Compression Test: Quick request for clarification

[Shortsguy1]

Just performed a compression test on my 1968 220D, but I realized that I might have done it slightly wrong. (all were 360 -390 psi via the glowplug hole, using the method I am about the describe).

I let the engine crank until the pressure no longer rose on the pressure gauge. But afterwards, I noticed that both the FSM for my car and the fine instructions in my Harbor Freight compression tester indicate I should only "crank the engine by 8 revolutions...".

So 8 revolutions at the crank is only 4 compressions of the cylinder in question (I think), so which of these three options is correct:

1) Crank the engine until the pressure no longer rises on that cylinder (maybe 12 compressions or so)
2) Crank the engine so 8 compressions occur on that cylinder
3) Crank the engine so 4 compressions occur on that cylinder

Thanks, and sorry if this seems like a dumb question. But it is hard to make sense of the FSM's 22-24 bar (319-350 psi) unless I know exactly how much to crank the engine. If it is option 3, my engine has low compression (maybe 200 psi). If it is option 2, my engine has mediocre compression (maybe 300 psi). If it is option 1, my engine has pretty good compression (the 360-390 mentioned above).

[t walgamuth]

I always crank until it stops rising. You cannot compress it more than the rings etc will allow.

[sixto]

FSM probably says to run the test on a hot engine too. How hot can the engine be by the time you remove all the glow plugs and test the last cylinder? You should be fine as long as conditions are the same or don't change much between cylinder tests. You're looking more for variance than hard numbers. Worst case is your engine needs the starter to run a little longer to fire

Sixto
87 300D^2

[vstech]

8 revs may work for a hot engine... I always crank until it stops rising...

[Shortsguy1]

Thanks all for your thoughts here. I am glad that I am not alone in my approach to compression tests. Basing a test on a specific number of cranks is not a repeatable process, so it must be (as said above), that if warm, 8 cranks is more than enough to achieve max pressure.

I am surprised that the specs for my engine say: "Compression pressure normal= 22-24 bar" (319-350 psi). In looking at many of the values posted on this website, many of you have pressures well above 350 psi. It seems odd that MBs "normal pressure" would not include the high pressure values for a new or well maintained engine.

[Stretch]

I think compression test data is a nice indication of engine health but it is very much dependant on the strength of the battery and the starter.

If you've just tested four cylinders consecutively you might find that the last cylinder you tested will test differently tomorrow after the battery has recovered!

For this reason I understand that the biggest difference between each of the readings you obtain is more important.

What does the manual / FSM for your engine say for that?

[cullennewsom]

Originally Posted by Army

... very much dependant on the strength of the battery and the starter.

If you've just tested four cylinders consecutively you might find that the last cylinder you tested will test differently tomorrow after the battery has recovered!

I've never hear that one before!

[funola]

Originally Posted by Shortsguy1

Thanks all for your thoughts here. I am glad that I am not alone in my approach to compression tests. Basing a test on a specific number of cranks is not a repeatable process, so it must be (as said above), that if warm, 8 cranks is more than enough to achieve max pressure.

I am surprised that the specs for my engine say: "Compression pressure normal= 22-24 bar" (319-350 psi). In looking at many of the values posted on this website, many of you have pressures well above 350 psi. It seems odd that MBs "normal pressure" would not include the high pressure values for a new or well maintained engine.

Mercedes FSM compression numbers are based on physical dimensions of the combustion chamber. Higher than spec readings can be:

1. Gauge not accurate

or

2. Carbon buildup in the combustion chamber

[Stretch]

Originally Posted by cullennewsom

I've never hear that one before!

Which bit?

The bit about a healthy battery?

The bit about a healthy starter?

Or (most likely) the bit about starting with the last cylinder you tested the next day / time? (Please note I did say MIGHT!)

The last one is based on my experience of helping out friends who have never liked spending any money on their cars - they've always been worn out => one step from the grave yard type of cars...

...in these cases it is hardly worth your time even starting with a compression test - skip straight to a leak down test - tell them which bit / bits of the engine is / are shot - explain to them that they need to get some machining done - watch them drive their car into the ground and off to the scrap heap!

[Shortsguy1]

The FSM for my 220D with a 21:1 compression ratio says "Compression pressure normal= 22-24 bar" (319-350 psi) and "minimum compression pressure ca. 15 bar" (218 psi). I worked from front to back and my minimum pressure was on cylinder 3 (360 psi). I have an old, but healthy battery, so I don't think it played a part in my case. The engine was not warm, so nothing was cooling off as I worked.

After the compression test, I adjusted all the valves for the first time. My intake valves were a little tight and my exhaust valves were a bit loose. I suspect the last mechanic set the exhaust valves at a bigger gap than specified in my fsm (0.3mm). I was able to get my 0.35 feeler gauge in there consistently before I made any adjustments.

I had planned to re-check the compression after the valves, but realized it wasn't necessary. From a compression standpoint, the engine seems healthy. It starts easily and idles smoothly (once I adjusted all the control linkages, that is).

The next step is to address the chain stretch. I only did the cursory measurement aligning cam marks and it came up at 7 degrees (yikes!). I repeated the process 3 times and got the same measurement, so the method is consistent (if not accurate). I plan to borrow a dial indicator and use the correct valve lift method soon. I will be curious how well the cam mark method compares. Many here pan the method as innaccurate, but I had trouble finding many comparisons of the two methods. It seems that if 5 to 10 of us checked chain stretch with both methods, we could end the debate one way or another.

If the chain already has a woodruff key, would that be evident in my cam mark alignment method? In other words, does that thrust washer with the cam pulley alignment mark also get shifted with a woodruff key?

[t walgamuth]

YOur compression is excellent. The 220 cold is required for reliable starting.

[Govert]

Compression tests should be done with the engine warm. Cold compression tests are useless as there is no reference with which to compare.

Compression test results can differ depending on the condition of the starter engine and battery, also on ambient air pressure. They have more use as a comparative figures (difference between cylinders, change from year to year) than as absolute figures. The best compression test is whether the engine starts in winter (providing you have cold winters).

Early engines had lower maximum numbers, the quoted 22-24 bars. Later engines had 30 bar as maximum and 18 bar as minimum. Also important in the early engines with pneumatic controlled IPs is to keep the throttle in the open position (accelerator pedal fully pressed).

[t walgamuth]

I'm not sure I agree that cold compression tests are worthless. After all, when you want to start the car it is cold.

[Govert]

The problem is that MB does not give a reference number for cold compression tests, so you never know if the result is OK or not. There can be 10 bar difference between a cold and warm compression test.

[Shortsguy1]

Is it safe to say that a cold compression test always returns a lower value than a hot compression test? That is my intuition, but perhaps there could be some strange effect where cold, thick oil could help seal some gaps.

As a funny aside, the car likely had the GPs installed 20 years ago, but for that length of time, one was only finger tight. I got my 22 mm wrench on it, expecting to strain while removing it after 20 years of sitting, but it moved before I put any pressure on the wrench. The probably only had 20,000 miles in that time, but it is funny that a finger tight GP was sufficient for it to run well.