Mysterious compression readings

[sterl]

Hello All.  I had the cylinder head rebuilt on my late 67 250SL (117k miles) because it had very varied compression test results. Also, a leakdown test on some cylinders showed valve leakage (and some bore wear). The compression readings (hot) were 1:127, 2:140, 3:140, 4:137, 5:150, 6:160. These results were repeatable.

The head had never been off before and the surface was very straight. It needed new intake valves and guides and 2 intake seats (cylinders 3 & 4) because the valves had buried themselves into the seats. The exhaust valves and guides were OK. The rocker pivots had wear crowns on them which were ground off. All valves were lapped in. There was no excessive carbon build-up in any combustion chamber. Bore wear was approximately 0.006" on all cylinders.

Got it all back together and set valve clearances cold (0.003" intake, 0.007" exhaust, between cam and rocker). After I got it running and warmed up I did a (hot) compression test with the following results: 1:131, 2:141, 3:130, 4:132, 5:135, 6:135. They are obviously more even but why have some of them come down making the average value significantly lower? Same compression tester (SnapOn), same technique. I was careful to time the cam correctly. Any ideas anyone?

Thanks, Mark.

Billericay, Essex, UK
Late 67 LHD 250SL

[JimVillers]

Mark .... Did you do the compression test correctly?  Wide open throttle, crank 7-8 cycles until the maximum reading is recorded.  It looks like you did not open the throttles.

Jim Villers
190SL, 230SL 5-Speed, 190E 2.3-16 Kompressor, MGB 5-Speed

[ctaylor738]

Thicker replacement head gasket?

Chuck Taylor
1970 280SL #14076
Falls Church VA

[ja17]

Hello Mark,

Hello Mark,

Cleaning the carbon out of the combustion chambers during head work may lower the compression a bit. In addition grinding the valves make them sit a bit higher in  the head, slightly increasing the volume of the combustion chamber in the head. Cranking the engine during a compression check also squirts unburnt fuel in the cylinders breaking the oil seal normally found in the cylinders. Try running the car for a few days and then re-check after taking the fuel pump fuse out so fuel is not injected.

Joe Alexander
Blacklick, Ohio

[sterl]

Thanks for the suggestions everyone. Here's a bit more info. I got the engine hot again this morning and checked the vacuum. 20 in Hg rock steady at 1050 rpm idle. The head gasket was genuine MB. Fuel pump fuse was removed. Yesterday I did not open the throttle during the test, so I repeated it this morning with the throttle wide open, engine hot, fuse removed, and got the following results after 9-10 cranks: 1:135, 2:149, 3:136, 4:135, 5:144, 6:139. Yesterday I did a couple less cranks so that may account for some of the differences too. Also the battery was running a bit low towards the end of this set of tests (tested in numerical order) - does cranking speed affect the readings?

Here's a quick summary of readings do far:

Before head rebuild:            127, 140, 140, 137, 150, 160
After rebuild, throttle closed: 131, 141, 130, 132, 135, 135
Throttle open, more cranks:     135, 149, 136, 135, 144, 139

I'm still surprised that number 6 has come down from 160 to 139 after all the work, and that 3 and 4 are no better or slightly worse despite their buried intake valves! Today I'll recheck the cam timing and re-adjust the valves clearances. Then, as Joe suggests, I'll use the car for a while then repeat the tests.

Thanks again,
Mark

Billericay, Essex, UK
Late 67 LHD 250SL

[glennard]

Hi, Just a question--.  You said the 'rocker crowns' were ground down.  Are these the ball studs?  Why grind?  New ones are a couple bucks.  New rocker arms, too.  Just wondering.

[sterl]

Glennard, new rocker pivots (or ball studs) are 25 Euros each, but I've read here that it's OK to regrind them when they wear, which the machine shop did for free and they agreed that they would be fine like that. You can no longer buy the upper ball stud itself, you have to buy the lower part too.

Mark

Billericay, Essex, UK
Late 67 LHD 250SL

[Shvegel]

hi,
A couple of things at work here. One is that when you clean the carbon out of the head or off the piston tops you will have lower compression. the other is when you machine a valve and seat you will actually move the valve up(up with the head on) which will give your head greater cumbustion chamber volume which will result in lower compression as well. It is quite probable that a healthy cylinder will have lower compression after a valve job.

[Benz Dr.]

You don't have to open the throttle to do a compression test on these engine. I'm positive about this.

Dan Caron's
 SL Barn
benzbarn@ebtech.net
 slbarn.mbz.org
  1 877 661 6061

[JimVillers]

Dan .... How can you get an accurate compression measure with the throttle closed?  How do you adjust for the vacuum in the intake manifold?  The compression pressures are different with the throttle open and closed.

Jim Villers
190SL, 230SL 5-Speed, 190E 2.3-16 Kompressor, MGB 5-Speed

[TheEngineer]

Do you have the stepped pistons? I have flat top Mahle pistons and my compression readings are in the 180's psi

[sterl]

They are the original flat top pistons. Bore is std. I can only dream of 180! If your engine is overbored I would expect higher readings.

What about cranking speed? It was getting quite slow at the end of the test.

I can't retry it at the moment as the pulley's off due to a shot fan coupling.

Mark

Billericay, Essex, UK
Late 67 LHD 250SL

[Benz Dr.]

No they aren't.  The whole engine will run at idle on the small amount of air that gets by the idle air screw.

During a compression test all the spark plugs are removed so you will only be filing the one cylinder that you are testing, right? So, how much air do you think that one cylinder will really need?
The amount that will fill six cylinders while engine is running at idle on the idle air circut, or WOT for one cylinder at starter cranking speed?
The results are; there is no material difference between closed throttle or WOT while doing compression testing on this particular engine. Other engines can have vastly different readings with WOT but not these particular ones.

Dan Caron's
 SL Barn
benzbarn@ebtech.net
 slbarn.mbz.org
  1 877 661 6061

[TheEngineer]

I second the observation of Dr. Benz and furthermore, when you do a compression check, the important thing is to have approximately equal readings in all the cylinders. If you have a few low cylinders, put a little engine oil into the cylinder via the spark plug hole, like maybe one squirt from the oil can. If the compression reading comes up substancially, you have leaking rings. Low readings are often caused by leaking valves. Many rebuilders surface the valves on a machine and machine the valve seat. They assemble the head without lapping the valves. The proper procedure is to lap the valves, then position the head with valves and plugs installed upside down. Then pressurize the passage from the outside of the head to the valve. You have to make yourself a rubber stopper with a tire valve in it. Pressure test it at about 15 psi. fill the head cavity with ATF fluid. Any valve leakage will show up as air bubbles. Lot of work? Yes.

[Benz Dr.]

I have a tool that puts a vacuum on the cumbustion chamber and tests the closed valves with the head removed. I can also test the cylinder with piston installed or the whole engine through the spark plug holes.
This is a very sensitive tool and anything on the valve seat or any leakage will give a low reading. The nice thing is you don't need to have the engine installed or running to make any tests. A portable air pig will easily do a V8 engine.

Slight leakage around the valves is common with older style valve grinders. Newer V block style valve grinders are far more accurate and the results are repeatable every time as long as the settings aren't changed. Collet syle grinders usually have dirt or wear inside which makes removing the valve a problem if you want to put it back in the machine for further grinding. The odds of getting the valve head perfectly centered or in the same spot each time are pretty close but the V block type grinders are 100% every time.
Collet style grinders can still do a good job but are nothing like a V block grinder.

Dan Caron's
 SL Barn
benzbarn@ebtech.net
 slbarn.mbz.org
  1 877 661 6061

[JimVillers]

Just so everyone understands the compression test procedure contained in the Manual.


Jim Villers
190SL, 230SL 5-Speed, 190E 2.3-16 Kompressor, MGB 5-Speed

[Benz Dr.]

That's why I said '' this particular engine ''

 The workshop manual is a guide or source for specs and assemby procedures. Sometimes information is of a general nature so that it will cover all general procedures. Engines with carbs may need to have the throttle opened but on these engines it make no real difference at all.

 Actually, I've stated this before and asked that someone do this test on their own. I guess I'll be second guessed by a workshop manual even though I know that it won't make any difference.

Dan Caron's
 SL Barn
benzbarn@ebtech.net
 slbarn.mbz.org
  1 877 661 6061

[Cees Klumper]

Hi Dan - below convinced me:

quote:Originally posted by Benz Dr.

... how much air do you think that one cylinder will really need?
The amount that will fill six cylinders while engine is running at idle on the idle air circut, or WOT for one cylinder at starter cranking speed? ...

Cees ("Case") Klumper in Amsterdam
'69 white 280 SL automatic