Leak Down Test Results

[mikeob1]

I am trying to diagnose small, dull pops in my exhaust. The pops are random and occur at idle and while accelerating through rpm range.

My vacuum guage is erratic at idle varying from 15 to 10.

1. I don't have vacuum leaks.
2. Ignition components are new
3. warm compression figures are (from cylinder 1-6) 140,135,120,125,130,140
4. Carb is clean.
5. warm valve lash is .08 intake and .014 exhaust.
6. Timing is set.

So I bought a leakdown tester to further explore the varied compression readings.

There are 2 pictures of the guage for each cylinder located at:
http://picasaweb.google.com/mikeob1/LeakDownTestResults/

The method i used to conduct the tests.
1. Crank piston to TDC.

2. Verify piston located at TDC by inserting pencil in spark plug hole and feeling top of piston.

3. Attach Leak down tester hose to spark plug hole.

4. Attach Leak down tester to hose.

5. Open regulator on leak down tester until needle gets to black line (far right).

6. Take two pictures.

7. Repeat for cylinders 2 - 6.

Results

The intersting thing is the results were similar for cylinder 1, 2, 5, and 6.
The needle was steady in the 0-10% leakage range.
The regulator on the tester was at 40-50 psi.

Cylinders 3 and 4

The cylinder 3 and 4 regulator on the tester was 75-80 psi, much higher than the others.

The needle was not steady on these tests. The needle bounced between 3% leakage and 12% leakage in cylinder 3.
Cylinder 4 also bounced, but not as much as cylinder 3.

Questions
Why would the input pressure be higher on these two cylinders?

#3 and #4 had the low compression readings (120 and 125). Could this be a head gasket leak between the two?

Do I have enough evidence to warrent pulling the head and inspecting?

Are there any more tests I could do?

Is my method correct?

There are two pictures attached...the first is for cylinder 6 showing a low input pressure. The second is for cylinder 4 showing the high input pressure. Check out the link above for all pictures.

 

[Aussie FJ40]

sounds like a valve on the way out

 

[malcb]

5. Open regulator on leak down tester until needle gets to black line (far right).

I am not sure the technique is right here, I try to make the input pressure the same of all the tests, 80 to 100 psi set using the regulator and then measure the leakage.

I might be wrong but it seem to work for me. Anyone have a better method?

PS - which black line?

 

[KLF]

Unless your leakdown tester is different than mine, the input pressure is supposed to be 100 psi every time you do the test. The leakage gauge (right gauge) is calibrated for this.

 

[mikeob1]

I am not sure the technique is right here, I try to make the input pressure the same of all the tests, 80 to 100 psi set using the regulator and then measure the leakage.

I might be wrong but it seem to work for me. Anyone have a better method?

PS - which black line?

Thanks for the input guys. I am also curious to know from others if there are other methods. I did what the manual says. Attached is a picture of the black line it set to.

The instructions that came with the guage indicate to be sure the regulator knob is fully counterclockwise. Connect to compressed air source with compressor regulator set from 7 to 100 psi. Never operate with air pressure higher than 100psi.

Then the guide says to put the adapter hose in the spark plug hole.

Then connect the tester to the hose hanging out of the spark plug hole.

Then it says, "Turn the Regulator Knob clockwise until the Leakage Guage reads "0" athe the end of the yellow band." This is the leakage guage far right.

Then it says, "You can now read the amount of leakage on the Leakage Guage as a percentage loss."

Also, as a sanity check, before i started the tests, i rotated the crank shaft to watch the tool's indication as I completed the powerstroke, exhaust stroke and intake stroke. And sure enough the needle reads 80% leakage when my valves open and returns to near zero when they are closed.

 

[ken_79-fj40]

I have the same leak down tester, and following the instructions my 2f that runs very well but burns about a quart of 20w50 every 4-500 miles had 5% leakdown pretty much across the board. I thought that was odd so I did a little research on the tool and discovered the instructions are printed wrong. Supposedly you set the pressure to 0 on the right gauge before you attach the tester to the hose threaded into the spark plug hole. Doing it this way I found about 50% leakdown, which made more sense. Using my mechanics stethoscope I was able to hear the leakage through the dipstick, but not though the intake or exhaust, so in my case I think my oil rings are severely stuck. I did a compression test on mine right before the leakdown and it's at 145-155 across the board pretty much. I would try it again this way and see what you find. You need something to listen for where the air is escaping too. I used to use a piece of fuel hose stuck in my ear before I bought the stethoscope.

 

[pb4ugo]

I think input pressures should be the same, to be consistent, but I really can't say since i've never done as cyldr leak down w/a real tester. I'm more old school, put the cyldr at TDC and apply air pressure and listen for air leakage. Put your ear to the exhaust, carb w/butterflies open,adjacent spark plug holes, oilfill or dipstick tube and check the rad for air bubbles. You will hear slight air leakage in just about every location. You need to determine which has more air leaking, for lack of a better description. You can also do a dry and wet comp test to help determine if it is the rings or in the valves/head. A shaky vacum needle points me towards a valve train problem or worn cam.

 

[H2o]

this is a harbor freight tool isn't it... I bought an air conditioning manifold gauge set and I had the same problem with the instructions telling me to do the opposite of what was supposed to be done. This made me thing twice next time about buying HF for such tools. A good set of instructions can make the world of difference for the weekend warrior.

 

[kulangot]

Sounds like a sticking valve to me or maybe a valve spring problem - that's just a guess. Look up vacuum gauge diagnostics and try to identify the problem via your vacuum gauge.

If you think that you have a leak between two cylinders, you can put compressed air in one and the listen in the other. Sears sells a compression tester with a air fitting on the end so that the gauge can pop and off. You can thread that into the cylinder while it's at TDC compression and then see if you can detect air getting into the next door cylinder using a piece of hose as a stethoscope.

I said earlier in Ken's post that I have learned several valuable lessons from cheap tools and I think when it comes to timing lights, compression testers, vacuum gauges, torque wrenches etc you need to get decent stuff - at least sears and not HF. If you do use cheap stuff, then cross check it with another (but then if your going to buy 2 then why not just buy one good one). Test your diagnostic tools when possible. You might be surprised.

You can build a leak down tester yourself with stuff from Home Depot and or your junk bin. Search the web.

 

[KOOTENAY KRUZER]

In the aviation industry the leak-down tester has 2 guages marked 0-100 psi. The left one is the supply pressure, you set the regulator to 80psi. The right one is the cylinder pressure, you take the reading from that one. A good cylinder will have a reading like 75/80. This is done at TDC on the compression stroke. If you have an unnacceptable leak rate you listen for airflow - if you hear it out the exhaust pipe your exhaust valve is not seated or the seat or valve is damaged. If you hear it out the intake (carburetor) it is the intake valve and if you hear it out the breather (PCV valve hole?) it is the rings. If it is a valve, especially the exhaust valve, tap the valve stem or run the engine and try it again, stuff can get stuck under the valve that will blow away next time the valve cycles with pressure in the cylinder.

I think even with the tester you are using the input pressure would have to be the same on all cylinders for the test to mean anything.

 

[malcb]

Thanks for the input guys. I am also curious to know from others if there are other methods. I did what the manual says. Attached is a picture of the black line it set to.

Looks like your into some more plug removal and engine rotating. I would be interested in seeing the results if you redo the test.

The way it was explained to be was the regulator sets the pressure, the air flow or leakage with no restriction will be 100% (valve open), with no leakage 0%, the leakage dial is actually measuring the pressure in the cylinder as opposed to the pressure supplied by the gauge. I have found anything less than 80 psi gives poor results and 100 psi is about right.

 

[mikeob1]

... so I did a little research on the tool and discovered the instructions are printed wrong.

Thanks for the help, guys. Ken, I thought the same thing and tried to call HF support line today. The instructions to make practical sense.

I agree and would think the pre-set to 0 is done before connecting to the cylinder. (When you know there is zero leakage).

I will try the tests again tonight and see what I find.

I do have my dipstick out, pcv line disconnected, carb blocked open. So once I know i'm doing the test correctly I'll be able to listen. I did listen with the valves open just to know what it sounds like.

I am ashamed to say, but this is the second one of these I bought from HF. I would love to buy a nice once, but sears didn't have one by my house. Maybe I'll check online or Napa.

 

[bsevans]

Thanks for the help, guys. Ken, I thought the same thing and tried to call HF support line today. The instructions to make practical sense.

I agree and would think the pre-set to 0 is done before connecting to the cylinder. (When you know there is zero leakage).

I will try the tests again tonight and see what I find.

I do have my dipstick out, pcv line disconnected, carb blocked open. So once I know i'm doing the test correctly I'll be able to listen. I did listen with the valves open just to know what it sounds like.

I am ashamed to say, but this is the second one of these I bought from HF. I would love to buy a nice once, but sears didn't have one by my house. Maybe I'll check online or Napa.

Don't forget to remove the radiator cap and check for headgasket leaks into the coolant.

 

[mikeob1]

Tried to Retest - no results

Hi Guys,

I tried the retest by setting the input pressure at 100. Then opened the regulator to zero the needle. Then connet.

I'm convinced the tool is junk. After doing that, the needle is jacked again. It doesn't zero out. That is the second and last one I'll buy from HF.

I like the avaition tool idea where both guages are numbered pressure guages. That would be fine.

I'll check around and see if I can find one. Until then...I'm going to run it with the little pops.

Because honestly, If I'm taking that head off, then the whole damn engine is coming out. I've been looking for an exuse to pull it anyway.

 

[grant5127]

If I'm taking that head off, then the whole damn engine is coming out.

 

[malcb]

Hi Guys,

I like the avaition tool idea where both guages are numbered pressure guages. That would be fine.

Too true - why they aren't like that is beyond my comprehension.

If you still have the old unit can you take the gauge from that and put it in the second one...... then again your not likely to mess around with it again from the sound of things.