When I had the oil changed in November, I was told they had to add
"quite a bit" of coolant to the reservoir. Since the car came from NM,
with a spotty service record, I had hoped that it was simply not topped
off ever. Now, three months later, with only moderate driving, I see the
reservoir is virtually empty again. I don't see excessive white smoke in
the exhaust, or smell antifreeze in it - although I haven't sniffed it
specifically for that, yet. The oil looks fine. I don't smell coolant
inside the car, either. It would be hard to spot a slow drip in this wet
Winter weather. How common are internal head gasket leaks in this
engine, as opposed to slow leaks in the reservoirs?
I took the car to a local shop today. They do free pressure tests (I
gave them $20 anyway). No leaks, including the cap. Unless I'm missing
something, it has to be the #@!@# head gasket, right? Assuming it's
going straight to the exhaust and not burning much, how safe is it to
leave it alone? Any suggestions?
go for at least a 20 minute run, then look at the coolant reservoir
while the engine's still running. [make sure you can see in, but don't
lift the hose above the coolant surface!!!] any bubbles? if so, it's
gasket. also, check under the car and look at the bottom of the timing
belt cover. any coolant? if so, it's the pump.
The pressure test needs to be performed in two places.
1) the cooling system, and
2) the combusion chambers.
the first is done in conjuction with the introduction of UV dye to the
coolant, this being detected on the outside with a black light. The
second consists of shop air being blown into each combustion chamber,
then the rad filler neck being then inspected for bubbles.
There is a third test that uses special paper test strips dipped into
the rad filler neck. These change color in the presence of combustion
gases in the coolant.
Your problem is NOT the rad cap. If it were, the reservoir would be
There is also the possibility of a worn water pump. In that case the
coolant would drip down the block under the timing belt cover and may
not be obvious. Findinig that leak means carefully inspecting the crank
Do not put this off. You WILL damage your engine, possibly badly. You're
supposed to check your coolant level every time you open the hood, or at
least once a week.
The procedure in the Honda manual online in the UK is to just use a
hand pump-operated pressurizer with guage, and see if the pressure drops
in the cooling system or cap. I don't know which they used at the shop...
What I'm going to do, assuming I can even afford the cost of a new
gasket done by the dealer, is to tell them to look for leaks, especially
at the water pump. If they find none, I'll have them replace the head
gasket. Shouldn't they install the new timing belt and water pump for no
extra labor charge (they will be providing the parts, not me) when they
do the gasket? I'm also having the front oil seals replaced as a
preventive measure. Thanks for your response. If you could answer this
post before this evening, when I drop it off, I'd appreciate it greatly.
On 2/20/08 6:20 AM, in article gWUuj.16648$FK2.1405@trndny08,
Seems like you're jumping the gun a little bit here. How low was it in the
reservoir? When you say reservoir do you just mean the overflow tank, or
In normal use, the system will settle out to where the overflow tank sits at
the "min" mark when the engine has cooled & the level will rise to around
the "max" mark when the engine is hot & has been running for a while.
I certainly would not spend money replacing a head gasket that has not even
been diagnosed. There is not enough info here to even say for sure that
there is a problem.
Where did you get this bit of misinformation? I've owned six Hondas and
the level in the reservoir never changes more than about 1", hot or
cold, summer or winter.
If yours changes that much, there's something radically wrong from what
the Honda engineers designed.
Sorry, misinformation is a strong word.
Volvos and Dodge products that I owned had pressurized reservoirs. The
coolant level in them went up and down with temperature, but the HONDA
isn't supposed to, since it's reservoir isn't under pressure at all.
Boiling hot coolant isn't supposed to take up much more room than cold
coolant, as long as the rad cap is keeping the pressure up.
Since about '92 most Hondas will self-bleed any trapped air because the
rad cap is the highest point in the system. When any trapped air is
forced out, the air is supposed to bubble through the reservoir and when
the system next cools, the level in the reservoir drops as coolant it
drawn back in. That's when the owner is supposed to add coolant up to
the MAX mark on bottle. The owner's manuals and the Helms service manual
all tell the same story.
However, none of this theory applies if there's a hole in a hose, the
head, the heater core, the water pump seal, or if the rad cap is shot.
sorry, that's fundamentally incorrect. coolant liquid most definitely
expands as it heats - for both systems. pressure doesn't stop coolant
expanding, it simply raises the boiling point.
but most antifreeze already has dye in it already. and dye does nothing
to diagnose a head gasket.
as said by someone else, the o.p. needs to take this vehicle to a shop
that /wants/ to fix it - clearly the current one doesn't. time to hit
I told them to look for signs of a leak. If they find none, they
will call me. How much should a dye test cost, and is it better or worse
than a cylinder pressure test? The car already had a system pressure
test, and passed.
you don't usually do a separate dye test for antifreeze since it's
already dyed. dye is more commonly used for differentiating different
types of oil leak [engine oil and mtf are similarly colored and hard to
tell apart sometimes] or for a/c systems.
the best thing at this time is to take the vehicle to someone that has
an interest in fixing the problem, not trying to tell you the vehicle is
too old to be worth bothering with. there are plenty of independent
honda specialists out there. find one and check to see if they have a
The two tests are meant to check for different problems. They are NOT
I'm assuming the "system test" means they checked the cooling system by
pressurizing it. If that test did not use dye, then the test was
incomplete. A dye test also will not pick up leaks that come from the water
pump, since those leaks take lots of time fo be seen from under the timing
If there is suspicion on a head gasket leak, you must do the cylinder
But it does. ALL materials expand with heat.
The reservoir level naturally rises and falls as the coolant heats and
cools. This is the purpose of the various seals in the rad cap, and for the
very presence of the (unpressurized) reservoir in the first place.
You don't have to take my word for it. Check your own reservoir's level at
dead-cold. Drive the car to full-hot, then have another look at the level.
It will be at least a half-inch higher.
Coolant is constantly traveling in and out of the reservoir with every
heat-cool cycle. It's a normal and necessary feature of a modern sealed
cooling system's operation.
my idle question is, how much negative pressure is required to suck
the coolant in. obviously, the spring keeps the in valve shut at zero
pressure, but when the coolant cools off, there's negative pressure in
the block which sucks the coolant from the reservoir. we assume the
pressure of the spring on the valve leading from the block to the
reservoir is what it's labeled as, but i just wonder how much residual
vacuum remains in the block when the spring shuts the input valve. for
no reason other than idle curiosity.
Motorsforum.com is a website by car enthusiasts for car enthusiasts. It is not affiliated with any of the car or spare part manufacturers or car dealers discussed here.
All logos and trade names are the property of their respective owners.