Swings in coolant temps

[PbFut]

So this is what I get for messing with something that was working fine.
Over Thanksgiving, I decided to drop the thermostat from 180° to 160°. Before the change, my ECU sensor in the head was hovering around 190° while cruising, and I noticed the engine seemed to run better in the high 170s to low 180s.
To swap, I drained coolant just below the intake thermostat chamber lip. I use a marked jug and siphon coolant to the mark, so in theory the heads should stay free of air. Since the swap, though, I’ve been fighting what I believe is an air bubble—but it behaves strangely. After a third thermostat install, I went back to the original 180° high-flow unit, and the issue still persists.
Radiator is full and cycling fluid to/from the overflow bottle.
Symptoms (note: everything worked perfectly before the swap):
- On startup, the ECU sensor in the head shows overheating (205–225°), while the factory sensor in the intake reads fairly cold.
- If I move the A/C lever to close the heater hose valve, the factory sensor rises and the factory gauge moves to about mid-point.
- This all happens within just a few minutes of warm-up.
- Heat gun readings show components closer to 140°.
- While driving, temps won’t stabilize—both sensors swing and rise.
- If I pull over and shut the engine off for about a minute, then restart, temps return close to normal. Sometimes I need to repeat this twice. After that, everything quickly settles into normal temps and stays fine until the engine cools back to ambient. Then the cycle starts over again.
You’d think if a bubble remained, I’d see issues while cruising—but once stabilized, it runs fine. I’ve tried all the usual bubble-bleeding tricks with no success change. The EFI ECU doesn't like the temp swings and the warmup tune is difficult to keep stable at this point.
Any other thoughts why this is happening?

[Rust_never_sleeps]

IDK what motor you have, but maybe crack the temp sensor loose and see if you can bleed anything out there

[franken]

The tstat just opens at it's temp. It doesn't and can't do anything to limit the running temp of the engine.
Think about it. How could a tstat keep engine temp from rising?
All it is is a temperature controlled valve. It's there to get the oil warm to limit engine wear.

[PbFut]

Originally Posted by Rust_never_sleeps

IDK what motor you have, but maybe crack the temp sensor loose and see if you can bleed anything out there

My intake manifold has 3 ports in the crossover chamber where the tstat is. One is for the heater core. 2 for sensor. The other i place a schader valve attempting to chase what I thought was a bubble. No help. I am now leaning towards the idea I'm not getting enough bypass through the heater hose. As I write this, it just occurred to me that I rerouted the heater hose exiting the manifold when I did the original swap. It was looped over the ac compressor to the heater control valve. I looped it under the compressor for a cleaner look. Is it possible the downward arch routing in this manner is limiting flow and I am getting a steam pocket in the head until the tstat opens and full flow clears the problem. I think I will experiment a little later today.

[my04]

Try jacking the front of the truck up as high as you can and run the engine with the cap off until hot and see if the trapped air will "rise" toward the radiator. This should provide a high point vent for air to escape.

I just re-read your post and realized that you may have tried this already, if so, then disregard.

[mike16]

Your thermostat may be fighting your fan clutch.
Your 160 degree thermostat begins to open at 160 degrees and should be fully open by 180 degrees.
Your fan clutch may begine to function at 180 degrees. Thus it cools the coolant and reduces its temperature below 180 degrees causing the t stat to begin closing. This causes a restriction which slows the flow of coolant causing coolant to absorb more heat and as the temps rise the t stat opens and the fan clutch sensing the increase in the temperature, it engages and begins to cool the coolant again causing a continuous cycle of thermostat to open an the fan to close it again. You see this occurring as a function of the guage.
The solution is to flush your system thoroughly. Keep your 160 degree thermostat and check your fan clutches operating or engaging temperature to make sure it opens around 200 to 210 degrees.
Also turn on your heater and feel the change in temperature through the blowing hot air temperature change.
People don't realize that the t stat and fan clutch should operate in harmony and to compliment each other's function. The t stat should begins to open about 20 degrees before engine reaches ideal operating temperature. The fan clutch should engage the fan when it senses an overheating condition of 200 to 210 degrees. If the t stat and fan clutch t operating temps are too close together they conflict with each other. And in my opinion that's what's happening as indicated by your guage

[mike16]

Sorry to blather on and on but where does air get trapped in a sbc? I've worked on sbc's since 1978 , on trucks, cars and vans, privately and commercialy and have never had to use any procedures to vent air from the cooling system.
I've done it many times on other vehicles and understand the procedure completely but never had to do it on a sbc.

[PbFut]

Air will trap in heads and heater core and slow the bypass effect needed to avoid hot spots and steam pockets. I went back to the original working for 8 months stat and still have the problem which makes me think that heater hose route change under compressor may be the issue. Ill get to it in a couple hours.

[HO455]

Exactly what engine are you talking about?

[cbeerman]

Sounds like your rerouting maybe the cause. If not, I was wondering if there is a bleed hole in the new thermostat.

[PbFut]

For those asking, here are the details:
This is a BPE 400 SBC with port EFI. The ECU temp sensor is in the passenger‑side head and controls two fans. The factory temp sensor is in the crossover passage next to the thermostat. The heads on this motor have a smaller sensor boss, which required relocating the factory sensor. Original and new thermostat have bypass holes.
Heater hose reroute test:
I tried rerouting the heater hoses, but on this first heat cycle it didn’t change anything. It’s possible another cycle and a drive up the hill might purge the remaining air — I still had about a pint left in the drain jug. The heater hoses are warming up, but they feel the same as before, which is why I hadn’t bothered changing the routing earlier. The heater itself blows at normal hot temps, so at this point I don’t think hose routing is the culprit.
Warm‑up behavior:
During warm‑up, the same issue occurred. The ECU sensor in the head climbed to around 240°F. The factory sensor in the intake only reached the midpoint on the factory gauge. I shut the engine off for about a minute, restarted it, and within another minute coolant flow began behaving normally. Within two minutes everything stabilized.
Once stabilized, the factory gauge sat just above the cold mark. Head temps cycled between 178–183°F, which matches the ECU’s fan‑1 on/off settings. Fan 2 (set at 190°F) does not trigger during normal operation, but it does trigger during this odd warm‑up phase — so at least the ECU logic is working correctly.
Working theory:
It really seems like a small air bubble forms in the head during warm‑up, creating a localized hot spot that escalates into a steam pocket. Both thermostats — the original and the new one — have bypass holes and are correctly oriented. But it’s obvious that coolant flow is poor until the system “primes.”
The big question:
What’s preventing proper coolant flow when the system is cold? That’s the million‑dollar mystery at this point.

[HO455]

Have you tested the coolant for exhaust gases? Maybe the head gasket (Or cylinder wall) is allowing compression gases into the coolant when the engine is cold and seals when warmed up. Thus causing a new bubble of gas on every cool down.

[Rust_never_sleeps]

Originally Posted by PbFut

[B
The big question:
What’s preventing proper coolant flow when the system is cold? That’s the million‑dollar mystery at this point.

Perhaps drill a small hole in the thermostat?

I'm interested in Mike16's analysis

[PbFut]

Originally Posted by HO455

Have you tested the coolant for exhaust gases? Maybe the head gasket (Or cylinder wall) is allowing compression gases into the coolant when the engine is cold and seals when warmed up. Thus causing a new bubble of gas on every cool down.

An excellent idea. I had not thought of that. I am not getting any milk in the oil that I can tell. But exhaust into the coolant? I think this may be the next step. Does anyone have a source to send for testing?

[PbFut]

After nearly 3 weeks of chasing this cooling issue, I finally got it resolved. In hindsight, I should’ve stopped, reset, and started over instead of stubbornly trying to make each step work.
The final fix really was to start from scratch.
I pulled the second 160° thermostat and launched it out the back of the garage. Then I reinstalled the thermostat housing with no thermostat and fully heat cycled the engine. With no stat in place, the temps were rock solid—settled right in the 170s with zero swings. The fans didn’t come on until heat soak from lack of airflow, and even then they behaved exactly as expected.
Next, I picked up a third 160° high flow thermostat. I drilled three 3/16" holes in it, same pattern as the 180° stat that had previously worked perfectly from the aftermarket. I left the rerouted heater hoses in place (over the A/C compressor instead of under).
I tested the new drilled thermostat on the stove: it started opening at 159° and was fully open at 166°. Not a lab grade test, but enough to confirm it was functioning within tolerance.
Installed the new thermostat.
It worked perfectly on the first run.
Because I was already frustrated, I didn’t test after each individual change, so I can’t say with certainty what the root cause was. Maybe running without a thermostat finally purged a stubborn air pocket. Maybe I got two bad thermostats in a row. I can’t confirm.
What I can say is this all started because I wanted to try a thermostat 20° cooler. Now that everything is finally behaving again, it’s clear the lower temp stat keeps the head temps in the low 170s, which pairs much better with the ECU tune I’m running and results in a noticeably smoother engine.