..There were some others brand water pumps for around 6000..He said he cannot guarantee the lifetime of those compared to the genuine.

Davy · May 25, 2014

Glad you managed to sort it out. They've only charged 1000 labour which is very reasonable I guess. +1 on opting for a genuine pump as well.

As for the temp gauge going up so quickly, well it gets it's reading from the Coolant Temperature Sensor (CTS), and when the pump is not working properly, the coolant from the radiator does not come back to the engine to bring the overall temperature down. But when you race the engine, the pump should probably be bringing back some amount of coolant to bring the temp gauge back to normal.

Also, the range from the mid point to the max of the temp gauge is actually the difference between a few Celsius. So even a couple of degrees is enough to show a visible change in the gauge.

Edit: Corrected typo

Edited May 25, 2014 by Davy

GayanR · May 25, 2014

On 5/25/2014 at 4:37 AM, Davy said:
Glad you managed to sort it out. They've only charged 1000 labour which is very reasonable I guess. +1 on opting for a genuine pump as well.
As for the temp gauge going up so quickly, well it gets it's reading from the Coolant Temperature Sensor (CEL), and when the pump is not working properly, the coolant from the radiator does not come back to the engine to bring the overall temperature down. But when you race the engine, the pump should probably be bringing back some amount of coolant to bring the temp gauge back to normal.

Also, the range from the mid point to the max of the temp gauge is actually the difference between a few Celsius. So even a couple of degrees is enough to show a visible change in the gauge.

Tanx for your help.Anyway hope this incident didn't do any damage to the engine.What they said was pump was seizing up slipping randomly when rotating.But from what they have experienced is pumps seizing up at high speed most of the time..They were looking at the electric fan both the time I went there.This time I asked them to remove and check the pump as well.

Edited May 25, 2014 by GayanR

vishkid · May 25, 2014

Gayan, were your previous repairs (incl. the coolant reservoir) also done there? They helped me to correctly diagnose a water pump issue on my Axela which had been misdiagnosed by someone else. I was late to spot this post, otherwise would hv commented on my experience before which was a bit similar to yours.

Plus, my fan control unit had a few capacitors blown which had reduced my fan speed from stock and that made my issue worse - replacement of water pump and FCU solved mine.

It was a pain to check all the elements of the cooling system!

GayanR · May 27, 2014

On 5/25/2014 at 5:54 PM, vishkid said:
Gayan, were your previous repairs (incl. the coolant reservoir) also done there? They helped me to correctly diagnose a water pump issue on my Axela which had been misdiagnosed by someone else. I was late to spot this post, otherwise would hv commented on my experience before which was a bit similar to yours.
Plus, my fan control unit had a few capacitors blown which had reduced my fan speed from stock and that made my issue worse - replacement of water pump and FCU solved mine.

It was a pain to check all the elements of the cooling system!

@vishkid Tanx for your help,Would have diagnosed it earlier with your experience.Yes I did the tank replacement there as well.The guy there said that the fan has 3 speed settings and sometimes fail to operate at the correct speed and hard to diagnose,May be he was commenting with your car experience.He first wanted to put in a another fan and try it for few days to check if the symptoms arise again.I was scared of risking another full blown overheat again,so asked them to check the pump as well.It was a pain, would have been easier if things just fail without failing randomly.