p0126 persists after new parts and various attempts

MortalMechanica

MortalMechanica

Greenie N00B Member
Greenie Member
Hey there folks. Wanted to bring up an ongoing issue I've been having for the last few months with the hopes someone here has insight into my situation.

A few friends and I have been trying to narrow down the issue with myself and another's P0126 (Engine coolant temp threshold) problem for the past 4 months. I'd appreciate any feedback on this as I've exhausted most of my capacity forward.

I blew a hose a few months ago and I ended up replacing it. no big deal, it happens. About a month after, I noticed the car was getting a p0126. Not the first time it's ever had it. I went ahead and changed the Thermostat, Initially, I was just being a cheap bastard and threw the 20 dollar Motorad on there. did a coolant flush, no problems. P0126 came back. Weird I thought. Monitored temps and saw that the temps were doing weird things. like jumping around a lot, going from normal temp to around 160. etc.

So I thought Must be the ECT going weird, checked the connector, checked grounds, replaced ECT, and replaced T-stat with an OEM one, because everyone in the speeds group was certain it was because of the motorad stat. no big deal.

Still kept getting the same behavior from the coolant. inexplicable fluctuations that would lead to a p0126. I bled the coolant more, did some more tests, figured it was "possible" for the oem t-stat to be bad. So I replaced it AGAIN with another oem t-stat just in case. It was at this point I did a pressure test and a HC test to rule out the obvious problems of issues with the motor/block. Everything seemed fine as neither test showed any problems. Changed the ECT sensor again for certainty. checked to see if the coolant was circulating or if the water pump was leaking/failing. couldn't find a mechanical issue.

We checked the temp with a FLIR camera and came to the conclusion that the ECT wasn't seeing or sending the right temp signal, that's why it was replaced again. all oem parts. Did the song and dance again. checked for leaks, checked for missing coolant. Nothing that would indicate that there was a physical problem with the engine.

Here's a few videos of the symptoms/trigger
Template public:_media_site_embed_youtube not found. Try rebuilding or reinstalling the s9e/MediaSites add-on.

Template public:_media_site_embed_youtube not found. Try rebuilding or reinstalling the s9e/MediaSites add-on.

and another

Template public:_media_site_embed_youtube not found. Try rebuilding or reinstalling the s9e/MediaSites add-on.

And a couple of datalogs.
Template public:_media_site_embed_googledrive not found. Try rebuilding or reinstalling the s9e/MediaSites add-on.

Template public:_media_site_embed_googledrive not found. Try rebuilding or reinstalling the s9e/MediaSites add-on.

I had a friend over at a Mazda dealership take it and he had it for a couple of days, he looked at it gratis. Called Mazda engineering, they swore up and down that it had to be the thermostat or an Air pocket. He also went ahead and tested the continuity of the wires and said that they weren't showing any problems. But The thing about the issue is that it doesn't always happen. I figured he tested it when everything was looking good and that everything was fine.

So. Here I am. still getting weird temp swings like in the videos above, with inconclusive tests on the health of the wire loom. So it's either a PCM or the Harness at this point, intermittently failing and reading open, and causing the coolant temp to do weird shit and throw the code. I throw myself on the mercy of the forum, in hopes that I can help both myself and a few other people with the issue I have.

TLDR
P0126 appeared
1. replaced thermostat (motorad) - no fix
2. Replaced thermostat again (oem) - no fix
3. replaced ECT (OEM) - No fix
4. Replaced thermostat and ECT (both OEM) - No fix
5. Pressure test (No leaks) HC Test (no HCs detected)
6. PCM Reflash - no fix
7. Harness shows continuity (at least during the testing of the harness according to my friend)
8. Problem not fixed - Currently unable to pass emissions because of it. I've driven it a few thousand miles over these few months as well.
9. Suspect either Harness or bad PCM
 
If it's harness, you should be able to trace the wire back and see if it's indeed bad. My thoughts are air pocket. Seen em hide over the past 10 years on forums. A high pressure flush and I think front end up will possibly find it. There are equipment shops use for this too.
 
So you've replaced the thermostat and ECT sensor twice, let's safely assume its not either of those.

There is a ground point around the HPFP that can fuck with a lot of things if it's not secured properly or frayed. Check that since it's in the vicinity of the ECT sensor.

Also, check your coolant reservoir cap. If it's not sealing properly and letting the cooling system pressurize, it will run colder/hotter and can potentially throw a code.

Last thing is on the ECT sensor, since you've replaced it twice make sure the pins are not bent or being pushed out of the pigtail. The ECU on our cars is very picky about even the slightest bad connection.
 
Last edited:
Just watched your videos, that temp jump is very drastic.

Those videos were taken with the A/C off? If you turn the A/C on, does it make any difference?

When your fans do kick on, how long do they stay on for?
 
I ran into this issue a while back; temp skyrocketed for no reason and then plummeted to like 140. I panicked at first, but after just wiggling the connector for the ECT sensor it started behaving normally again.

Had a similar issue with the MAF sensor before my engine let go originally too; wound up wiring up a new connector (for the MAF), which might be your best route forward. If my ECT does this again, I'll probably just buy a wholly new harness and be done with it but I don't think that's a viable option for most people.
 
Raider said:
If it's harness, you should be able to trace the wire back and see if it's indeed bad. My thoughts are air pocket. Seen em hide over the past 10 years on forums. A high pressure flush and I think front end up will possibly find it. There are equipment shops use for this too.
Click to expand...

I mean. I'm out of options at this point. I've followed the FSM procedure to bleed the coolant on the vehicles of other speed owners, but I've never had this issue with mine. I've been as far as having jacked the car up as well, but I suppose I can try the airlift and see what it finds. The harness and the connections to the ECT tested fine, and no matter how much harness wiggling I do, I can't seem to get a consistent fault. pins look good and the pins on the ECU look good. Went ahead and checked all my grounds too to boot including the one by the HPFP and the one that pigtails off of the ect itself.
 
Mauro_Penguin said:
Just watched your videos, that temp jump is very drastic.

Those videos were taken with the A/C off? If you turn the A/C on, does it make any difference?

When your fans do kick on, how long do they stay on for?
Click to expand...

I read both of your posts Both the continuity check in the loom, the condition of the ECT plug, and the connection to the ECU are in near pristine condition, no pins damaged no pins pushed out, everything clicked and tight. Wiggling the wires also doesn't seem to prompt the issue. I responded to Raider up above Re: using an airlift tool to see if perhaps some air is trapped.

The videos were taken with the AC off, but I have tested it with the heat on, with the AC on and the conditions seem similar. I don't think air would cause a jump so drastic unless there was a huge air blockage. But no. the AC being on and normal cycling doesn't make a difference, There's no drop in heat quality from the core either.

When the fans do kick on, they kick on for normal cycling time 25ish seconds at my sloppy count before turning off and then starting up as the system builds pressure in the AC system. If you meant how long they run when the motor hits 219F? probably about 8-10 seconds until the coolant hits less than 210 and the fans stop. The fans seem to be functioning normally and within normal spec.
 
Enki said:
I ran into this issue a while back; temp skyrocketed for no reason and then plummeted to like 140. I panicked at first, but after just wiggling the connector for the ECT sensor it started behaving normally again.

Had a similar issue with the MAF sensor before my engine let go originally too; wound up wiring up a new connector (for the MAF), which might be your best route forward. If my ECT does this again, I'll probably just buy a wholly new harness and be done with it but I don't think that's a viable option for most people.
Click to expand...

I can't seem to get a consistent fault wiggling any part of my harness. I just spent about an hour shaking and examining what I could and there was no fault detected, engine warmed up to 187 in about 10 minutes, however, after a brisk drive and a break to have some lunch, starting the car a second time (car sat for about an hour and coolant temp dropped to about 175 normally) the coolant climbed up to 178, dropped to 150 for about 10 seconds (which caused the motor to enter the higher warm-up idle) and then jumped back up to 180. With no changes on my part.
 
Is it possible to install the thermostat backwards on these cars?
 
No, it's part of the housing for the hose. you swap the tstat you swap the part that goes on the hose too.

What I am interested in is if anything changes if the heater is left on (or off if already on), though that shouldn't matter because the sensor should only ever see outflow from the head.
 
Enki said:
No, it's part of the housing for the hose. you swap the tstat you swap the part that goes on the hose too.

What I am interested in is if anything changes if the heater is left on (or off if already on), though that shouldn't matter because the sensor should only ever see outflow from the head.
Click to expand...

Makes no difference. I figured if the heat was swinging aggressively as it was, I would feel a drop in the cabin heat somewhat, but nope. Same with if I leave it on or off regardless. This is partially why I think it's PCM or Harness related, having serviced every other part. And to my memory, air in the coolant tends to cause upwards fluctuations. not downwards fluctuations.

For now I'm going to throw an air-lift on it and see if somehow there's a cavity of air trapped somewhere it's not supposed to be. Pending that. I think my only options are to try a known good harness or a pcm.
 
Unplug the sensor and plug it back in a dozen times or so.
 
Just an update/bump for anyone interested.

Problem persists. Replaced the ECT and the Thermostat with working units from another car to run some tests. Temperature fluctuations still present and in some cases vary wildly from 160 to 230 while cruising. in some cases enough to trigger a "cold coolant" code. The car feels like it takes an unusually long time to heat up sometimes and the rise in temperature isn't linear. Sometimes it will get to 170 then drop back to 160 then rise to 171 then drop to 165 within the span of a few minutes.

Took it to a friend at a Mazda dealership nearby and he had a look at the harness. The signal from ECT to PCM and PCM are showing no damage or loss of connection during testing. No U codes. No signal loss. Nothing. ECTs were tested just to be certain and they all show the same and they operate within spec. I suppose it's possible that the harness is damaged somewhere that is only apparent when the car is moving? But it's done this while idling and cruising so.

I still need to try to put the system under a Vacuum as @Raider suggests. I do notice that if I leave the cap off long enough, once the engine gets above Thermostat full open temp (closer to 200F+), I'll get a steady stream of bubbles but I figure these are because the coolant is boiling because the system isn't pressurizing with the cap off. This only happens when the engine is hot as well. Tested the bubbles for hydrocarbons as well and they showed no combustion gasses.

Template public:_media_site_embed_youtube not found. Try rebuilding or reinstalling the s9e/MediaSites add-on.

I'm not entirely convinced it's trapped air given how many heat cycles the engine has been under while properly sealed. I'm gonna fire the parts cannon at her again and change out the water pump and see what happens with that. Mostly because I don't have access to a vacuum system without buying one myself for cheaper than what a local shop would charge me for the time out here.

I'm open to any other suggestions. Having ruled out most of the typical solutions. I ran a leakdown test and it's definitely getting close to rebuild time. with compression showing 155, 150, 130, 185. I didn't see any bubbles in the coolant during the leakdown, but since the engine was already hot from a long drive, I didn't think to take the coolant cap off, and the leakdown showed loss via rings. I didn't think to depressurize the cooling system before the test. So... maybe?

Bear in mind I've seemingly lost no coolant at any point in my testing. no smoking or burning of coolant detected. Coolant has been added after changing the thermostat to bring it back up to level again, but I'm drawing straws here.
 
Spent about an hour exploring some thoughts from the above. Still getting a similar result as the above video with air seemingly in the system. I spent about an hour with the cap off on the system and used the rad fan to keep the coolant from boiling, but hot enough to keep the thermostat open.

I'm starting to have suspicions that this might be head gasket related, mostly because of the wild temp swings. (tonight it would vary from 170... 190...205...178...187...238 in the span of about a minute) The number of bubbles coming out of the smaller hoses on the radiator indicates to me that (assuming the coolant travels clockwise in the system) There must be some kind of combustion leak.

The p0126 seems like an underlying symptom to a more complex problem. The MIL lights up most likely because, if there is air in the system that is only visible in the reservoir AFTER the Thermostat opens (this is consistent with my findings) Then during warmup (between 0 and 176ish) the bubbles/air that is being generated or trapped in the system are settling at the high points, which in this case seems to be around the ECT area and likely others. Only once the Thermostat (which seems to open and close at the correct temperatures) opens enough for the contents of the Surge tank to begin swirling, indicating that it's open, do the bubbles now have the ability to flow from the Upper hose (on the driver's side) through the radiator where the bubbles are rising and then enter the Surge tank via the two smaller hoses on the right of the tank.

I guess the big question for me is. Where is that air coming from and why isn't it stopping? I've also noticed that revving the engine at any point after the thermostat opens often results in a healthy rush of bubbles/air. I'm going to do another hydrocarbon test after a longer test period.

The other question is... why does my system pass a pressure test? I had assumed if there was a leak somewhere that the pressure would be lost from that point, but there's no pressure being lost.

I'll also likely purchase an air-lift system to see if I can vacuum the system out which gives me another pressure test at least. And if there is somehow a MASSIVE amount of air trapped in the system (I'm not sure this is the case) then it should be removed.

At this point I get the feeling I'm seeing the result of air pressure (likely combustion?) entering the cooling system that is only apparent when the thermostat is open and allows the air to enter the tank. But since the leak/break in the head gasket isn't massive, it's not a consistent rush. what I have noticed is that after long drives my Coolant tank is terribly distended, and is showing stress marks/cracks at the corners. Almost as if my cap isn't releasing that pressure either? Though I've read also that it's possible that once the system is pressurized to 21psi, if the crack in the gasket is small enough it's possible that the air has enough resistance to prevent the leak.

though I am curious why I haven't been losing any coolant, and why the HC test didn't show any combustion gasses the first time... I saw a lot of bubbles like in the above video tonight. I have a feeling I'll wake up in the morning and the system will not have lost any coolant. I've heard that water pumps can have a one-way leak through the seal... but this seems wild to me. I know that on some engines it's again 'possible' to pressurize a system on a combustion stroke and not suck in coolant on the intake stroke.

Aside from the above, I'm open to any and all suggestions. And I hope for those of you who seem to have had this problem and have reached dead ends from the other forums, stumble upon this and find something that might help.
 
Last edited:
The difference between maximum engine vacuum and average combustion pressure is, quite frankly, insane. If you think about it, it's also far easier to tear a circle (head gasket) apart from inside out than it is to compress one from the outside in (in which case any minor tear should re-seal with minimal losses).

Food for thought.
 
Enki said:
The difference between maximum engine vacuum and average combustion pressure is, quite frankly, insane. If you think about it, it's also far easier to tear a circle (head gasket) apart from inside out than it is to compress one from the outside in (in which case any minor tear should re-seal with minimal losses).

Food for thought.
Click to expand...

That's true that's true. The only thing that doesn't make sense is the lack of HC's being detected. But I'll test that again tomorrow. Either way, the engine is on borrowed time. After 150k miles I think she's just tired, and all I'm really accomplishing with fixing this is pushing the rebuild a little further down the road.

I think it's likely there is an air pocket. But I think that pocket is being created from either combustion gasses or some other air incursion not found by a coolant pressure test, which I feel rules out a leaking hose or junction somewhere.

With how much air has come out of it, every time I try to burp the system unless that pocket is massive, it feels air is being replenished somehow. Even if it was a trapped air pocket, the bubbles rising in the tank would indicate they were being displaced with coolant.... unless somehow the air is constantly bleeding into the system. which is probably why the coolant level never changes despite seeing all of this air.

Yeah, this has been a frustrating journey, that would probably be fixed if I had the engine rebuilt. but I'm cheap right now, and I figured if I could fix this then I could get a few more months out of her.
 
as far as the lack of hydrocarbons being detected, with a small leak and relatively few miles on fresh coolant the contamination may just not be high enough to be detected by your testing method.
 
You must log in or register to reply here.
Back
Top