Help me interpret compression/leakdown! - PeachParts Mercedes-Benz Forum

#1 11-02-2006, 01:17 PM

My car started running very rough last tuesday. I took it to a good indy and they did a compression and leakdown test. Could somebody help me interpret the results?

Compression test: # 1-210 #2-95 #3-170 #4-225
Leakdown Test: #2-90% Leak-intake valve

The shop wants to do a valve job on the head, but I want to be as sure as possible there is nothing else wrong with the engine before I give them the Ok. Are these results enough to be reasonably sure I don't have lower end work in my future?

cheers,
dan r.

EDIT: Car has 174k miles, original head, replaced timing chain and tensioner at 90k

david s poole · #2 11-02-2006, 06:21 PM

what they are saying makes perfect sense.

Engr693 · #3 11-02-2006, 08:56 PM

Quote:

Originally Posted by Dan Rotigel

Compression test: # 1-210 #2-95 #3-170 #4-225
Leakdown Test: #2-90% Leak-intake valve

Notice the difference between the 4 cylinders. That 95 reading is out of spec. all four cylinder pressure readings should be close to each other in value. Not exact but close. Say within 15% or so (don't know the exact % that is correct)

Did they do the compressions test again after squirting oil into the cylinders? If not, by doing so it could bring the 95 reading up. If it does, I would suspect you may have compression ring problems on the piston in that #2 cylinder. Number 3 is a little off too, but I would be more worried about #2.

Merkey · #4 11-02-2006, 10:11 PM

Yep, makes sense.
A compression test is a quick start, but only tells half of the story.
When you do a leakdown test it tells you pretty well the other half.

They would be able to hear the hissing of the pressure escaping back into the inlet manifold, so telling them where the leak is.
If it was a compression ring or holed piston, the pressure would vent into the crankcase....
If it was a blown headgasket into the cooling system... the pressure would go into the cooling system... bubbling would be heard.

And a Exhaust valve would be heard in the exhaust!

I hope its just a valve.
Good luck!

Matt L · #5 11-02-2006, 10:45 PM

Quote:

Originally Posted by Engr693

All four cylinder pressure readings should be close to each other in value. Not exact but close. Say within 15% or so (don't know the exact % that is correct)

20% is considered acceptable for most engines. Closer than that will result in a smoother and more efficient running engine. You won't get all of them to be exactly the same.

#6 11-03-2006, 08:10 AM

Thanks all,

I didn't understand how the mechanic could tell if it was the upper or lower end that needed workd-now I do. What I don't want is to have them open up the engine and say 'oops, this is going to be another 3k, sorry!' I'll call the shop today, talk it over with the owner and most likely go ahead and have the work done. 2.5k for a valve job, assuming no valves are shot (most likely one will be!), they also put in a new tensioner. Sound reasonable?

cheers,
dan r.

t walgamuth · #7 11-03-2006, 08:18 AM

they are supposed to be within 10% of each other. is number three leaking past the valves too? if not then the cylinder is suspect. seems most likely to be valves i suppose.

is this the 16v you are taliking of?

the compression values are very high for a gasser. the combustion chambers must be full of carbon.

tom w

cjlipps · #8 11-03-2006, 10:44 AM

Ditto Tom's comments. As a rule of thumb you can use the compression ratio to ballpark what ideal compression should be. A 10:1 CR with standard atmospheric pressure of 14.7 psia would yield 147 psig if everything was good. Not an exact representation by any means but it serves to illustrate that 210 psig for a gasoline engine compression reading is horrendously high. And even if your gauge is faulty and #s 1 and 4 actually have normal readings, #3 is bad and #2 is dismal.

#9 11-03-2006, 10:51 AM

Yep, its the 16v. I'd been running the car as it is (very rough) for 4 days previous to the test (first apt. they had). Perhaps that was the possible source of the carbon-Not sure what else it could be? Before this the mechanic suspected low compression on a cylinder just from hearing it start, but the idle was smooth and i was content with leaving it alone for a few months.

The shop owner said that the head gasket could be blown between the two leaking cylinders, causing leak through that way.

T-Is there any way I can find out of the cylinder itself is ok before I give them they go ahead to do the headwork?

cheers,
dan r.

cjlipps · #10 11-03-2006, 11:25 AM

Dan, I don't know of a way to quantify your problem but you could get closer by doing a compression test after squirting some engine oil into the cylinder. If the oily test is significantly higher than the dry test, the rings are suspect. If there is minimal or no change, it's the valve(s) only. A leakdown test would also point you in this direction but with a major intake valve leakdown it can be difficult to tell if there is also leakage past the rings into the crankcase. Listen at the oil fill hole for this. As far as a headgasket leaking into the next cylinder, the large disparity in the compression readings between #2 and #3 make this unlikely. Or at least even if it is a problem, it's not the only one. If it was the only problem, 2 and 3 would be essentially the same reading. One last thought. I had a small block Chevy that idled a little rough after a rebuild. The compression readings on one bank of cylinders was 145 across the board. The other bank was 120, 110, 80 and 45. The mechanic who refurbished the heads put the wrong valve springs on all the exhaust valves on that one side. (Chevy had two different spring heights possible for this engine) It was a simple matter to pull the valve cover and change those springs for the correct ones. Fixed the problem. In other words, you might get lucky and have a broken or collapsed valve spring or two. They can check that without head removal. Good luck and I'll stay tuned.
Chuck.

mpolli · #11 11-03-2006, 12:31 PM

My first impression was the same as Gary and Chuck, that you need a wet compression test also. There is another thread ( http://www.peachparts.com/shopforum/showthread.php?t=169145 ) on here right now just like this. He was low in 2 middle cylinders, did the top end, only to find out it was rings in #2 and 3 (4 cyl).

Mike

#12 11-03-2006, 01:52 PM

Thanks All,

I just called the shop, said 'go for it' and the owner said 'ok, first thing we do is double check the previous diagnostic and check out the rest of the engine.' When I drop the car off on monday, I'll make sure he will do wet leakdown (this shop is quite good, nothing but MBs and very busy). Hopefully things will be over in two weeks. Till then, I'm biking to work =). It will be interesting to see the outcome-I urge everybody to enjoy the voyerism of somebody else having a valve job done on a 16v-second documented case on this board i think.

cheers,
dan r.

Merkey · #13 11-04-2006, 02:01 AM

Just for interest...
Last week I did a compression test on my son's 1988 300TE.
It has 68000 miles on it.

1/ 180
2/ 180
3/ 185
4/ 185
5/ 195
6/ 190

I dont know what the compression ratio is on the 12valve 3.0 I6 but I thought these numbers were good bearing in mind the milage!!

cjlipps · #14 11-06-2006, 05:16 PM

Man, apparently my rule of thumb for compression values is based on flawed logic. I know the method for calculating compression ratio based on volumetric measurements would point to a direct relationship between engine compression and atmospheric pressure but apparently real-world measurements don't bear this out. I guess the main thing is that the values are near what the mfr. says they should be and aren't too far apart from each other. In other words, we still have a problem with the center two cylinders here.

Tomguy · #15 11-06-2006, 05:26 PM

Quote:

Originally Posted by cjlipps

Man, apparently my rule of thumb for compression values is based on flawed logic. I know the method for calculating compression ratio based on volumetric measurements would point to a direct relationship between engine compression and atmospheric pressure but apparently real-world measurements don't bear this out. I guess the main thing is that the values are near what the mfr. says they should be and aren't too far apart from each other. In other words, we still have a problem with the center two cylinders here.

My gauge agrees with your numbers. I was expecting my heads to give me a 9.2:1 compression ratio but apparently I have 10.0:1 - I got 146PSI on my #6 cyl (the only one I measured) after swapping to hotter heads. Either they may have higher compression or the gauges are off (or both?)