Head gasket replacement: What did I do wrong?

Yes, and the boss now allows me to eat indoors now. The compression gauge is for the car, not our digestive system.
 
There was the beginning of coolant boiling in the plastic reservoir.

I think the coolant in the overflow tank was just on the verge of boiling, but I shut off the engine (both times) before that could really get going, so not sure about this.

How can you tell if coolant in the reservoir tank is beginning to/on the verge of boiling? It's either boiling or not. There should be no bubbling of coolant in the expansion tank (or in the radiator), & if there is, you have 1 of 2 things going on:
  1. Compressed exhaust gasses are being pumped into the cooling system, possibly caused by a blown head gasket, a crack or pinhole in the cylinder head, or a crack or pinhole in the engine block.
  2. There is an air leak within the cooling system which is not allowing the system to pressurize, thus allowing the coolant to boil. Typical cause for this is a loose hose clamp, a hose with a pinhole or a failing radiator cap seal.
What was the initial reason that you removed the cylinder head from the engine?
 
quite a typical over heating causes on 850's are;

the radiator cap not sealing. 850's use a longer radiator neck, so the cap is a little longer than normal (longer than 124 or X19). simple enough to do a quick check of the radiators neck depth against the cap

the water bypass hose not allowing water to circulate / air bleed out of the system. If the two way connection where the temperature sender mounts is blocked/ bypassed then air doesn't bleed out of the water pump. Is the temperature sender the correct type for this bypass/ bleed system? (with the internal hole to allow water circulation)

Unless you paid big money for a racing head gasket, all you got was a regular one... they do come in various thicknesses, and yes that is to make up for any material that has been skimmed from the head surface when reconditioning. In the old OHV's this just makes it easier to maintain proper lash adjustment / pushrod and valve train geometry. There's nothing "reinforced" about them, they are simply thicker. You do need to torque and then retorque the head bolts after a heat cycle as they are composite gaskets.

If the engine starts / runs and idles well, and then restarts on all four cylinders - not starting on 3 and then clearing to 4 cylinders - then you might be lucky and have not overheated anything too badly so far. If yor shutting it off and then it begins to sound like a boiling kettle that's heat soak causing the boil, and hopefully you havent done any damage to the gasket / warped the head again.

SteveC
 
Question on the temperature sender: It has the single hole at the bottom end, then a through hole through the center. May I assume the angular orientation of that through-hole doesn't matter (since it's controlled by how far the sensor threads into the head)?

Jeff, what I meant by 'on the verge of boiling' was 'I thought a saw a few bubbles forming in the reservoir, and I didn't want to go any further'.
I removed the head because it had been running too hot for a couple of months, I could barely keep it out of the red zone after prolonged highway driving. I had already torqued the head bolts as high as the spec would allow, replaced the thermostat/temp sender/radiator cap (which made no difference, the old ones were still good), and I couldn't think of anything else to check. As I mentioned at the beginning, the machine shop did say the head was warped.

FWIW all the cooling system parts I replaced were sourced from Midwest-Bayless, as were the original parts I took out, so I'm confident they're all the correct ones for my car.

Oh, I made a mistake in my initial post: My temps were around 115C and climbing (not 130C, which is the top of the red arc) - my bad.

At this point, the best hope is to entice Magic Carl over to my garage (by offering beer and refried beans), confident in the knowledge that he will diagnose the problem within Mere Minutes. Yes America, he's just that good.
 
Last edited:
Jeez, I owned a brand new 850 coupe in 1967, haven't played with one since!
I am sending Dave a prescription for the proper medication to fix the problem, permission to go buy a used Miata.
 
This really sounds to me like an unsealed coolant system. Stupid question, but you have the radiator cap and reservoir cap fully installed when doing the run tests, right? It's also worth replacing both caps to rule them out.

How can you tell that the coolant is starting to boil in the reservoir if you have the cap installed? If the caps are off or not sealing, then the temperature will continue to rise and rise.

EDIT: What's the condition of the radiator?
 
Yes, both the radiator and reservoir caps were properly installed during my two brief op-checks. Both times, immediately after shutdown, I removed the reservoir cap and peeked in and thought I saw a few bubbles starting to form. Not conclusive, but I didn't want to risk any more damage.
Radiator condition is a question mark.
 
Yes, both the radiator and reservoir caps were properly installed during my two brief op-checks. Both times, immediately after shutdown, I removed the reservoir cap and peeked in and thought I saw a few bubbles starting to form. Not conclusive, but I didn't want to risk any more damage.
Radiator condition is a question mark.
Assuming a properly functioning cap and a properly sealed (aka no detectable leaks) system, Removing the cap in and of itself would have caused the evidence of boiling you saw....IOW you created your own "problem."

Depending upon which chart you find on Google, these values may vary slightly:
50-50 mix of ethylene glycol coolant boils at 226 degrees F at zero pressure
50-50 mix of ethylene glycol coolant boils at 248 degrees F at 8 psi
50-50 mix of ethylene glycol coolant boils at 259 degrees F at 12 psi

So if we assume you have a 50-50 mix of EG antifreeze and a good rad cap & overall system that holds about 10 psi of pressure, your coolant would not boil until about 253 degrees F aka 122 degrees C.

You stated that your gage was showing 115C when you shut off, below the expected boiling point at 10 psi of 122C. So there should not have been any boiling.

By removing the cap the pressure went from about 10 psi to zero, instantly reducing the boiling point of your coolant to 226F aka 107C. Since the indicated 115C is more than the no-pressure boiling point of 107C, presto, instant boiling.
 
Myron, the story Dave told me was the orange antifreeze pretty much turned to jello in his cooling system. But I'll let him give you the first hand gory details.
 
The story on my previous X was, the orange Dex-Cool mixed with the standard green antifreeze already in the system, creating a thick gray sludge that ruined the entire cooling system, including the heater core, and ultimately caused me to sell the car as a test mule to a 'programmable EFI' developer.

And Dan: I did not remove the reservoir cap while the engine was running! I only removed it right after shutting down. That did NOT cause the overheating. Also, I had made a radiator drain plug repair and wanted to check the integrity of that under pressure before adding a full 50% of antifreeze, so the coolant at that time was mostly distilled water - I'd guess probably 70-80% water. Thus the boiling point would have been a bit lower than your numbers for 50-50. Regardless, the temperature was climbing out of control.
 
Last edited:
The story on my previous X was, the orange Dex-Cool mixed with the standard green antifreeze already in the system, creating a thick gray sludge that ruined the entire cooling system, including the heater core, and ultimately caused me to sell the car as a test mule to a 'programmable EFI' developer.

And Dan: I did not remove the reservoir cap while the engine was running! I only removed it right after shutting down. That did NOT cause the overheating. Also, I had made a radiator drain plug repair and wanted to check the integrity of that under pressure before adding a full 50% of antifreeze, so the coolant at that time was mostly distilled water - I'd guess probably 70-80% water. Thus the boiling point would have been a bit lower than your numbers for 50-50. Regardless, the temperature was climbing out of control.
I am not stating that removing the rad cap caused the overheating, I am stating that removing the rad cap would reduce the boiling point of your coolant instantly. If the coolant was already boiling with the cap on, taking it off would only exacerbate the problem.

A better way to test the integrity of a cooling system repair would have been to get one of those cooling system pressure test kits that the major chain auto stores loan out to customers.
 
Dave said he removed the overflow reservoir cap, not the radiator cap. Removing the reservoir cap wouldn't do anything, regardless of the engine running or not.
Evidently the 850 cooling system is a conventional rad system with an overflow reservoir....not an expansion tank like an X1/9.

So in the famous words of Emily Litella:
1626575761453.png
 
UPDATE: Carl the Magic Man, driving his trusty Rat-X with the remarkable FAZA Overhead Air Scoop, came over to assist with troubleshooting.

Results of compression check: 135/135/140/150 psi. Conclusion: Head gasket probably sealing fine.

Dirty-ish new oil: Carl didn't think it indicated anything untoward, probably just dirtied from the oil pan. Conclusion: Oil and coolant appear normal.

Test-run: Same climbing temperature as before, but this time we checked radiator and hose temps by touch and IR thermometer. While upper radiator hose and top of radiator got hot, the lower radiator hose never even got warm. Carl's deduction: Radiator is clogged.

So I will remove the radiator and get it tested and repaired if needed. Assuming I can find a radiator shop in Northern Virginia.
 
Concur. The thermostat was removed for this test so the motor should have idled away forever without even reaching normal temps. But after about seven minutes it continued to heat up and reach for the red zone. The bottom hose and bottom of the radiator was cold. Dave did admit it was the original radiator. Ergo, therefore and henceforth my money is on the rad being the problem. At the very least a good rad needs to go in before any meaningful further investigation could be done.

And Dave gave me a good excuse to get the X out, first day in about a month that the temps were in the mid 80s and not the high 90s.
 
Back
Top