Configuration; MS3Pro on a "built" '99. BW 6758, ID1050's, DIY coolant reroute, aluminum radiator - other specifics supplied as needed.

My problem; As summer temps arrived, I started to get coolant boiling into the overflow tank even though the CLTs were indicating max of 180* F even while sitting in traffic, something new.

My "solution"; I switched out the CLT sensor (aftermarket - of course) but nothing changed. So I installed a K-type thermocouple into the return line, downstream from the stock CLT sensor, just under the intake manifold and am running the signal into the same CAN module that is reading my EGTs (on a spare port). Since I trust the thermocouple more than I trust the CLT sensor, I re-calibrated the CLT sensor to (more or less) match the thermocouple using a custom INC table.

The results; The temps from the thermocouple are within +/1 50* F of the CLT sensor (see the scatter plot on the left) which is good. The weirdness comes in looking at the graph on the left I'm expecting the scatter plot on the right to be at a 45* angle (more or less) which it is except when the "CLT Check" temp (the thermocouple) gets above about 175* F, then the CLT's go wild (eg. the "Broccoli Cut"). Does this mean that the CLT sensor is "bad", or is something else happening? I have bled the system more than once (suspecting air bubbles) so I think that's off the table.

So, any suggestions are welcome...Here's the plot.

 

Perhaps I'm being ---- about this whole thing...BUT I'm pressing on anyway.

I'm thinking that what I'm seeing is some form of cavitation where the "stock" temp sensor is located. My re-route is a DIY affair, with an adaptor from God-knows-where that I've had on the car for probably 15 years where the CLT sensor taps in. On top of that is a fitting from a Kia Sephia where the return hose clamps onto. Perhaps I should bite the bullet and get a REAL re-route adapter from FM or 949 and see what that gets me. I have two reservations with doing that;

1. Busting my knuckles getting the old adaptor out and the new adaptor in, and...
2. Introducing a water leak because I might not get the new adaptor gasketed properly or tightened properly.

Of course, I could leave things as-is and use the thermocouple output as the REAL CLT.

Thoughts?

 

Has the car been overheated recently? My only thought is head gasket leaking at coolant temps above 180, hence the bubbling coolant. There’s kits to check this (“block tester”), but in my experience, you get a lot of false negatives, so if you try it, don’t assume the head gaskets fine even if you don’t find exhaust in the coolant.

you could also check for leaks with a pressure or vacuum tester, and replace your rad cap. If you wanna be really fancy, install a pressure sensor in your coolant system and log it.

 

Now you got me "skeeeert" Curly!
There were a few occasions before I installed the thermocouple where the overflow tank was bubbling but my coolant doesn't have any oil slicks and my oil is oily (and not milk-shakey). I have even used the trusty old Doorman coolant bleeder funnel several times and only get gushes of coolant when revving - and no bubbles.

I know that none of that is definitive, so I'll scare up a pressure tester and give that a go. I HATE to think that I've blown a head gasket, or worse yet, warped the head.

 

To be fair to head gasket leaks, milk shake coolant/oil is only if the oil and coolant passages are combined. Bubbles in coolant is just combustion and coolant.

 

To be fair to head gasket leaks, milk shake coolant/oil is only if the oil and coolant passages are combined. Bubbles in coolant is just combustion and coolant.

 

I'm in line with curly's statement here as first path to diagnosing what's going on.

When my BP4W had bubbling coolant in the overflow observed after sustained highway cruising, I tried replacing the pressure cap first (no fix) and ended up putting a new HG in it for the fix.

 

Let me clear up one thing...the bubbles in the overflow tank ONLY occurred when the indicated CLT at the time exceeded 220 degrees. After I fitted the thermocouple and got "real" temps, I rarely exceeded 200 degrees and I don't get any bubble in the overflow tank any more.

That doesn't mean that I have escaped a blown head gasket, but I'm trying to convince myself that I don't have to go that far.

I'm going to get a tester to check for exhaust gasses in the coolant (noting curly's warning that false positives are possible) and go on from there.

 

Also...I'm going to throw this into the "mix". Here is a trace of CLT & the thermocouple on a 20 minute drive into work on Thursday;

The white trace is from the stock sensor, the red trace is from the thermocouple in the return line. As the engine is coming up to temp the two are more or less the same, with the sensor tracking a bit higher (which is factor of the calibration). Then, about halfway through the drive the sensor starts to drop below the thermocouple but manages to "recover". Then, at about the 15 minute mark, the sensor values start to go berserk, dipping into the 120's while the thermocouple remains relatively steady (yeah, it's "fuzzy" because the signal line for the thermocouple is not sheilded and it runs pretty near the coils).

My point in all of this is: Why are the sensor temps relatively stable early on, but get wonky later on? I'm going to do a longer run tomorrow and see what happens (do the sensor temps "recover" again, or do they stay wonky?). This behavior makes me think that I have a sensor problem and not a HG problem - yes, I grasping at straws, can you blame me?

 

So get a new sensor, they are cheap (I think), and it will mean you will need to re-burp the car, properly with the big funnel and nose in the air.

I think in my build thread, or elsewhere here, I have reported the grief a failing CLT sensor caused me.

 

You know, it's AMAZING what a difference good wiring makes...

This morning, after having dreams last night of having to pull the head to put in a new HG, and how wildly enthusiastic I was to do that, I took a look at the CLT sensor wiring just for ***** and giggles. What I found was not good. Those wires had experienced some sort of "heat event" (i.e. the wrap over the wires was melted and the insulation was slightly melted. I think that was a result of a small engine bay fire a couple of years ago when I left off the turbo heat shield durng a track day at Atlanta Motorsports Park where the brake resevoir melted and puked brake fluid all over the hot downpipe. Anyway, I thought I had repaired all of the damage done during that - evidently not.
So, I spent some time this morning repairing those melted wires and securing them away from a future "heat event". This is the result (after re-calibrating the thermistor table);

Thanks everyone for your input and scaring me enough to do some REAL critical thinking about this problem.

On to the next "mystery"!

 

Coolant at atmospheric pressure around 215-220F (depending on coolant supplier), so this type of feedback would be indicative of a pressure cap (or other gross leak) issue.