Hi, I've just rebuilt my B16A and I'm having trouble with the cooling system. The radiator fan is switching only above the middle mark at the dash temp gauge, almost 3/4 from the C mark. It stays on for about 15 seconds or less and switches off. The temp decreases to about the middle mark and never drops below that. When cruising (only have done about 15 miles of driving, nothing more until this get solved) it stays around the middle mark and rises just above at a stop, never a 1/4 of the way up from C as before. Thermostat opens as usual, coolant was bled per service manual instructions before the first drive, and then twice after I noticed this issue. One time following the service manual and then another time with the heater on, and the cap off. There are no leaks at any of the sensors or hoses. I did see some bubbles coming out at this third try after the fan switching on the second time, but the issue remains. I did have the OEM sensor previously, but unfortunately I've lost it and replaced it with an MTE-Thomson 3139 that replaces the OEM part number. I don't know what else to test, the rad cap seems fine but another one (OEM) is on the way. It also doesn't spill water out of the reservoir and there are no signs of boiling or anything like that. Coolant is a good organic type without borates or silicates (as per Honda's recommendation) mixed 50/50 with distilled water.

 

Test or replace the ECT sensor.

 

How many times did you let the fan cycle on and off before you test drove the car?

When I do engine work that involves rebuilds I try to replace the sensors that I know are a PITA. If you aren't getting new parts its always a good idea to do as wunfstgsr said "Test or replace the ECT sensor."

If the thermostat is opening up and closing as it should, I too would suspect your ECT.

 

The fan cycled twice during the bleeding process before the test drive, every time I've bled the system the fan turned on twice at the same spot on the dash Gauge before I finished the procedure. The fan switch is new and the old one was also used to test, same thing happens. The temp sensor that sends information to the dash is also new, MTE-Thomson 3139. The ECT sensor (assuming you're talking about the sensor that sends the readings to the ECU) was not replaced, but I don't see how it would affect the reading at the dash or the fan switch. If it does, I'm curious to know how. Another thing I've noticed is that the car used to keep the rad fan on for a while even with the car off with the key at the 2nd position if the fan was previously on. It doesn't do that anymore.

 

Man I’ve had so many gremlins with the fan coming on reliably that I just always paper clip it now. I’ve replaced every sensor and paid hundreds to have the same problem repaired professionally twice only to stop working correctly within weeks. Paper clip the thermostat plug and run the fan permanently.

Just my two cents, call me an idiot if you’d like. I’m prone to worry. So I really like knowing for a fact my expensive time consuming machine ain’t gonna f*ck up.

 

I *think* in the stock ecu there is a fan override that turns it on at a certain (higher) temp to cover a fan switch that has failed. Not 100% sure, it's a feature I do apply in Hondata/etc. though. I couldn't see any other reason for there to be a cooling fan output from the ecu if it was going to rely 100% on the fan switch to control it.

Have you tested the ect gauge sensor, per the service manual? Also, what ecu are you running?

 

You're right, there's pin A12, if I'm not mistaken, which is labeled FANC on some ECUs, but, I've had the fan switch fail and the ECU didn't override it in any scenario. I've been told is not active on any stock ECUs. I'm running the original P30 ECU. It's getting chipped soon so I'm definitely using this feature. I did not test the gauge temp sensor yet, but at this point maybe it's giving the dash a higher reading for a given temperature. I'll try to check the temperature with an infrared thermometer on both hoses, radiator, block and head to see if the reading is consistent with the switch. Are all gauge temp sensors the exact same spec between 92-00 cars, regardless of engine/model trim?

 

It seems weird because your temps appear normal with the car moving but that might be the next logical step. I looked at the catalog, the sender is the same for pretty much all 88-01 cars.

 

Consider that the head gasket is blown and allowing exhaust gases to enter the cooling system.

 

I drove the car a few miles earlier to borrow a thermometer. The temps stayed normal (about 1/4 of the way) during the drive. When testing the car at idle, the fan would kick in at the same spot (little less than 3/4 up) and temps would stay about halfway and a little over. The fan switched normal the first time (93 degrees C) but quickly switched off without reaching it's specified 88 C. Then it would turn on at around 94 C and temps would remain a little over 93 C even with the fan cycling 2 or 3 times, and the gauge reading a little above halfway which is not ideal. During the test, I've also measured 3-4 degrees C of difference between the upper and lower radiator hoses. With the gauge about halfway or just below, it would show around 80 degrees at the upper hose, so it seems that the gauge is not reading higher or anything. On the way back, temps would get just a touch below the middle (hotter than on the way there) so it's clear the system it's struggling to cool down after it has reached a higher temperature. I've switched the fan relay, and tested the fan with a jumper wire, it stayed on as long as the wire was there, so it rules out anything electrical. I'm thinking the thermostat is stuck halfway open, so the cooling system is not working as effectively as it could, hence the trouble cooling down from a certain temperature. It's also fooling the fan switch reading somehow. I'm ruling out the actual fan switch because the old one was known to be good, and the new one has the same behavior (tested it with both).

 

The engine was recently assembled with a new gasket. There are no signs of coolant boiling or filling up the overflow tank, and the car maintains its temperature while cruising. No coolant loss (checked it every time the car has been driven, before and after), and no signs of burning coolant from the exhaust. It seems unlikely, but I'll try and investigate it further.

 

there is a test for that....

 

So, I've replaced the thermostat, and to my surprise, the previous owner just left the top portion of an old thermostat, so until now the car was running without one. New thermostat in place, bled the system again as follows:
- Loosened the bleed screw
- Filled the radiator up until coolant was running through the bleed screw without any bubbles
- Tightened the bleed screw, started up the car with the rad cap off and heater on
- Air bubbles came out, let the fan cycle 3 times and the car was shut off.
The fan still cycled briefly, but more frequently. The first time above the middle and the other 2 with the needle right on the middle mark. Temps stayed right at the middle mark with the car stationary. I've let the car cool down a little, and went for a ride. When cruising, temps remained 1/4 of the way. As the engine is new, I've been doing the break-in procedure (short acceleration bursts in gear, varying rpms, then let it engine break). When at a stoplight or when applying load, temps would raise up at the middle mark, and
with engine loads increasing, it would rise up to about 3/4 of the way. Stopped it, jumped the fan and temps remained at about 1/4 of the way even with load. This is definitely not normal and I don't know what else to try. Maybe it's running lean because it's heating under load. The fan definitely doesn't like to stay on for over 15 seconds, unless it's connector is jumped. Still no signs of coolant spilling or overflowing the tank, and coolant levels remained the same before and after the tests.

 

Rent a block tester kit from AutoZone.