Replacing Thermostat

[thetyrant]

Ok due to my weird cooling issues (see airlock thread) im going to replace the thermostat to either fix the problem or at least rule it out, ive picked up a new mahle stat from motor factor this morning so hopefully get it swapped over tonight.

It looks pretty straightforward from what ive seen so far and not havin AC pump in the way on my car is a bonus it seems, ive had a search for diy/tips on the job etc but not found much other than this video which didnt really show much i didnt know but handy for others maybe - https://www.youtube.com/watch?v=nZvyRuYw7C8

Im hoping i dont need to drain all the coolant out and just clamping hose and leaving cap on res should prevent too much escaping, no doubt i will get wet though lol  if anyone has done it and has tips let me know

Wish me luck

[Carolyn]

Use the drain on the engine.  You'll still get a bit wet.  The thermostat has a small hole in it, which goes to the top.  Very important!!

I still reckon the turbo plumbing is at the heart of it.

[thetyrant]

Use the drain on the engine.  You'll still get a bit wet.  The thermostat has a small hole in it, which goes to the top.  Very important!!

I still reckon the turbo plumbing is at the heart of it.

Ok thanks is that little tap on back of block under manifold  ?

I appreciate what your saying about Turbo but that is not in the main coolant route via radiator, its tee'd into the heater return circuit and all that is working fine after bleeding, also the coolant temp has always been mid 90's long before the turbo went on so i dont think its the turbo myself, however i can easily take the turbo out of the cooling loop if needed for testing and to be honest at the moment im thinking about putting car totally back to stock!....the tried and tested way of less hassle ...then i feel the boost though and want that lol

Either way i need to get coolant circulating around the system first, hopefully the stat is the issue or at least taking it out might release a big air bubble at stat end which is my other possible thought....or the pump not working as well as it should....we can look at those things next and stat is easiest thing to rule out and its 103k miles and 15 years old so might as well.

[Carolyn]

That's my entire point.  The heater return circuit allows the thermostat to open!!!!!!

If the return circuit is not spot on - the thermo won't open.  Closed thermo means the return from the rad won't work.  Hence the rad not heating up properly. 

It delivers warm coolant to the back of the thermostat.

I think your plumbing is short circuiting the thermo in some way.

But returning the plumbing to bog standard is the best thing to do.  I'd do that before messing with the thermostat.

I bet you a hot buttered doughnut that it does the trick.

Yes the drain on the block under the manifold.

[thetyrant]

OK thanks yes i see and that makes sense, i can easily take turbo out the loop as further test if needed but its not doing anything other than extending the pipework coming back from heater and ive got hot coolant at both ends where it joins into system, in fact all the pipework under bonnet seems nice and hot from what i can remember but i will check it all again with thermometer gun.

Also it doesnt explain that i had same issue before the turbo install though with car running in 90's ? understandably i cant prove there was no flow out rad as i still had frunk on back then but temps were identical, i guess if i pull the stat and it works perfect in kettle water and have same issue with new stat i need to dig further, ive got the stat now and changing it isnt too much hassle so will do that and at least i will know its all good at that end, we can go from there if still no circulation.

So plan is change stat and bleed system, if still no joy remove turbo pipework from coolant system returning it all to stock coolant routing and see what happens there.....if that fails push it outside and throw a match on it

Wish me luck

[SV-3]

OK thanks yes i see and that makes sense, i can easily take turbo out the loop as further test if needed but its not doing anything other than extending the pipework coming back from heater and ive got hot coolant at both ends where it joins into system, in fact all the pipework under bonnet seems nice and hot from what i can remember but i will check it all again with thermometer gun.

Also it doesnt explain that i had same issue before the turbo install though with car running in 90's ? understandably i cant prove there was no flow out rad as i still had frunk on back then but temps were identical, i guess if i pull the stat and it works perfect in kettle water and have same issue with new stat i need to dig further, ive got the stat now and changing it isnt too much hassle so will do that and at least i will know its all good at that end, we can go from there if still no circulation.

So plan is change stat and bleed system, if still no joy remove turbo pipework from coolant system returning it all to stock coolant routing and see what happens there.....if that fails push it outside and throw a match on it

Wish me luck

@thetyrant Make sure you get some sleep

[thetyrant]

@thetyrant Make sure you get some sleep

Tell me about it lol...last few days ive been waking up at 3am thinking about next plan of attack and googling related issues

[thetyrant]

Ok new stat tested off car in pan of water and opens as it should so removed old stat ready to swap it over, thankfully didn't get too much coolant in my face, up my arm and over garage floor!.... fitted new stat refilled res bottle, attached bleed extensions, opened screws and did a little massaging to get a few bubbles out but its going to take awhile to bleed it all I think so leaving it for the night to hopefully let gravity do its thing.

Old stat was pretty grubby looking but did seem to work when tested in pan of water although slower than new one to start moving, maybe its ok but at least I know that part of the system is working so once its bled we will see what happens tomorrow!

[thetyrant]

Still bleeding!....taking some getting all air out no doubt as engine block was drained its got a few nooks and crannies to pump around, to be fair ive only been doing a little before and after work etc and will get a proper crack at it tomorrow see if i can get coolant circulating.

[thetyrant]

OK so i have the new thermostat in and system is fully bled now im pretty sure, however everything is just the same cars warms up as expected but coolant is still not circulating through the radiator and back into engine!  gets to 93-95c and just sits there!

All pipework i can access is nice and hot including the turbo coolant pipes that go back onto coolant bypass to feed the thermostat, only ones that are cold are rad exit pipe back to engine/ stat housing.

So i think Carolyn is onto something (hot buttered donut if you want one in that its something to do with coolant bypass pipework that feeds hot coolant to the stat to open it, i dont think its related to the turbo though as coolant in and out of that is as i would expect and problem was there before the turbo im sure,t i think its something further down the line preventing hot coolant back to the stat housing, if i had only got problem after fitting turbo i would of been more suspicious of it but it was there before im sure and pipes in and out are as exepcted temp wise, i will remove the coolant pipes to turbo to 100% rule it out though once it cools down.

Question is what else is in that pipe run?, its hard to see due to intake manifold and other wires/plumbing etc but from what i can see it starts at side of head unde PCV and then disappears around back of of head around under intake, there must be a pipe from the IAC heater circuit and pipes going down to oil/water cooler on filter housing, they seem to have hot coolant in from loosening the clips and squeezing pipes to get it to leak out into my facse! so think its making it that far im pretty sure, is the feed to the cooler inline or a take off ?  as in if i had a blockage there would coolant still get past to the stat housing ?

If anyone has a picture of the front of engine out of car especially under the intake to show how the coolant bypass goes that would be handy ?

[thetyrant]

Popping up to see Dick Sloan this afternoon to have a poke around an engine out of car so i can see whats going on with coolant bypass setup, hopefully it gives me some ideas and maybe Dick will have a thought on it as well

[Carolyn]

Do as you suggested and take the turbo plumbing out of the system.
There's a good chance it's stopping the last of the air behind the thermostat from coming back up the steel pipe.  What's the temp of the steel pipe going into the top of the thermostat on the engine?  I bet it's not right.

There's a bypass actually inside the head - but it's no use if there's air stuck behind the thermostat.

The oil-heater coolant circuit comes of that bypass and returns to the steel tube (on a euro car).  That circuit doesn't exist on a JDM car.

Either way - you've air-locked the thermo housing with your turbo plumbing.

But- hey - what would I know?? 

I'm female and only  had a dozen of these in bits....

[thetyrant]

All valid points and I am taking notice not ignoring you im just covering all bases   main thing that battering my head is same issue of running at 95c was there long before turbo went on and that has me stumped!

Ive been to Mr Sloans today and had a look at a engine fresh out of car (literally was just dropping it out as I got there!) so I could see the way piping is back to thermo, its as I thought but just wanted to see to be sure, nothing that made me think "ah that could be it!"

I drove the MR2 up there which is about 35miles round trip and it did run a bit hotter than previous, but maybe im just watching it more as temp gauge never moved from middle its just I had laptop running to see actual temp, basically it sat around 97-98c occasionally dropping to 95 coasting down a hill,  I also saw it hit briefly hit 100 for short time, still air in the system I thought like you say.

So got back  bled hot system and got some more and got air out rad, then rad started to get hot right across which got me thinking im getting close so lets get some more heat in the system see if it helps... or gets worse, switched fan off on the map and let it get upto 98c sat in garage which I could not get it to do before, rad is hot all over and the coolant started to circulate nicely with all pipes nice and hot....  Ive cracked it I think but not so! ...let it run like that for few minutes steady 97-98c all pipes hot but didn't drop temp, switched fan back on which cooled rad down fast and then coolant stopped circulating with temps staying around 97c....arrgh....went for drive kicking its arse a bit to get it hot and stopped to checked pipe under bonnet but rad cold from air flow and not circulating

Its like its got a 98c thermostat but ive tested it and know it opens at 82c as is it should, weird.

Tomorrow im going to remove turbo coolant pipes as you say and see what happens, it could still just be air in there but its acting very strange as it all circulates properly if you let it get very hot!  time for some food and a beer now

[Ardent]

I am not sure if any of this helps or not as my turbo setup is different to yours. But with this thread in mind. I had the Torque Pro rigged up to monitor my temps. 85 86 87 tops.

[thetyrant]

I am not sure if any of this helps or not as my turbo setup is different to yours. But with this thread in mind. I had the Torque Pro rigged up to monitor my temps. 85 86 87 tops.

Yep thats the temps for stock car and adding turbo should not affect things too much if everything is good, i just need to get mine down!...it was same when NA and sat around 95c all the time.

[thetyrant]

With all the bad weather yesterday i couldnt be bothered to fiddle with the car again, i did check coolant level though and it had dropped quite a lot over saturday night so it obviously burped some air out overnight, took it for a test drive in the hope everything would be sorted but from short drive i did (10miles or so) it wasnt much different, unfortunately the roads around me were so water logged i couldnt go much further or give it any hammer, it does seem to be back to the sitting at 95c and never above so looks like its back to where it was before i air locked heater circuit, checked levels this morning but no drop over last night.

Will work up the motivation to remove turbos coolant plumbing from circuit this week and see what that does....hopefully it does something!

[mr2garageswindon]

Is the temp sensor telling the truth? Have you tried another sensor to rule it out?
You could be chasing a fault that isn't there...

[thetyrant]

Is the temp sensor telling the truth? Have you tried another sensor to rule it out?
You could be chasing a fault that isn't there...

This is something ive thought about and trying another sensor is on the cards, i should of picked one up from Dicks on saturday...doh!

[thetyrant]

Been too cold to start changing pipes around to take turbo out the loop but did a few more tests where i didnt need to get wet!

First up i wanted to double check again that the coolant temperature shown on Link G4 software and that reported via the OBD from stock Ecu output on Torque app matched right through the range, from stone cold to fully warmed up, ive just done that tonight and they are within 2 degreed right from 4c to a reported 100c in fact the OBD output is slightly higher right at top end of that with Link data showing 98c but it was 100c on Torque app. 
I could fine tune that to get a perfect match via sensor calibration in Link map but its close enough i think and proved what i wanted for now.

Next test was to try and determine the ACTUAL temperature of the coolant/engine to see if its sensor reporting incorrect numbers as mentioned on previous posts above, its difficult to get accurate data of coolant temp externally  with infared gun on outside of the engine but i can report that with Ecu showing 98c the temperature of the cylinder head right next to the coolant temp sensor is only 87c, not a definitive test i now but its looking like it might be the sensor which would explain a lot, i found a spare sensor from my old Celica GT4 so will swap that over when i get chance and see what happens.

Would be nice if that is all it is but im not holding my breath, next course of action is still to remove turbo from coolant route then everything is back to stock in that regard.

[jonbill]

Are the g4 and Toyota ECUs sharing the coolant temp sensor?

[thetyrant]

Are the g4 and Toyota ECUs sharing the coolant temp sensor?

Yes they both use the stock sensor.

[jonbill]

Are the g4 and Toyota ECUs sharing the coolant temp sensor?

Yes they both use the stock sensor.

Is it possible to disconnect the toyota ecu from it as an experiment?

The ecus have a bias resistor that's involved in the correct calibration of the sensors. Gets screwed up with two connected. (although I just read the g4 may have some config options to correct).

[thetyrant]

Yes its a possible but the way its setup in the Link to monitor the voltage rather than resistance i dont think there is any interfearence issues, i can change the Link to work via resistance as well if needed and your right it can be adjusted to take into account the pull up resistor on stock ecu.

I was digging back through data logs last night to make sure the old ones i was looking at from OBD/Torque app were definitely before i touched the wiring harness to install the the Link or plumbing for turbo, i have found  one that was 100% done the week before i started changing engine harness for the Link and ive attached a graph of coolant temp over a 25 minute drive so you can see problem of either hot running or dodgy sensor was there well before i touched the stock cars wiring or plumbing,i just never paid it much attention at that point.

[Beachbum957]

The fans should be coming on above 93 C, so if the chart is correct, the fans should be running continuously.

The thermostat should start opening at 80-84C and be completely open at 90C, so I would expect the normal running temp to be in the 90C range.  I know on ours the temp gauge shows in the middle, but as soon as it goes above the middle the fans come on.  This happens quite often on a hot day when stopped

[jonbill]

Yes its a possible but the way its setup in the Link to monitor the voltage rather than resistance i dont think there is any interfearence issues, i can change the Link to work via resistance as well if needed and your right it can be adjusted to take into account the pull up resistor on stock ecu.

I was digging back through data logs last night to make sure the old ones i was looking at from OBD/Torque app were definitely before i touched the wiring harness to install the the Link or plumbing for turbo, i have found  one that was 100% done the week before i started changing engine harness for the Link and ive attached a graph of coolant temp over a 25 minute drive so you can see problem of either hot running or dodgy sensor was there well before i touched the stock cars wiring or plumbing,i just never paid it much attention at that point.

Easy enough to prove, as you say, with a replacement sensor, or cutting the wire to the toyota ecu. The ideas about bias resistors I've picked up on megasquirt and speeduino forums, and at least speeduino measures voltage but you still have to get the bias resistors right.