2 gal. compressor + compression tester = head gasket test! (1 Viewer)

This site may earn a commission from merchant affiliate
links, including eBay, Amazon, Skimlinks, and others.

Joined
Nov 22, 2005
Threads
34
Messages
208
Location
Minnesota
Hello,

This may be common knowledge, but I worked out a way to test the headgasket failure on a recently aquired '95 4Runner with the 3.slow engine (3.0 V6) which ought to work on any vehicle. I have been trying to find a cylinder leakdown tester in town but half the parts stores (Napa, Checker) have no idea what a leakdown test is (much less the tool), or they don't have it in stock (Advanced Auto). I ran a compression test, but it isn't necessarily going to show what's wrong other than compression isn't right. Soooooo, I stared at my compression tester (from sears, about $20), stared at my 2 gallon mini air compressor from Menard's, and realized I could put the two together! The best part is, I didn't even need to buy anything more. I figured I'd share this since most people probably don't have a leakdown tester, but many people probably do have an air compressor, and a compression gauge is much easier to find in stock and cheaper ($20 vs. $75+) than a leakdown gauge. I wasn't feeling too good about paying $75+ for a tool I'd only be able to use for specific tests. A person can just about buy a 1-2 gallon air compressor and a compression gauge for $75, and there will be many other uses for them (at least for the air compressor).

My main goal was to confirm if I could pressurize the cylinder with air and push air back into the cooling system via the suspected cylinder. I figured I had a blown headgasket and didn't really care what leakdown rate was, I just wanted to know that it leaked. Any air escaping the cylinder into the cooling system = very bad, right?

The gauge on my compression tester unhooks via standard air tool quick disconnect fittings, so I took the gauge off and just used the hose to hook up the air compressor. I made sure the cylinder was at top dead center (to close the valves so air can't escape back through the intake or exhaust system). The tank on the compressor was already filled up to about 10PSI, so I didn't turn the compressor on. I hooked it up as-is, and immediately I heard glug glug glug glug gluuuuuuuuuuuug through the radiator (cap removed).

I then turned the compressor on and let it run until the tank was up to 25PSI, then I shut the power off and just let it drain. The line pressure never got above 15, and the tank drained within about a minute in a half. I could hear water rushing through the intake, coolant tubes, and glug glug glugging out the radiator. I think I can call that a blown headgasket. Because of how little air volume exists inside the cylinder, especially at TDC, I can't see any reason to need to buy a huge 60 gallon $$$$ compressor for this type of thing.

Please note that an actual leakdown gauge would certainly be better, especially if your headgasket is only barely leaking. I figured mine leaked, I just wanted to make absolutely sure. The 4Runner blew about 1 gallon of coolant out the overflow reservoir and actually blew the cap right off. There's a terrible miss, theoil cap has milkshake residue on it (oil still looks good though), and there is copious white clouds of smoke coming out the exhaust within about 3 seconds of starting up a cold engine (no, I wasn't stupid and ran it very much after the gasket blew--and it went fast). The suspect cylinder would collect water droplets on the end of the compression tester and give compression readings anywhere from 160 to 230PSI, while all other cylinders were around 165 and just got oil on the end of the compression tester hose/fitting. Oh, and I used a mini wand LED light and was able to stick it into the spark plug holes to see the tops of each piston. The suspect cylinder had a cleaner piston than the others. Lastly, the spark plug from this cylinder was cleaner than the others, and turned a but rust colored (likely from water sitting on it and causing rust stains on the metal).

I'm no expert, but I figure ANY air getting pushed into the cooling system from the cylinder is VERY BAD. If you try this yourself, I wouldn't recommend going over 80-100PSI. That's the range I've read on most websites that give basic instructions on doing leakdown tests. I didn't have to go anywhere near this high. My pseudo-leakdown test also won't be super helpful for small leaks or diagnosing between intake or exhaust valve leaks or ring leaks. In those cases, a real leakdown tool with an incoming line pressure gauge and a cylinder pressure gauge would be best, to tell you how much pressure is being lost. Again, I just wanted to confirm either air was getting into the cooling system or it wasn't. Also, make sure the cylinder is at top dead center, or you'll be losing air out the intake or exhaust valves accidentally, and the cylinder won't pressurize enough to test the cooling system. If you do try this, make sure all spark plugs are removed. Also, I've read that higher PSI such as 80-100 will rotate the engine (even quite fast), so you may want to put a breaker bar on the crank bolt and fasten it to the frame or something. At 25 PSI, the engine never even thought about rotating. Lastly, make sure the battery is disconnected before doing any of this (standard precaution for most any kind of work).

I know the 80's have their issues with headgaskets, but man, I really HATE :mad::crybaby: the 3.0L V6's. I've owned 6 4Runners/pickups with this engine, and 4 of them blew headgaskets. Two of them had already been on their second set. And despite a special service campaign (aka recall) on most 1990-1995 V6's, Toyota says my '95 isn't covered (bastards). The 3.0L is notorious for blowing the gaskets at the rear cylinders and the theory is because there is a cross-over exhaust pipe pumping a lot of heat into those cylinders. Combine that with the fact toyota was forced to stop using asbestos in their gaskets in 1988 (first year of the 3.0L), they had some poor gaskets.

If my procedure is a little flawed, please feel free to chime in. It was so easy though, I feel dumb for not figuring this out years ago. :doh:

Oh, one last thought. If you use a compression gauge and it's like mine, it will have a schrader valve at the very end where the spark plug hole fitting is, and another schrader release valve up by the gauge itself. My air compressor hooked up to the end of the guage's rubber hose where the gauge itself went, and the air must flow backwards through this connection to the spark plug fitting connection and into the cylinder. The schrader valve that is found at the spark plug hole fitting should be removed since it's not meant to leat air flow backwards through the hose stolen from the compression gauge. Leave the other schrader valve in (the one by the gauge itself), because you don't want air escaping out before it even gets to the cylinder. You can also skip the compression gauge assembly and make your own hose assembly assuming you have hose, a fitting to go into the spark plug hole (napa), and a quick release air line fitting to connect to your air compressor hose. I basically just stole the rubber hose assembly from my compression gauge assembly, but you can make one too. If you thought you could thread it in without this short length of hose, you could get a spark plug hole fitting and a screw in schrader valve and hook your air compressor line right up to that. My air compressor hose is that curly yellow stuff, so I'd never be able to get it that close to the spark plug hole.
 
Last edited:
If my procedure is a little flawed, please feel free to chime in. It was so easy though, I feel dumb for not figuring this out years ago. :doh:


nope sounds good, you have determined for sure that there is a leak from that cylinder into the cooling system, considering the 3.0's history that is a pretty solid HG diagnosis.


you have owned 6 3.slows? are you a glutton for punishment?
 
Yes , I have owned six:eek: 3.slows. I admit I have a problem, and I am addicted to 4Runners and pickups. Please sign me up for the 12 step program.

I love the pickups and 4Runners, especially 4Runners, but it is VERY hard to find them with 22RE 4 cylinders. My beloved 1990 4Runner does have a 22RE and a 5 speed (best of both worlds in my book), and I've hung on to that one for 10 years now. I don't drive the 1990 in the winter though. This vintage rusts horribly. Mine is rust free, and I intend to keep it that way. I typically drive other vehicles during the winter. My '85 pickup (with a 22R) is now mainly reserved as a beater to be used for occasional trips hauling dirt, class 5, stuff to the dump etc....

I think this will be the last 3.slow that I own.... Yes I own and 80 now too, but I think one cylinder head is easier to manage than two.
 
did you try this on all cyclinders or just one? i would wonder how they might differ
 
good job!

been meaning to do this for a while and was always too laz... errrr... busy :D to do it. (Also too scared to find out there is something wrong with my HG. Ignorance is bliss...) Now, I just might do it the next time I have the plugs out... thanks for the writeup!


100 Psi: think about the area of the cylinder, guessing 5 sq in (no idea), we're talking 500 lbs of force... no wonder it could move...
 
I did try it on all cylinders because I wanted to see what is "supposed" to happen on a normal cylinder. Air will leak out is what will happen, but not through the coolant. Cylinders aren't perfectly sealed systems, and there will be some leak down even on a new engine. The actual leakdown test gauge's purpose is to see just how bad the leakdown rate is. I didn't care what it was, just wanted to see if I could push air into the coolant. Air did NOT move into the cooling system on any of the other 5 cylinders. I started with the passenger side head (remember, I'm talking about a 4Runner 3.0L V6, not an 80) and worked my way over to the driver side head.

First I ran a regular compression test on everything. Cylinders 1, 2, 3, 4, 5 had compression within 165 +/-3, with cylinder 6 being anywhere from 160's to 230's. I then removed the gauge off the tool, removed a schrader valve off the hose, and ran the pressurizing test (I probably shouldn't call my test a leakdown test), and only cylinder 6 pushed air into the coolant. I never used a PSI higher than about 25 on any of the cylinders. In my case, the headgasket leak was obvious enough and I probably didn't need to go up to 80-100PSI. I have heard of shops doing leakdown tests on barely leaking headgaskets only to not "find" any problems. I'm sure there can be cases where the headgasket only leaks when the engine is hot, or possibly only when cold. Mine just plain leaked, and like I said, it billowed clouds 3 seconds after after starting a cold engine.

Once I tested the remaining cylinders, I was able to get the engine to rotate at about 25PSI once or twice. It doesn't spin wildly at this PSI, but it did rotate maybe once. My guess is, the bad cylinder's leakdown test never caused the cylinder to pressurize enough to rotate the engine since air was obviously leaking into the cooling system.

12 hours of labor later, and I am just now getting to removing the lower intake manifold. The 3.0 has a bazillion little hoses, tubes, and wires EVERYWHERE!
 
should be easy to improvize a leakdown test. Fill a smallish tank with a known pressure -say 25psi. Connect to cylinder. Open tank. Time how long it will take to go down to -say- 10 psi. Repeat on other cylinders. Note differences.
 
Your test is exactly what I used everyday for testing the leakdown on new/used 4 stroke outboard motors. The only difference is that my tool had a regulator to take the 300psi shop lines down to 25psi and I had to use a clamp on the flywheel so that it wouldn't spin the engine over when i did my testing.

It was then up to me to listen for where the noise was coming from=
Intake = Stuck open/damaged/carboned open intake valve (remedy was to use combustion chamber cleaner first before teardown)
Exhaust = Stuck open/daamged/carboned open exhaust valve (remedy was to use combustion chamber cleaner first before teardown)

If leak down was present but no sound was found from either the exhaust or the intake, then I had to start removing anodes from the blocks cooling water passages to find it, which if the unit was under warranty we would replace the block or cylinder head, or both.

This was done on an outboard engine which is generally similar.

So I think your test is pretty sound and awful cheap. As long as you pressurized the tank to the desired pressure and timed it like said in above posts, you should come to an average leakdown per cylinder. My rule was always having cylinder compression within 10% of each other, and leak down was no greater than 15% per cylinder. You should time it over 60 seconds per cylinder and you might think about putting a lossless valve in-line so that you can watch the gauge on the tank.

That's my 50cents. Good work!:flipoff2:
 
I remember a couple of years ago I wanted to do a leakdown test on my 22RE before I tore it apart, to decide if I should get into the bottom end or not. I was just doing a PM HG replacement, and adding a few other goodies. I debated about buying the tester, I finally caved in and bought the $70 one from Summit. I figured I'd rarely use it, if ever again.

I've since used it at least 5 times. Really glad I have it now.

Not that I have a problem with your method, but I also like that the dual-gauge tester will tell you HOW MUCH leakage you have, help you make a decision on overall engine health.
 
I am sure I will eventually buy a real leakdown gauge, however, I made the fatal mistake of assuming I'd find one in town to purchase. The headgasket just failed on Monday and I had a 3 day weekend this weekend (plus next weekend) so I got right into things and wanted to have it torn apart by today and put back together by Sunday next weekend.

I did get it all apart today, and sure enough, the infamous #6 head gasket errosion from a coolant passage right next to the cylinder wore away and cracked. The fire ring was still intact, but there was pretty much no gasket next to it. Interestingly enough, the 2 headbolts on the backside of the affected cylinder head felt a tad loose compared to all the other headbolts. Toyota said my VIN # wasn't covered in the recall, but I'll be darned if this isn't the exact same failure that happens to "affected" VIN's. I may frame the driver side headgasket and send it to Toyota as a Christmas gift.
 

Users who are viewing this thread

Back
Top Bottom