bryan.0831

So I was performing a leak down test with my civic today..
I started to take off my very old spark plugs and when I got to Cylinder #3 this happened...
SOB broke on me!! Gave me a mini heart attack. LOL


I got that thing out easily.. Just had to literally run to Sears and got me a screw extractor.. the only problem is I can't take the extractor off the spark plug thread housing thing.. oh well.


Anyways I ran the test twice on all the cylinder just to make sure..
Here's the result:
Cylinder #1,#2, and #3 has 18% leakage.. pretty much all looks like this


But Cylinder #4 has 14%..


I have a 2005 Civic EX with 161,000 Miles, blew the head gasket twice. Never done anything to the bottom. I did the test hot/warm, right after a nice drive. I used a Snap-On MT324 Tester. I have heard that this is very sensitive compare to other testers.

Is 18% an OK result for an engine with 161,000 miles? Need info yo!
I have read that below 20%, you shouldn't worry about it too much.. but at above 30%, that's when you have major problem and time to rebuilt.

ezone

Did you do a compression test that led you to do a leakdown test?

What were the results of the compression test?

Or is there a specific problem that led you to use this test?

bryan.0831

I'm waiting for my Compression Tester. I figured to do this first since I have it.
I'm just planning to run 6-7 psi of boost. I already have everything except the ARP Head Studs. I have read/heard to do this sorts of test to determine the condition of my engine.

ezone

Ok, I never pull out a leakdown tester without prompting by a lousy compression test. I gotta have a reason.

A decent tight engine should give 10% or less after the rings are broke in.
An engine with Total Seal%u00AE "gapless" piston rings could give 3% or less. (Not even sure if those exist anymore, LOL)
The miles you have, I'd SWAG that less than 20% is probably fair, but could be better.... I'd say to do the compression test though, see if they are fairly even.

HTH

bryan.0831

Thank you sir. I will do a compression test and see from there.

ezone

Stupid question: What brand of spark plug was it that broke off?

bryan.0831

NGK. Not sure what exact part# but it's just stock ones.
I replaced it with Denso IK22.

bryan.0831

Hey ezone. I just did a compression test. They seem high. It seems like I have worn piston rings. The engine was hot when I did the test. I tried 4 cranks and 7-8 cranks, same reading.
Results: Dry
#1 - 210
#2 - 210
#3 - 210
#4 - 220

Results: Wet
#1 - 235
#2 - 235
#3 - 235
#4 - 240

ezone

Those are mutually exclusive assumptions. Unpossible. Worn rings will lower compression, not raise.

When I do a compression test, I keep cranking until the gauge stops climbing. Plus a few extra hits to be sure. I watch how much it climbs on each hit, but I don't stop until it is definitely maxxed out.

I expect well over 180 on almost all of Hondas engines from the last 15-20 years. Over 200 is good.
Your differences are well under 10% between cylinders.
I'd say it's ok.

Ring wear? Probably some, it has a bunch of miles. Does it smoke? Does it use a lot of oil?

A compression test cannot check oil control rings. Ever.

HTH

bryan.0831

Oh yea, my stupid assumptions. Lol. I was assuming because when I added a few squirt of oil the reading is higher, but when its calculated percent of different its not that bad.
The gauge climbs every crank. Can't really see how much since Im peeking through the crack of the hood to see the gauge stop climbing. I'd say like 30+/- not really sure.
I just checked the oil. It's till on the top hole of the stick. It's been 2,000 miles since last oil change. It doesn't smoke or anything. Just a P0172 code that is running rich. Don't know what cost it but it's off topic.
So I think I'm good to run 6-7psi. Yay!

ezone

Compression numbers should always go up with oil added. Oil makes the rings seal better.
You normally would do this if you had a low cylinder on the regular dry compression test, squirt oil in and rerun the comp test to see if the low reading was caused by rings not sealing...
You didn't seem to have any one cylinder that jumped waaaay up on the wet test, so I'd assume the rings are roughly equal.





Got freeze frame data?
The usual cause is tight valve adjustment, if the engine appears to run ok.
I look at MAP voltage@idle, too high is a biggie clue.
There could be other causes too, such as EGR sticking open.

HTH

bryan.0831

The engine appears to run ok, idling at 600-700. I adjusted the valves two weeks or less ago at .006 intake and .008 exhaust. (I have a Stage 1 Cam) Very slight drag on the feeler gauge. My LAF sensor is fairly new, like bought this November.
When I reset the ECU, the code doesn't come on again until a few days for driving, like 75 miles more or less.
Here's an idea..


If the Map Voltage @ idle is high, how do i fix this?

bryan.0831

I don't know why I took a picture. I could have use Print Screen. Haha.

ezone

Whut tha hell is that?
That causes more questions than answers for me.
You got a kinda generic set of data parameters, and only a few are usable.

There is no MAF on a stock engine,
Is that really the MAP instead?


The O2 values are off the chart like this was taken on a cold engine. Is "Bank 2" supposed to be the rear sensor? There is no bank 2 on a stock engine.

The "fuel trim"...far too negative. Is that LTFT or STFT?

The cam alone, I really don't know anything about. Is it a dual lift deal with stock profile for the low end?
I wonder what the data would look like with a stock cam.
A bunch of overlap would kill vacuum and cause rich codes with a stock PCM. Actually, almost any deviation from stock could affect how the stock PCM thinks.

bryan.0831

LOL. Sorry. I just Plugged in the AEM FI/C for a minute to see what's going on. I don't really use it I keep it unplugged. I just wanted to see the voltages but I never really run with it, just saving it for future use.
And yea, I just started the engine and ran it for a few minutes. Sorry again.
The fuel trim is calibrated because I changed the injectors to a Type-S. So STFT.
I should've never posted that. LOL

ezone

I guess you have to ignore the rich code then, or get it all dialed in good once it is all set up.
Fat injectors and a cam should cause richness that the computer don't like. I can see it, anyway.

bryan.0831

Yea man. I've been ignoring it for a while now. Thanks for the help!