Overheating or dodgy sensor?

[m4rkw]

Hello friends

Seasons greetings!

On our way to visit some friends for christmas I saw the temperate gauge high again, this time though there’s no loss of coolant. I kept an eye on it and it seemed to not be too high but then later it got near the red so I pulled over. Both fans are working, no loss of power and no milky residue under the oil cap. My suspicion is a dodgy thermostat which would mean it’s ok to keep going, is there any way I can determine this?

Have called the RAC anyway

Cheers,

Mark

 

[steve2006]

The only real check is to remove the stat and then lower it into a clear Pyrex jug of freshly boiled water then see if it opens. The rubber seals tend to perish so a replacement would be wise after 15 years.

[m4rkw]

Cheers Steve, I'm hoping the RAC can confirm one way or the other whether the block is actually overheating and then I'll know if it's safe to drive it home or not. Don't really want to chance driving it with the needle on the red. That's if they manage to turn up of course, not really the best day to need a callout! If they don't I guess I'll just have to limp it home and keep an eye on the gauge and stop if it gets near the red again.

[m4rkw]

It’s been sat for two hours and the block is cold, temp gauge still points at red. Has to be a bad thermostat 

[m4rkw]

I’m assuming it’s not dangerous to drive it in this state, might lean the mixture out a bit but hopefully not cause any damage 

[m4rkw]

It's definitely the coolant temperature sensor, after 3 hours sat to cool down the gauge was still pinned to the red. I drove it a short distance and it dropped back to the bottom and the check engine light came on, so presumably that was the sensor dying.

How hard is it to replace? might have a go myself if it's simple

[m4rkw]

Hmm toylexparts.com appears to be dead, anyone know of an alternative?

[m4rkw]

Right, I've found the part number for the coolant temp sensor and the crush washer and located it on the engine block. Am 50/50 between getting the part and doing it myself or booking it in with the garage. It looks easy enough to replace, I'm just not sure how much torque to apply to the bolt, worried about doing it up too tight. Would be handy to get it fixed tomorrow because mum is coming to stay with us, I doubt the garage will have time to look at it this week.

[Herbie]

In the absence of a proper torque setting I'd go hand tight plus a quarter turn with the spanner.

[m4rkw]

I've watched a few videos and some of them make this look really simple i.e. just do it quickly and you're fine, others suggest draining the coolant out and then bleeding the system when refilling it which seems a lot more complicated. Not sure.

[m4rkw]

Success, toyota was still closed today so I got a sensor from euro and swapped it over quickly. A little bit of coolant leaked out but not much, car now runs up to temperature normally and the check engine light is cleared. I checked it carefully for leaks as it came up to temperature and then took it for a drive, seems perfect. Not a bad result for £20 👍

[m4rkw]

I'm slightly paranoid about the torque on the sensor, doing it up with the rachet it felt loose until the last bit and then less than 1/4 of a turn until it was really tight. I went and checked it again just now and it didn't take much movement at all to become free so I just did it up again about the same. It's not leaking so hopefully tight enough but not too tight. I always over think stuff like this

[Herbie]

5 hours ago, m4rkw said:

I checked it carefully for leaks as it came up to temperature and then took it for a drive, seems perfect. Not a bad result for £20

I know - gives you a nice warm fuzzy feeling when something like this works doesn't it?

Our Beko washer/dryer recently died; no lights, no actions, no sounds except for a very, very, faint bleeping.

A bit of Googling later and I found it was 99% sure to be the printed circuit board, with new ones costing £80. A bit more Googling revealed that it was most likely a diode that had gone for a poo. The photo below is my actual board with the diode arrowed.

I live about a mile and a half from https://cpc.farnell.com/ so off I pootled in search of a replacement. Turns out they don't sell them singly so I had to buy a pack of 10 for the huge sum of 40pence including VAT!

Once home, armed with the desoldering station and the soldering iron, out came the board and out came the faulty diode. Replacement in, and the end result was a fully working washer/dryer for just 4p and a bit of effort 

[m4rkw]

Nice 🙂  I have ordered a low-range torque wrench from the bezos empire just to set my mind at ease with the coolant sensor, I'm pretty sure it's on correctly but always best to be sure.

[m4rkw]

How far over torque spec can you go before causing damage? Still waiting on a torque wrench, Ive noticed after driving there’s a tiny bit of water on the underside of the nut on the sensor but it’s very slight and could easily be rainwater. No idea if i’m over or under the 20Nm spec 

[m4rkw]

The coolant level isn’t visibility going down and there’s no wet patch under the sensor at least 

[m4rkw]

Ok so I confirmed after a long drive that there was a very slight coolant leak (dabbed tissue turned red), got the torque wrench and set it to 20Nm and tried to turn it but there was very little room to move the handle, I managed to move it a bit but the amount of force felt disconcerting and the torque wrench wasn’t clicking, then because of the limited space I managed to accidentally pull on the connector with the socket and pulled it slightly to one side. 

I didn’t try turning it anymore because it felt very tight and I was nervous about the torque wrench being off because 20Nm is it’s lowest setting. Drove to euro to get another sensor and checked it again and found that after the tightening it was no longer leaking. So my plan now is to get a second sensor and keep it around just in case but not change it now and go through all that hassle again, the connector moves slightly more than it should in one direction but it works perfectly and the connector is pretty snug, I don’t think there’s any chance of it breaking off. 

Alls well that ends well I guess! Moral of this story is probably “if you’re going to wrench, buy decent tools”. I still don’t really know if my torque setting is over or under but next time it goes in for a service I’ll ask the garage to check it.

[m4rkw]

Checked again in the morning, no leaks, ran it around a bit up to temperature, still no leaks, I’m calling this a win. 👍

[m4rkw]

Another thing that's just occurred to me, when the problem initially started and I'd left it parked up for 3 hours before looking at it again the temp gauge was pinned to hot. It was very hard to start because thinking it was hot it would have created a lean condition which could possibly have caused damage if it'd stayed like that for long. Shortly after it started the sensor failed completely (either that or the PCM decided not to trust it) and the CEL came on and then it defaulted to assuming cold. I now realise what I should have done when it was pinned to hot was to disconnect the sensor then it would have immediately assumed cold and run rich rather than lean. Live and learn etc.

[m4rkw]

Interestingly though despite it being very difficult to start when it was assuming hot, once it was running it seemed fine, no power loss, ran smooth. I wonder if it somehow automatically compensated after realising the ratio was too lean?

[Delphius1]

Mark, you're probably spot on. It started with the lean mixture as per the temperature readings. But as the Oxygen sensor reports the Lambda readings back to the ECU, it will trim fuel levels to compensate. If the short term fuel trim goes over a certain threshold the engine management light is lit reporting a problem. 

It's one of those issues that is a bit of a black art to diagnose, especially if the ECU temperature sensor is separate from the gauge sensor. You don't see what the temp sensor is reading, so you have to dive a bit deeper. Fuel trims can tell you part of the story, then looking at live data from the sensors against measurements with (in this case) a contactless temperature gun should show the sensor isn't reading correctly.

[m4rkw]

Sounds like I got lucky with my diagnosis then!