: Cylinder leak-down test procedure (with percentage gauge)



km7648
01-08-07, 09:54 PM
I've got a little smoke on start-up and before doing the anticipated valve stem seals, I wanted to be sure I didn't have any other problems.

Well I bought an engine leak-down test kit (with poorly translated instructions) but haven't been successful using it.

Do any of you know how to test using a setup like the one in the picture?

The left gauge show PSI and the right shows percentage loss. While it sounds easy enough, I'll be darned if I can find the right procedure anywhere (and I've looked).

So far, my engine is either perfect or totally pooched <grin>

Ranger
01-08-07, 10:01 PM
What are you testing for? Head gaskets or burnt valves? What engine?

km7648
01-08-07, 11:12 PM
It's in a sand rail but from a 1996 Seville - a Northstar.

While I think the motor is OK, I wanted to be sure the smoke is the valve stem seals that I expect.

Ranger
01-08-07, 11:23 PM
I don't see how that is going to tell you anything about valve stem seals. That is more for testing compression, rings & valve seats. You could also use the separate hose for a cylinder pressure test to check for bad head gaskets.

km7648
01-09-07, 12:54 AM
I "know" that I need valve stem seals and I don't need the leak down test to tell me that.

I just want to be sure that I don't need anything else.

I could still use help re. the proper usage of the leak down tool with a 2nd percentage gauge.

zonie77
01-09-07, 01:10 AM
Found these quickly.

http://members.boardhost.com/MustangSteve/msg/1160701958.html

http://vmaxoutlaw.com/tech/leakdown_tester.htm

http://www.tavia.com/08015_instructions.html

clarkz71
01-09-07, 10:41 AM
To be accurate, you need to regulate the supply air to 100 psi, then a 10% reading on the leak-down tester is actually 10 %. A leak-down tester will help determine if an oil consumption/smoking problem is valve guides/seals or rings. If you hear air coming from the oil fill hole in the valve cover, and the percentage of leakage is high, you have a ring problem. Likewise, if air is heard from the throttle body, intake valve, exhaust pipes, exhaust valve. You want to see all 8 cylinders within 10% of each other. To use this tester, you need to have the piston at TDC on the compression stroke.

km7648
01-09-07, 10:42 AM
Thanks Zonie77

My problem is with the style of gauges that I've got. The left one is a standard PSI gauge but the right one shows percentage of leak.

I'm just trying to figure out how to calibrate and use it.

The only links that I've found (and those you sent) show a PSI gauge on the left and right. Kinda' thinking I should buy one of those anyway...

clarkz71
01-09-07, 10:45 AM
The left gauge is the supply air, set the regulator to 100 psi. The right gauge is the percentage of leakage. IE: 50 on the gauge would be 50% leakage. (bad) You want to see no more then 20 or 30%.

clarkz71
01-09-07, 10:54 AM
I don't see how that is going to tell you anything about valve stem seals. That is more for testing compression, rings & valve seats. You could also use the separate hose for a cylinder pressure test to check for bad head gaskets.

Larry, a leak-down tester puts air in the cylinder at a lower psi then direct shop air (120 psi or more). You can hold your finger over the fitting. It measures the percentage of air leaking past the rings & or valves. If the leakage is 20% or less and you have an oil consumption problem, you know it's the guides/seals. If it's high, 40% or higher, the rings are suspect. You'll also hear the air coming from the crankcase.

Ranger
01-09-07, 12:27 PM
Understand that Clark. Just don't understand how it relates to valve stem seals. Shouldn't be any pressure there, right?

clarkz71
01-09-07, 12:34 PM
Right, if you don't hear air from the crankcase. If leakage is in specification (20-30%), and you have an oil consumption problem, it's the seals & or guides.

km7648
01-09-07, 01:13 PM
OK, I contacted the manufacturer and indeed, the instructions are in fact incorrect.

For those following this thread, the correct procedure is:

1) Bring the cylinder to be tested to top dead center, valves closed.
2) Thread the hose into the the spark plug hole.
3) Set the incoming air pressure to the tester at pressure 30-100 psi
4) Slowly turn the regulator on the tester until the percentage gauge reads "0". This calibrates the gauge.
5) Connect the hose from the cylinder to the tester.
6) Read the percentage of leakage on the gauge that previously read "0" when calibrated. This will be the percentage of leakage.

The instructions had step #5 before #4 and it made no sense to me.

*Thanks to those that offerered assistance.

clarkz71
01-09-07, 03:52 PM
Your gauge must be set up different then mine. Let us know what readings you get.

km7648
01-15-07, 11:26 PM
First, I did a compression test and for all cylinders got reading of 195-200.

Then I did the leak-down test and per the gauge, got readings right in the center of "low" leakage for all cylinders. The gauge indicates that the middle of "low" is about 30% OK for their gauge but lousy by other 100psi gauge tests.

I did each of these tests while the engine was at about 60 degrees so I understand these readings would be even higher/better if the motor was warm.

The attached photo shows the gauge dial face. You can read the numbers if you squint just right <grin>

Then to be paranoid, I cranked up each cylinder to ~80psi and listened for any leaks, hisses, gurgles in the coolant etc. and - nada.

Sounds like the motor is A-OK and I should go ahead with my valve seals and not tear down the motor?

zonie77
01-16-07, 12:12 AM
A common problem with N*s is rings sticking in the piston grooves because the engines are often babied. There is a GM procedure of pouring solvent in the cyls, letting it set, cleaning them out, change the oil, then test (by driving). Possibly this engine is a candidate?

We don't hear of valve stem seals going on N*s but it is possible.

km7648
01-16-07, 02:32 AM
Thanks.

I was doing the valve seals due to a puff of smoke that I get on startup - a classic valve stem seal problem?

Doesn't a ring problem give you the puff on hard acceleration or no smoke but ongoing oil consumption?

clarkz71
01-16-07, 06:16 AM
Remove the oil cap & listen, if the rings are leaking you'll hear air from the crankcase.

zonie77
01-16-07, 02:27 PM
Thanks.

I was doing the valve seals due to a puff of smoke that I get on startup - a classic valve stem seal problem?

Doesn't a ring problem give you the puff on hard acceleration or no smoke but ongoing oil consumption?

Yes, I was suggesting the ring cleaning for the high leakdown. I should've clarified that.

myronf
10-13-07, 06:34 PM
I bought the same gauge set as KM7648. Makes no sense to me. When the right gauge is zero'd there is only about 10# on the outlet side of the regulator. I thought you wanted much higher pressure. Also, when there is no pressure on the inlet the right gauge reads75%, just like KM's. Comments?

97Concours1
10-14-07, 11:38 AM
I bought the same gauge set as KM7648. Makes no sense to me. When the right gauge is zero'd there is only about 10# on the outlet side of the regulator. I thought you wanted much higher pressure. Also, when there is no pressure on the inlet the right gauge reads75%, just like KM's. Comments?

Your right "Cylinder Leakage" gauge has been damaged by too much pressure. It should read 100% with no inlet pressure. It is only about a 30 psi gauge. You must always turn the regulator out BEFORE connecting the inlet air hose. If the regulator is turned IN, the right gauge will peg due to too much pressure and will be damaged. If you try to use the tester now you will not get correct readings.

97Concours1
10-14-07, 12:11 PM
I've got the same tester from Harbor Freight. The instuctions are wrong. I think the problem everyone is having is they are trying to set the tester while it is connected to the cylinder adapter hose (as the instuction INCORRECTLY state). The tester needs to be disconnected from the cylinder adapter hose to set the pressure. There is a valve that blocks the outlet while it is disconnected. This allows you to set the "cylinder leakage" gauge to zero. THEN you connect it to the cylinder adapter hose and take the reading.

FYI: A trick for finding TDC of a cylinder (ON THE COMPRESSION STROKE and not the other one). Put a long tube on the cylinder adapter hose that comes in the kit. Screw it into the cylinder you are testing. Put the other end of the long hose in your mouth. Crank the motor over, BY HAND, until you get air coming out of the hose. Keep going until the air tapers off and then begins to vacuum. At that point, crank back and forth slightly to find the middle between a slight pressure and a slight vacuum. You're there. Works like a charm.

FYI2: My leakage numbers were: 1-43%, 2-55%, 3-35%, 4-40%, 5-40%, 6-30%, 7-40%, 8-40%.

Ranger
10-14-07, 12:53 PM
Why not just drop a 1/4" dowel rod down the spark plug hole. It will rest on the top of the piston. Then you can watch it rise and when it stops your at TDC.

km7648
10-14-07, 01:53 PM
Here's a quick update for everyone to let you know how things turned out. I know I appreciate hearing the Happy Ending after all the advice.

By all accounts (leak down, compress etc.) the motor is fine. It's been running great in my sand rail since last year. The leak-down tester that I bought and asked about is "different" that the other gauges but worked fine.

It turned out the the PCV valve setup on this engine was unrestricted. When driving, it would suck too much oil vapor into the intake manifold where it would collect. At idle or after sitting, the warm intake would allow the oil to drip into the cyclinders where it would be burned (on startup or prolonged idle) causing the smoke.

There is one major problem with that gauge set: If you connect unregulated shop pressure to the gauges, it will damage the inlet gauge by spinning it beyond its maxiumum range.

97Concours1
10-15-07, 12:18 PM
Why not just drop a 1/4" dowel rod down the spark plug hole. It will rest on the top of the piston. Then you can watch it rise and when it stops your at TDC.

That's not a bad way to do it either, but it's a 50:50 chance that you are between the compression/power strokes where you need to be. You could be between the exhaust/intake strokes where the valves are open slightly. With the hose you know for sure, and then you can go right to the leakage test because the adapter hose is already in place.

The dowel rod is also difficult with a non-hemi type engine where the plugs are at an angle. My first experience with using the leak gauge was on an Aerostar van. I was by myself trying to find TDC on the cylinders while turning the engine over from underneith at the harmonic balancer. It was impossible to stick something down the cylinder, turn the engine over at the same time, and feel for TDC. In the van you can't even see some of the plugs you are working on. The hose method worked great in that situation.