Leaking head gasket confirmed by home CO detector
Ray L
lots of carbon monoxide in the air I collected from the
overflow hose.
I put a carbon monoxide detector with digital display inside a
plastic bag, with a piece of paper towel to absorb any
coolant that got in. I then put the coolant overflow hose
into the bag, sucked the air out of the bad, and slowly
released the radiator cap until all the gas was expelled into
the bag, plus a little coolant. When I plugged in the CO
detector, the display showed about 410 ppm, and
eventually the alarm went off.
I tested the CO dector with clean air plus an antifreeze soaked
paper towl, and it read 000, showing that it was the gas from
the radiator that set off the alarm.
Someone else can now use this technique to diagnose
a slow head gasket leak at home.
2001 Subaru (made in USA) Legacy Brighton Wagon,
well under warranty.
How can I make sure the dealership doesn't just
put a new gasket on without also checking for bad
mating surfaces? I don't want it to leak again
after the warranty is up.
OOR...@usmc.ret
> How can I make sure the dealership doesn't just
> put a new gasket on without also checking for bad
> mating surfaces? I don't want it to leak again
> after the warranty is up.
Very nice, and I think the result is valid and defensible. I usually
breathe through a hose and do the fingernail test. 8^)
I would have the dealer certify what's been done, claiming to have
checked the mating surfaces. Odds are that if the surfaces are found to
be bad a year from now, you'll be able to hold his feet to that fire.
Good luck, and thanks for the tip.
.
<<<jmarsince2.at.yahoo>>>
Ray L
<OORAH!@USMC.ret> wrote in message news:3D444D83...@san.rr.com...
Delaership: We can find nothing wrong with your car.
Me: Did you test for carbon monoxide in the radiator?
Dealership: No.
After I arrived home, I left the engine running, and
opened the hood. I looked at the coolant level in
the recovery bottle, and the surface was moving a
bit. I opened the lid, and saw a small stream of
bubbles bursting at the surface.
I took a turkey baster, sucked some of the
air from within the recovery bottle, and
squirted it into the plastic bag containing
my home CO detector. The reading went
from 000 to 999.
hmm...
Bill Putney
Ray,
That's a cool idea. If you don't mind, I'm going to post that on a
couple of automotive forums - will give you credit as Ray L since I have
nothing else to go by.
It'll be gratifying to you when you see atrangers posting advice about
using your idea to detect a cracked head or bad head or leaking gasket.
I won't tell them about the air bubbles you saw, because, undoubtedly,
someone will say "Well, you don't need a CO detector if you've got
bubbles coming up". However, there are certainly situations where it
may not be so obvious (i.e., the leak only occurs under certain
conditions that you can't duplicate), and your little trick can verify a
hunch.
One suggestion for proving your method: I know you got 000 reading from
just plain antifreeze. But to really reach a more scientific proof, you
should capture air in the bag from a car that is known not to have a
leak to confirm that you get 000 reading.
Bill Putney
(to reply by e-mail, replace the last letter of the alphabet in my
address with the first letter of the alphabet)
Ray L
Bill Putney <bpu...@moonstzr.com> wrote in message
news:3D471D06...@moonstzr.com...
> One suggestion for proving your method: I know you got 000 reading from
> just plain antifreeze. But to really reach a more scientific proof, you
> should capture air in the bag from a car that is known not to have a
> leak to confirm that you get 000 reading.
Yesterday, I did that, armed with a turkey baster
to sample the air, and got 000 from a different car
that had recently been run. I also tested my car after
it had sat for an hour, and got 000 from the
overflow tank, but gas taken from the top of the radiator
cap gave a 500 or so reading. The machine takes
about 5 minutes to respond.
For newer cars, with the pressurized expansion
tank, you just have to allow the engine to cool
a bit, then open the pressure cap and
suck out a sample.
There are test kits available where you collect
a gas sample and bubble it through a test liquid,
which quickly canges color if CO is present.
This CO detector, by the way, can operate directly off 12VDC,
so I can take it to the dealership and show them what
is going on, right inside my car. Other models may
require a DC/AC inverter to use in a car.
These things only cost about us$40. A turkey baster
is about us$2.
good news:
2nd dealership has a set of new head gaskets on order.
He said they have seen a lot of them go bad, and there's
a new design to replace them.