leak down test results

[llm4]

I been fighting an annoying stumble/miss. I have done all the maintenance possible (adj. valves, ip timing, new injectors, etc) with little or no change. Decided to re-check compression and do a cylinder leak down test. Adjusted valves again and here are the results:

cylinder/compression/leak down %
#1/395/10%
#2/395/10%
#3/395/12%
#4/395/34%
#5/375/24%

On-line manual say max leakage for valves is 10%, rings 20%, and total is 25%. All cylinders have similar amount of noise passing rings (assume 10%, each cylinder).

Cylinders #4 & #5 had a noticeable amount of noise coming out of turbo (passing by intake valves, I thought). I decided to pull the head and have a valve job thinking the intake valve seats were shot.

After the valve job (grind seats and valve faces and exhaust guides), I put the engine back together. I drove it about a month and the car still has the same symptoms, weak cylinder. I took another set of "leak down" and compression readings, no change??? Some how the air pressure is still getting into the intake manifold and escaping from the turbo inlet. To verify, I put a blue rubber glove over the turbo inlet, started the leak test and the glove blew up like a balloon very quickly.

Question: Any ideas on how this could be happening? Can a prechamber be cracked and cause this? Or do you think its a cracked head?



1983 300CD, motor is a 617 with 165,000 miles.

thanks!

[leathermang]

Please Explain EXACTLY ALL the procedures you did to accomplish your leak down test.
Do not worry about using up all the electrons in your computer....
or how long it will take me to read and evaluate it...

[josha37]

Please say you had the head checked for cracks

[leathermang]

Originally Posted by josha37

Please say you had the head checked for cracks

LOL

[Chas H]

Cylinder heads rarely crack into the intake tract and the prechamber is upstream of the intake valve seat. I think your valve job was done incorrectly.

[josha37]

Rarely or not it is irrelevent. When the head is off it should be checked for cracks.

[Chas H]

Originally Posted by josha37

Rarely or not it is irrelevent. When the head is off it should be checked for cracks.

Yes, the head should have been checked for cracks.
Do we know for sure that it was or was not?
Tell us about the last head you saw that was cracked into an intake port.

[josha37]

About 1 in 5 chevy heads is cracked from intake valve to exhaust valve. What are you getting at?

[Chas H]

Originally Posted by josha37

About 1 in 5 chevy heads is cracked from intake valve to exhaust valve. What are you getting at?

For starters I'm getting at this a Benz diesel forum. But thanks for sharing.

[barry123400]

Where the head and block checked for flatness? If one intake valve is open with the other one closed for the leakdown test either adjust the valve to close the other or pull the rocker.

Then do a new leakdown test. Even if you had a crack in the one cylinder in the head casting I would not expect a simular crack in the adjacent cylinder. Unless the crack is between cylinders in the head. The glove will expand if their is leakge between the adjacent cylinders with one intake valve open. If you move to the other cylinder for a leakdown test then the other cylinders intake valve may then be open. You can pretty well establish if this is the senario visually before disabling its valve .

If closing the other valve stops the glove expanding you are either leaking across the head gasket or a crack between cylinders in the head .If not stopping after disabling the valve in the other adjacent cylinder then the intake valves are not seating on those two cylinders in my opinion. The chances of cross leakage between those two cylinders and both intake valves either not seating or having cracks in the areas of both their seats are pretty much about the same odds as winning the lotto. Unfortunatly it is already fairly likely that leakage between the two cylinders is there. Testing is your answer.

It would take two cracks at least in my opinion otherwise. One between the cylinders. Most likely is the head gasket is not sealing well perhaps if not the valves or a crack between the cylinders.

Hope I got this right as I found it a mouthful. For what you describe with the glove simple air pressure applied may enable the test as well.. I would have to think this through though. The glove as a detection test was a good ideal.

It could be a combination problem but I doubt it. When I used to get my heads back from a shop. I would by habit put some very light oil around the valves with the head upside down and no cam installed if overhead cam of course. If the light oil went away on one I knew it was not seating. Machine shop work seems to have become sloppier with time.

[llm4]

Thanks for the information! One factor that seems to lean towards a crack(s) is the symptoms/readings are exactly the same after a valve job and new gaskets. Not sure what the machine shop did other than grinding seats and installing exhaust valve guides. I know they did not remove the pre-chambers. I assumed the head was ready for installation.

Prior to testing, I had the valve covers off and made sure the cam lobes were not against the rockers and piston on TDC. While preforming the test (with 34% leakage), I tapped the intake valve several times hoping to seat the valve and get better readings, It didn't make any difference.

Another clue. After I removed the head, when removing the intake valve from the #4 cylinder I noticed there was a tremendous amount of carbon (or crusty oil) build up on the backside of the valve and intake port. It was so bad it was almost blocked! This was not normal and I was convinced the valve may not have been seating properly due to build up and leaking by. Maybe this is an indication of the problem.

Barry123400, I think I will try your idea and close #5's intake valve and retest #4 cylinder. Also, may pull intake/exhaust manifold and see whats going on inside the port while testing.

This was my first post on this site and amazed by the response!

Thanks again guys

[leathermang]

Except you did not answer the first question.
You are giving ' results'....
but have not described your procedure....

[llm4]

Sorry forgot that part...

Here is the test procedure (from memory):
- started the car and let it warm up
- adjusted the valves
- removed glow plugs and used those holes
- started on #1 cylinder and worked rearwards
- did compression test first on all cylinders then leak down test
- rotated to TDC on each cylinder ensuring both valves were closed
- using leak down tester, applied 100 psi to cylinder
- removed radiator cap and listen for noise at radiator cap, turbo, exhaust pipe, and oil fill cap (with valve cover on)
Also followed on-line manual.

Please let me know if I made any mistakes or need more info

Thanks!

[leathermang]

Ok, now we are cooking with gas...
nice that you used 100 psi... makes percentages lots easier...
but am interested in how you determined those percentages you reported.... (?)
Was there some time function involved in your procedure ?

[barry123400]

If there was a lot of buildup in an intake port on the head. It almost sounds like some of the ajacent cylinders exhaust gasses where being processed through that chamber causing the buildup.

Your description of not running right may be an understatement. A crack or non sealing intake valve might produce simular results but not as bad as you had. The crack if there where one would be too small. If the valve sealing was that bad the cylinder possibly would not fire. Plus before doing the valves you had fairly good compression. It almost also sounds like the problem may have been lurking there for a long time.

Then again I know very little about this area. If I were a working mechanic my intuitive senses would be far better honed.So always take any of my thoughts with a grain of salt so to speak.

You cannot beat actual hands on day to day experience in almost any field. This would be a good one for our Mr Hunter and other working mechanics to comment on.