m103 compression interpretations + valve guide questions

Page may contain affiliate links. Please see terms for details.

Brabus3.6

Active Member
Joined
Apr 9, 2004
Messages
109
Location
London/Oxford
Car
1989 Brabus 3.6 Saloon
Well my Brabus 3.6 has eaten its way through yet another headgasket :(, thankfully I have an MLS gasket that PUMPISH (guy the mercedesshop forum who's putting out 600bhp from m103s!) made for me. I've been losing oil in this car at the same rate (~1L/300mi) I had done before the rebuild last summer (installed gapless 2nd rings having re-bored a new block out to 92mm). I gave it a compression test today with the following results for cylinder numbers
1: 12.5 bar
2: 12 bar
3: 11.5bar
4: 8.5 bar
5: 13 bar
6: 10.75 bar.
I expected adjacent cylinders to be low on compression. I added a teaspoon of oil to cylinders 4 and 6 and re-checked the values but the compression changed by maybe a quarter of a bar which I don't deem to be significant considering the gauge I used is a Gunson's gauge. My car has manual valve lash adjustment which I adjusted less than 3,000 miles ago. The valve guides were changed about 20,000 miles ago; the last time I had the head off (~5,000 mi) my engineer told me that the guides were fine.

The car idles slowly (just over 500rpm) and rather grumpily with a slight misfire that clears up at around 2000 RPM (it no longer has the hesitation which I have previously posted about though). On the motorway there is little perceivable difference between how it runs now and how it ran 30,000 mi ago. It's very torquey and 90 to 140 is still very impressive. Low down, from a standstill I guess it is slightly slower than its optimum.

I'm wondering whether the valve guides on cylinder 4 are shot (certainly I remember them being darker in colour than the others) and whether this is the cause of my oil consumption. Furthermore, if this is the case then perhaps the worn guides are preventing the valves seating properly? The spark plugs are all dry, the 4th plug has a burgundy colour and the 5th is black, sooty. I'll be opening the engine up tomorrow and having my engineer check the bores for wear and piston deck height (for a possible bent conrod) on Monday. Ideas & guidance would be much appreciated.
 
Before removing the head perorm a cylinder leakage test to ascertain whether the leakage is due to valves or rings.
 
Thank you for your replies.

grober - I can't deny that I've read that but I honestly don't buy it. My engine didn't start consuming oil until around 110,000 mi and KLP's also didn't seem to have any such problems. The bore can take it, the trick is in getting the headgasket to seal on the now wafer thin webs (and perhaps moreso, in getting a headgasket!).

Dieselman - can I perform a leakage test using only my Gunson gauge, if not what else do I need? I was assuming that if the engineer checks the bores for wear and finds none then it is the headgasket (or the head). Can a feeler gauge not be used on a head which has been removed to test for an incomplete seal?

Are cylinders 4 & 6 my only problems?
 
Excessive oil consumption has been linked to something as simple as hardened valve stem seal O-rings. This wouldn't explain your lack of compression tho.
 
Cylinder leakage tester works off compressed air, so some garages will have them.

Ian Walker In action.

Use a straight edge and feeler gauges to test for flatness of the head.

You could make a copper gasket to size then seal it with special thermo sealant both sides.
 
Do you mean the valve seat ring, number 11 in the pic here. I'm wondering whether the people who installed my valve guides got them mixed up (intake and exhaust). Is there a visible difference (as with the seals)?

Loss of compression would then simply be down to my headgasket which unfortunately, I know is gone.
 
Ferriday actually already have the file for the brabus 3.6
They make specialist copper headgaskets and IIRC are just up the road from you :)
http://www.ferriday.co.uk/copper/gasket_list.shtml

I would get the gasket from Ferridays, roll my sleeves up and figure out how it all comes apart :) By all I mean strip and measure the head, check the guides, cut the seats all yourself. If you do it once yourself and take loads of time over it you will likely get it right. As 80% of the cost here is labour you could save a fortune, learn a lot and be sure that everything is as it should be. Actually fitting new guides is best left to an engineering shop though.

Check the state of the block face very carefully, Use a deck plate/engineers level and make sure the block is completely square. If you have a couple of thou discrepancy on the block (esp between cylinders) this will provide a weak spot ofr the next head gasket blow.


Sorry if any of the above is teaching you to suck eggs :p


Mark
 
They didn't have that on file when I inquired! I have an MLS gasket for the car so that base is covered. Do I not need any special machinery to cut the valve seats and valves then?

Dieselman - since my headgasket is gone wouldn't a leakdown test be useless, I would have thought the path of the headgasket breakage is the easiest for the air to traverse.
 
MLS gasket will work fine if the surfaces are perfect , copper may be better if there is any doubt or you haev a history of blowing gaskets.

Low compression may well be valve sealing but could equally be any block scoring. I dont use gapless rings in any application and instead gap conventioanl rings accurately. Any blowby is easily contained with a decent breather system (and in some of my high boost turbo engines there can be a reasonable amount of blowby!)
Valve seat cutting for new guides really requires a seat cutter, but if the guides dont need replacing thenn you can gring the seats in the old fashioned method.

The difference is that you know you have done it, you know the seats are accurate and time taken to achieve the results wasn't an issue - just quality of finish.
I will go to an engineering company every time I need specialist machining work - but all they do is machine or replace what I ask them to do. I strip everything, measure everything, and determine what actually needs replacing. Then I build it all back up again :)
This cuts down on engineering labour charges but more importantly you know there were no corners cut in any way.

We rebuilt my mate Eddies 16v Honda head last Xmas, he had never done it before and it took 2 whole days of my supervising his labour before I was happy to put it back on the car :D
 
No I mean the valve stem oil seals. I think item 14 is the valve seat insert in the head ?? the inlet and exhaust valve stem seals are different sizes and should be installed using assembly sleeves to avoid damage.
 
Last edited:
Dieselman - since my headgasket is gone wouldn't a leakdown test be useless, I would have thought the path of the headgasket breakage is the easiest for the air to traverse.

Obviously if the gasket is definately leaking the readings will be low for those cylinders, but a leakdown test will ascertain whether there is valve or ring leakage even on those cylinders.

If the valves are leaking you will hear air being forced into the manifolds, if it's rings/bores then you will hear the air passing into the breather system via the crank-case.

As Mark said, if you can, do your own labour, it's unlikely that any commercial enterprise will take as much care as you will when building up the engine/head.
 
A picture speaks a thousand words, so here are a few of the internals.

Headgasket break
Similar breaks are repeated in a couple of other places, I'm guessing this is where my coolant is disappearing into the combustion chamber.

Underside of the head
I've already noticed some dirt on the under side of number 5's exhaust valve, the part that is cut and should seal with the seat, I doubt this can seal properly in its current condition. My fuel distributor was overhauled by Ken Mills Injection Limited a year or so ago. Perhaps cylinder 5 is this this sooty due to a leaking injector?

A typical cylinder
The visible strips look to be what Mercedes describe as friction marks due to frequent short journeys and is not a worry according to their literature.

As for the valve stem seals, I replaced them when I put the engine back together last summer. I've just checked its disc diameter, they are standard Mercedes valves. I will check the valve dimensions against the Merc specs. Anything else to check for on those? How about the seats, when do they need changing?
 

Attachments

  • Brabus (0).jpg
    Brabus (0).jpg
    40.9 KB · Views: 76
  • Brabus (1).JPG
    Brabus (1).JPG
    76.8 KB · Views: 75
  • Brabus (4).JPG
    Brabus (4).JPG
    80.3 KB · Views: 79
My guess from the picture would be that cylinder no 5 is running appreciably cooler than the others-maybe due to a coolant leak into it. This cools the exhaust valve appreciably allowing deposits to build up on the valve head and valve seat. I have heard of head corrosion problems on engines using add on "water injection systems"so a careful exam of the valve seat may reveal some recession. The other possibility is some form of mechanical "stiction" preventing the exhaust valve from seating properly altho I can't see why this would cause it to run cooler? That 5th spark plug was sooty you said-it didn't look "oily" black did it ? as opposed to "dry powder" black. Cylinder head gasket leakage has given me "oily looking " plugs in the past. Its not oil of course but damp sooty deposits. that's my guess anyway.;) This is a good little article on CHGs from Reinz http://www.reinz.com/pictures/praxisinfo_3_eng.pdf
 
Last edited:
Looking at cylinder No5 I would do as Dieselman suggests. Do a leakdown test. Something not quite right there. Start at the basics first. Mark300sl has the right idea, do all that you can yourself. Not only will you know its right but its therapeutic. You will also save a fortune.
 
Looking at the design of the gasket it looks like there isn't enough supporting metal just near the compression rings, which is causing the gasket material to fracture.
I would definately be inclinied to go for a solid copper gasket to ensure it doesn't blow again.

In the picture of the valves, no5 look oily. Are they?
 
Looking at the design of the gasket it looks like there isn't enough supporting metal just near the compression rings, which is causing the gasket material to fracture.
I would definately be inclinied to go for a solid copper gasket to ensure it doesn't blow again.

I'm with Will

We have similar problems on big bore Hondas under boost, there is an overbore point where we just won't turbocharge a particular bore size because the lack of gasket material means they are always blowing. However - as you have the gasket from Pumpish you might as well try it :)


Oil Control

Oil can only normally enter the combustion chamber past the rings or valve guides, (assuming its not breathering heavily back into the intake tract). If the guides are good, and the valve stem oil seals are also good then it just leaves rings.

Who assembled the pistons and rings last time ?
Did you use a bore micrometer to check for ovality?

I use specific rings to combat oil loss problems as they are exacerbated under boost. Gapless rings are not on my shopping list, ring end gaps are critical as is ring end gap placement. I use second compression rings with an oil scraper as well as the conventional 3 ring oil scraper for the 3rd oil control ring.
From my experience the gapping of rings and positioning of the gaps has a marked significance with regards to oil control.

If it were me - I would be tempted to take one of 2 routes
1. Have the whole lot apart and check bores, rings, ovality and rebuild head
2. Do a total and thorough rebuild of the the head, run the correct size reamer down the valve guide to ensure non of the valves are sticking, Recut the valve seats and lap everything to perfection, replace the stem seals yeat again and have the head skimmed by the absolute minimum amount.

Pay particular attention to block cleanliness and flat check it until you are sick of doing it :)
If you are absolutely sure the head is perfect then any future oil control issues must be back to the bottom end
 
I have attached pictures of

Piston #5
Combustion chamber #5
& Inlet port #5

It looks like number 5 the dirt of combustion chamber 5 is oil, looking at the state of its inlet port. Given the good compression of cylinder 5 while taking account of the oil in it I'm pretty sure it's a shot valve guide. How much oil can only a couple of shot valve guides account for? I will certainly be disassembling the head and measuring the valves myself.

Inlet ports 4 & 6 are dirtier than 2 but better than 5. I will be measuring the valves today. If any are out of spec should I order both a valve and valve guide in repair size? How about changing the valve seats?

The rings in there are gapless 2nds, they were installed by an engineer, Middlesex Re-boring co., the same company that the dealers use. I would have though an MLS gasket is stronger than a composite one in sealing combustion pressure, not as good as an o-ringed block with a copper headgasket but o-ringing my block would be a joke I guess. Mark300SL - my car running at slightly higher compression compared to a normal 300E (10:1 versus 9.2:1) but I doubt it is under the same strain as a turbocharged car. I will take your route of rebuilding the head.
 

Attachments

  • user25488_pic335_1218367436.jpg
    user25488_pic335_1218367436.jpg
    83.4 KB · Views: 45
  • user25488_pic337_1218371249.jpg
    user25488_pic337_1218371249.jpg
    41.3 KB · Views: 44
  • user25488_pic339_1218372649.jpg
    user25488_pic339_1218372649.jpg
    76 KB · Views: 44
A verdict of sorts

I thought I was onto something yesterday as I was dissembling the head. The intake valve guides were all tight but the exhaust guides had some play in them. However my engineer inspected them and told me that they're all perfect and the extra play allows for the expansion of the much hotter exhaust valves. The valves themselves all measured up within tolerances. I must reiterate that he is used by Mercedes of Colindale and so has seen many m103 heads.

He then inspected the block, immediately telling me that there is no wear in it but that it has become glazed and looks like a 100,000 mi engine (despite having a new bore 5,000 mi ago). He told me to lap the valves in again and get the pistons to him. The bore will be honed and a new set of rings put on. The engine was running very rich for 500 mi after the rebuild which he thinks caused the glazing.

Mark300SL - what rings do you use? The original oil control ring was a 2 piece, a ring and an expander. The Total Seal oil control is a 3 piece. Are you a suggesting usage of a 2nd napier ring will help further with oil control?

P.S. the new m103 valve stem seals are all the same (both intake and exhaust), just bought them from Mercedes. An improvement or cost-cutting?!
 
Last edited:

Users who are viewing this thread

Back
Top Bottom