'97 Honda Civic LX, 5-speed, 211k miles

I left my car sitting for two weeks while I was away on a work trip. Checked my coolant overflow tank when I got back, and it was a little low (maybe 2" below the "full" line). Filled it and didn't think much of it.

About two weeks later (today), I checked it again. It was again about 2" low. Filled it, and this time I checked the radiator. Instead of being up to the top of the radiator, the level was about 1" below. You could see the "fins" inside the radiator, and the coolant was a little above the level of those fins.

Rented a coolant pressure tester, and pressurized the system. The tester made a good, but not 100% effective, seal against the radiator. You could hear a slight hiss from where the tester was not sealing 100% on the radiator. It pressurized it enough to see that the system is holding pressure, and there are no other visible leaks.

The system already had UV dye in it (I replaced the radiator when it was weeping last time - about 10,000 miles ago), so I looked around very carefully with the blacklight. I checked the upper and lower radiator hoses, the various hoses behind the engine block, the heater core hoses, and then I crawled under the car and checked the weep hole on the lower timing belt cover. No apparent leakage of coolant on any of these places.

What could be causing it to become low? At this stage, should I worry about it? Or should I just top it off and call it good? Car hasn't overheated on me, and I changed the timing belt / water pump 30k ago, so those should be good.

I took off the overflow tank hose and placed it in a glass of water, to check for air bubbles (which would signal a head gasket issue). No bubbles whatsoever.

I'm kind of lost, because I have to take this vehicle to New Hampshire on Saturday and I don't want to be worried / stranded while on this two-week trip.

 

Need to check for exhaust gas in your system.

 

Forgot to mention that I did a block test with the liquid that changes color when the exhaust gases hit it, and that was fine. (Luckily).

 

could it be a bad radiator cap and some fluid is escaping that way?

 

Did you shine the black light around the heater core itself under your dash board in the car?

If it's a pinhole leak, it probably won't leak enough to escape the heater core box so may not show anything as the coolant likely to evaporates inside the heater core box long before weeping out any cracks or seams.

 

I will check that tonight, actually. Thank you for the suggestion.

And I changed my radiator cap ~15,000 miles ago, with a Genuine Honda one, so I'm doubtful that the cap is the issue (although it is still a possibility).

 

Are you checking the level when the engine is cold or after it's up to running temp while idling?

 

For the overflow tank, I've checked both times. When I found I needed to fill it was after checking it while cold. I also checked the radiator itself while bone cold, and that's when I had to add the 1" of coolant.

I've also checked the overflow tank after the car had been running, and it was fine both times I checked.

Like I said, I'm taking this to New Hampshire Saturday so I'm hoping this doesn't turn out to be an issue.

Shined my black-light in the passenger footrest area and up behind the dash and didn't see any signs of coolant.

 

My daughter's '98 Civic EX does the same thing. Cold = low (radiator) and overflow, At running temp the level is fine (radiator) and overflow is fine. When the car is at running temp the coolant system is fully pressurized and expansion brings the coolant level to the top of the radiator and overflow at the proper levels. Maybe this is just a Honda thing..I've had other makes of cars and don't recall there being a difference when checking cold or hot.

 

How do you know the radiator level is fine while hot? You can't pop the cap without scalding yourself (or risking it). I assume that the radiator level is fine while running, because it hasn't overheated on me at all. The overflow has remained at the proper level, but when I checked the radiator itself (with the car completely cool), there was about 1/2" of air before getting to the radiator fluid.

Maybe I am just being paranoid, and there's nothing to worry about. Radiator fluid does expand when hot, so maybe that's the explanation right there.

 

Sure you can. Get it up to running temp and while idling use a rag to loosen the cap (about a quarter to half turn) but, do not completely remove it and leave it that way for a minute (or until pressure is released) to allow overflow (probably won't though and if it does it should go in overflow tank). After a minute fully remove the radiator cap and check to see if coolant is flowing properly at the top of radiator...flowing water (but not spilling out of radiator) is a good sign the water pump is working properly.
.
Obviously, do not put your head/face over the radiator cap when initially opening it. Here's a video... the engine is not idling (but is warm/hot) but, it's the same procedure. Also, use a thick rag and not paper towels:

 

Ive had this on two different cars.. the civic and my toyota. Both had old reservoir tanks, and the cap had warped over the years. Allowing the coolant to evaporate! Once I replaced the tank, no more coolant loss! Just a thought to check out. See if that cap on the overflow reservoir tank is tight.

 

That makes sense. Next time I go to the junk yard I'll pick a couple caps up and try it on my daughter's car.

 

In my opinion, because the leaky tester confounded proper interpretation of the result, you should repeat the test with a non-leaky tester. If you cannot do a pressure test without loss of pressure at the radiator seal, perhaps you found the problem.

 

Seems highly unlikely. The reservoir cap is designed to be loose to accommodate constantly changing coolant volumes in the reservoir.

 

FYI coolant does not evaporate, unless you only had nothing but water in there than I guess we found the problem.

 

ECG might help

 

I have a '97 Civic with the D16Y7 and 216,000 miles. Water pump / timing belt was replaced about 30k ago, and radiator was replaced about 20k ago (if that). Radiator cap is also about 20k old, and is genuine Honda.

Today I was sitting in traffic at a stop, and I noticed the needle going up to about 3/4. I turned off the A/C and put on the heat full blast, and it came down immediately. Pulled over and parked in the parking lot.

I first checked the upper and lower radiator hoses, both of which seemed to be equally hot. The coolant overflow tank was full, and I opened it and left the expansion tube hose in there to see if there were any bubbles (which could indicate a bad HG). No bubbles present during this test. I also revved the engine to make sure.

I noticed the cooling fan for the radiator wasn't coming on (as it was overheating) so I checked the under-hood 20A fuse (good) and also swapped the cooling fan relay with a known good relay (the A/C condenser fan) and turned the car back on. The cooling fan was cycling on / off, and I thought this was my culprit. Swapped the relays back to their original locations, and the cooling fan once again turned on like it should. So I don't believe it has anything to do with the relay.

I let the car cool down for about two hours and opened the radiator cap. The coolant level was below the fins (not sure how far below, but I had to add a decent bit of coolant to get it all the way up to the top). Filled it up and drove home with the A/C full blast, trying to get it to overheat. It stayed right on the 1/2 mark the entire time.

Last summer, I had a similar problem. The car started overheating after a very long drive to Chicago. I first suspected the radiator cap, and that's when I replaced it with an OEM Honda one. Then, I pressure tested the radiator itself by hooking the upper / lower hose openings together. It lost pressure, so I figured I'd replace the radiator. Looking back on it, though, I think it was more an issue of a crappy O'Reilly pressure tester (I never have been able to get them to hold pressure like they should - the needle just drops verrrrry slowly).

My Honda mechanic friend thinks it could be an intermittent sticking thermostat, since I've never replaced it since I've had the car (70k ago). He also suspected the radiator fan switch (sticking out of the thermostat housing) and suggested I "jump" it to be sure it functions properly.

Any ideas? I am supposed to drive to Wisconsin tomorrow and really don't want to take my dad's Prius (although that's my last resort). Could it be possible that it has a REALLY slow leak (like, started last summer slow) and when it finally gets to a certain point, exhibits the signs of overheating?

Any reason that the fan wasn't kicking on after the 1/2 temperature point had been reached? After I topped the radiator off I have had no overheating, although it's only been about 15 miles.

 

Unrelated?

https://honda-tech.com/honda-civic-d...t-why-3243069/

 

Touche. Could be the exact same issue reappearing. Almost exactly 2 months ago, though. Slow leak / loss? Crazy thing is I can't for the life of me figure out where. It there any harm in just running it and just making sure the coolant stays topped off?

 

Is my edit^ accurate?

 

Related threads merged.

 

You may very well have the same issue as Ron pointed out being you haven't located the leak.

However, I would still be concerned about the fan not kicking on even when the temp gauges is over half. The fan thermo switch is easy to replace and if you jump it and the fan runs (as the relay works) then you know it's not working properly. As to why it's not working properly is questionable as it might be an air pocket. But it's still something to keep in mind especially if you thoroughly bleed the system and it still doesn't kick the fan on.

This is the fan thermo switch.

 

Hate the bugginess of trying to reply when this is done. *bleh*

 

That's the thing - the only time the fan didn't kick on was when it was exhibiting the overheating symptoms (temp gauge was rising to 3/4ths). At 1/2, it kicks on and runs fine. If there were a significant air bubble in the system, could the fan thermo not kick on the radiator fan, due to the bubble?

Is there any harm in jumping the two terminals in the female plug that connects to the fan thermo switch, causing the fan to run constantly? Any way to do this that's less ghetto than a bent paperclip? I feel like that would fall out too easily.