Britbike forum Forums British motorcycles in general British motorcycles in general Cylinder Leakdown Testing ....

I've always had a few questions to myself whenever I do a cylinder leakdown test on one of my bikes.

1) The leakdown percentage gauge has a "Green (OK)" range of 0 - 30%, a "Moderate Loss" range of 30 - 60%, and a "High Loss" of 60 - 100% marked on the gauge face. However, the few tech articles that I can find always say something like "5 - 8% is good, 20% means trouble, and >30% mean your engine is hosed." However, even on good-running, easy-starting engines of mine, I seldom read anything less than 25%, and engines that are in comfortable middle age and still starting and pulling good show 40 - 50% leakdown rate, which is in line with the markings on the gauge.

Besides that,

2) How can the same leakdown gauge work for any engine size and type? If I'm doing a test on my Honda 125, am I expecting the same "Good - Moderate - Bad" ranges if I'm doing a cylinder on my 428 Ford V-8? How is the orifice size in the gauge that is used for comparison to the air loss from the cylinder able to know whether I'm doing a 50cc tiddler or a massive truck engine cylinder .....

Very theoretical, I know, but .....

Lannis

The primary air supply needs to be maintained at 100 psi....then opening the second air supply valve will show air loss as a percentage of 100 psi..

Well, I'm doing that - I have a regulator on my air compressor tank that supplies 100 psi to the gauge.

But how do we figure out what is ACCEPTABLE leakage from a cylinder? Is it simply a relative value, such that I check all the cylinders and make sure that they are close together and that one is not going bad with a burned valve or whatever? Is it just something that I track over time to see how fast it might be degrading?

I don't see how I can hook it up to a V8 and see, say, a 10% leakage value, and hook it up to an old Brit and see a 50% leakage value, and have that absolute value mean anything to me. Maybe the old Brit runs just great with a bigger leakage percent and it's not time for any work?

It's a great tool; I guess I just don't know how to interpret it ....

And while we're on the subject - Compression tester. If I hook up a compression tester to any engine, do I simply expect to see atmospheric pressure (14.7 psi) X (Compression Ratio), and anything less than that indicates a bit of a problem?

My M21 is 4.9:1 compression. 14.7 x 4.9 = 72 psi and that indicates as-new compression? My A10 is 6.5:1 . 14.7 x 6.5 = 95 psi. My Firebird is 9:1, so I should see 132 psi on the gauge when properly tested on a new engine? If not why not?

Lannis

Those are very good questions and I've wondered the same about leak-down testers. I get up to 7-10% leak down using a NAPA leak-down tester with 100 psi showing before I connect the quick-disconnect on my newly rebuilt B50 (bored 8mm oversize) land speed bike and I'm told I should not get more than 4 or 5% that the "pros" get. But I've got a 92mm piston while the "pros" are measuring are measuring leak-down with a 75mm piston. I've always figured that surely with more perimeter of rings on the larger bore, I'd expect to get more leakage. And then, do all testers have the same size orifice? Most are probably made in China or India and who knows?

I recently bought a new one on amazon because my NAPA one is in Colorado and I'm stuck in Nova Scotia. The new one is definitely not American made, had a crappy % gauge, and leaked air at every connection. I got the air leaks fixed and the % gauge blows out on the first try. So I just bought a 100 psi replacement gauge, put it on and figured it would just show a lower number representing the leakage and on two motors I tested, I got like 2 or 3% leakage on motors that I honed in a bucket of solvent myself. So I wouldn't worry much about leakage of 10%, but at more than 25%, I would listen closely at the exhaust, the carb inlet, and the breather because if the gauge is reading correctly, that's a lot of leakage.

As for compression gauges, I've always wonder how I could get 175 psi with only 10 to 1 compression, especially since the piston doesn't even start compression until 65 or so degrees after bottom dead center. So one of the "pros" tells me it's because it gets "bumped" up above the theoretical 145 psi. But I'm an engineer, and that still doesn't make sense. I figure it's actually the momentum of the air due to the succession of rotations (I use about 10 kicks over tdc) that causes you to stuff in more air than the theoretical 145 because it doesn't seem possible to get more than the compression ratio would generate no matter how many times it goes down and up again. I guess we need another "professional" opinion, as I've only rebuilt about 50 motors.

Tom

The FAA has some basic guidance on this, as it is one of the tools mechanics use to determine an engines health during an inspection. One thing they do is specify a different orifice size depending on bore size. The gauge with the larger orifice WILL produce different results than one with the smaller orifice on the same engine.

The mechanic’s technique used to perform the testing can produce varied results. I would only use this tool for diagnostics, not as an absolute measure of how worn an engine is, or isn’t. An engine can still make full power even with terrible numbers.

Much like a compression tester, leak down is a relative thing measured over time unless of course you are getting 100% in which case some thing is drastically wrong but of course you would be able to hear & feel the escaping air.
As for compression testers when they have to be matched to the engine
Not so much of a problem with a 500cc single but a big one with the 25cc blue smokers I spend all of my days attenpting to fix.
Ideally the head / nozzle what ever you want to call the bit that screwed into the engine should occupy the same place and the same combustion space as the spark plug.
Otherwise you are changing the volume of the compressed charge.
And yes air has weight and therefore inertia when moving so you can force more air into the tester kicking the engine over very quickly than you would doing the same thing slowly.

And of course I doubt that any guage under a couple of hundred dollars would actually be all that accurate.

I would really only be looking to take the Leak down tester out of the tool box if I had a fault that required diagnosing. Hooking it up to a good running engine is like setting fire to money. When I do have occasion to use it ( On multi cylinder engines mostly ) I find watching the pressure gauge as you introduce air into the cylinder ( And comparing it to each cylinder ) is just as useful as looking at the percentage gauge after a given time.

Same thing for compression testers. Unless the figures are ridiculously low I'm more concerned with them being even across the board. Freaking out when it does not match a particular figure ( Even on a brand new engine ) based on running the numbers would be like pulling a super model and then leaving the Hotel room because you spotted some cellulite!

I've found it hard to keep the pistons at tdc. Crank wants to slip a bit and slightly open a valve. Putting it in gear and standing on brake didn't work.

Aircraft mechanics have a propeller to grab onto to hold the engine at TDC. Even then, the engine will try to turn if you aren't right at TDC. (I've done plenty of differential tests on aircraft.)

A socket and breaker bar on the drive side crankshaft nut should work fine. And it allows you to turn the engine back and forth a slight bit to help the rings find their best spot. I have seen values improve significantly by doing this.

Also, an aircraft tester uses 80 psi so you could try the lower pressure since you are more looking to pinpoint leakage and not so much an absolute value.

On the BSA, put it in high gear, then rocked it to TDC with a straw in the spark plug hole (which gets pretty close), then put a 30 pound weight on the brake pedal. When I applied pressure, the piston didn't move enough to move the rear wheel (and I don't know how much slack the drive train has in it, probably a bit, but does it have enough to crack open a valve without moving the rear wheel? Instinct sez no).

Thanks for all the info, guys. It pretty much all points in one direction!

Lannis

There is some slack in the primary and drive chains. Sounds like best practice is to lock the crank with a wrench.

I use a breaker bar with a socket on the nut on the rotor like Stuart does, but then rotate the crank just past TDC and support the end of the breaker bar on a solid block of wood in front of the primary, as that is the way it will try to rotate when you apply the pressure. But make sure you have the bike in a wheel chock or on a lift with the front end tied down or the the torque on the breaker bar will try to turn the bike over. If it rotates even several degrees, it doesn't make any difference as the valves will stay closed for over half of the stroke. And to tell the truth, you don't even need the leak-down tester if you can make a fitting out of an old spark plug with a quick disconnect to fit your air hose, and no, you don't need 100 psi. It the motor is in decent condition, it won't leak much, but if it does, listen closely at the intake, exhaust, and breather to determine if it's valves, rings, or blown head gasket.
Tom

I've gone through that process before, and it demonstrates one thing for sure - spark plug are built unbelievably tough, mechanically. Electrically, they're as delicate as a flower; if you gas-foul one, it may never work again no matter how you clean it, and a tiny carbon track will keep it from firing. But you'd better have a heavy hand hammer and a tough fixture if you want to break the center out of one!!

Lannis

Not everybody is as hardheaded as I am. For those, there is this item on ebay:

303793694881

Tom

The numbers of a leak down test aren't much to worry about below 20%, it is a tool that we rarely used when I was turning wrenches in motorcycle dealerships. The way Motorcycle Mechanic's Institute taught us in 1991 was to use 60-80psi and to figure out where the air is going over 8%. I've seen cylinders with over 15% leakage blowing past the rings provide plenty of compression. The test that is more telling of piston/ring/cylinder problems is a compression test, and the main thing to pay attention to is how fast and far the needle jumps on each compression stroke. You should get close to halfway to the desired pressure on the first compression stroke with good bore, good valve clearance, and usable rings. -Mike

Anyone use the TDC locating pin to prevent crank rotation?

Are thinking it might bend? i just lock the wheel with a bar through the spokes.......
This is the leakdown on my 650 land speed race Triumph....The engine was run to warm it up.I repeated the test several time....By comparison, a compression test was 120 psi.......most like to see less than 10 percent on a high performace engine...

Hi Lannis,

it is quite easy to get the center out. Put the spark plug in a lathe and cut off the peened steel part that keeps the porcelain center in place. A file or an angle grinder could be helpful as well if you had no access to a lathe.

Then saw the center electrode off, one proper blow and the center is out.



Cheers!

Phil

"Easy", says the guy with a lathe in his shop and the skills to use it. I can only dream ....

Hi Lannis,

to be precise, I had written 'quite easy'.
The point I tried to make is that without cutting off the peened part holding the porcelain center it is quite tough to get it out. As stated, in absence of a lathe a file or an angle grinder might prove helpful to remove this.

Sorry for highjacking your thread discussing the useability of lathes.

Although my lathe skills are very limited, I find such thing extremely handy. I had always smiled when my mates stated a life without lathe was possible but pointless.
Then I stumbled over one and bought it at a rather low price.



It's an old Myford ML 10, even Made in England.

After owning it for 5 years I second my mates' statement.

Cheers!

Phil

I got tired of hearing that on the BSA board, "just turn it up in your lathe." Like we all have one.

I've had a lathe for at least 35 years and now own 3. My first one, that I traded a magneto for, honestly was crap, a flat ways Craftsman made by Atlas with babbitt spindle bearings. but I still got a lot of work out of it. All that to say, you don't need all that good of one to learn on. They are out there; you just need a spot to put it in your shop.

And frankly, I do not see how a brit bike hobbyist can get by without one. I use it all the time and not just for machining. They are very useful for straightening and polishing things too.

Ok, I'll get off my soap box now. Thanks for listening.

I got a lathe about 12 years ago, bougt a 70 year old Davenport Model A.Cost $750.... I have a modern tool holder and a fair amount of tools....I am a crappy machinist but can accurately make all the spacers,adapters and other bits need for my modified bikes and trucks. My wife even does a bit of wood turning on it....For some a mill might be more useful but they need to be big enough and they heavier and a lot more money..
Back to the leakdown tests....In my opinion over 20% is too much. The engine will run ok and power loss won't be an issue with some riders...But that 20% is not pushing on the piston and just leaking past valves and or rings...

Lannis,
I used a hacksaw and cut around the plug at the indicated point, pounded out the center and filed two flats on the round portion.
I use it a base for a dial indicator, but might work for your purposes.
Adam

I'm no Joe Pyzinski myself, and sorry for not letting this rest but really, have you (and I'm sure you have) looked closely at much of the machine work on our old british bikes? A lot of the cycle parts aren't very nicely machined at all, and the crudity even extends into some of the engine internals! So, even a someone with a few things to learn should be just fine as long as they are careful. And what I really like is it totally expands the capabilities and gets one to thinking in a whole new way.

And back to leakdowns, 25% leakage ie: 60 over 80, is the limit for an aircraft engine.