Valve/motor diagnosis

[Dave535Phoenix]

I was kinda putting my foot into my '93 535im yesterday passing 5k in 2nd gear when the acceleration went from strong and smooth to hesitating and not smooth. I backed off and the motor dropped to idle normally, but revving of the motor without load still produced vibration and stuttering.

It didn't feel healthy so I pulled off, opened the hood and listened to it idle. It seemed that there was some undesirable mechanical noise from the middle cylinders. A straight-6 running on 5 cylinders could feel like my ailing car does now.

Vital stats; 1993 535im, M30 engine, 162k miles. Stock except for a (great!) De Sylva chip. Oil changed regularly. Was half-way through a 25-mile trip; car was nice and warm. The oil light never came on, the temp gauge never moved, no cockpit indications that anything was wrong. Oil level was right in the middle of the range.

Possibly important; it blew it's head gasket about 2 years ago and I replaced it myself. I drained 11 quarts of coffee-colored oil/coolant mix from the motor before fixing it. I had the head machined flat and was pretty thorough putting it back together - after 2 years of problem-free running I thought I got it right, but I have heard that coolant in the oil can cause accelerated wear in crank bearings. The pistons and valvetrain looked to be in very good shape.

I accelerate in the car like this in 2nd all the time (I don't do it in 3rd anymore since that 67-in-a-40 ticket) ; it was not an unusual experience for the car.

I'm pulling the valve cover this weekend to take a look. Any ideas of what to look for and how? My best guess is that a valve spring broke. I'll check compression as well, then maybe pull plug wires individually to see if a cylinder is indeed dead, and which one.

TIA

Dave

[George M]

Hi Dave,
Agree...first step is to pull the VC and see what you find. My guess is a broke rocker if you have harsh mechanical noise. Sometimes cracked rockers arms are hard to detect as the deflect under load. I would then pull all the plugs for inspection...see if you can identify a mis-firing cylinder and perform a compression test to determine level of valve sealing. Be sure to check your valve clearances as perhaps one of the eccentrics came loose and you dodged a bullet.
I would not drive the car of course until you pull the VC and have a look as you want to keep any damage to a minimum.
Good Luck,
George

[Dave535Phoenix]

Well George, you were only half right. *Two* rocker arms were busted.

Both rockers of the #1 cylinder broke in two. They pieces were just sitting there next to the valves. Everything else looks fine.

Soooo.....now what. I'm almost certain I have to pull the head to get these replaced (yeah, I'll be replacing all of them, probably the valve springs too). Someone tell me I don't have to! Please!

I already did that. It was fun, satisfying, but a buttload of work and kept my car out of service for a month. The busted rockers are *right there* in front of me.

Maybe I'll just pull the plug and injector and run my car as a 5-cylinder :(.

(I'd moved the pieces before this pic was taken)

http://members.cox.net/davegpz/rockers.jpg

http://members.cox.net/davegpz/fullhead.jpg

[DueyT]

Dave, that stinks! At least this was a "fail-safe" mode, as opposed to a "fail-in-the-worst-and-most-expensive-way-possible" mode!

As those rockers aluminum? If they were ductile iron I'd recommend getting the set magnafluxed to check for cracks and if OK, just replace #1 cyl. If they're aluminum, I don't think magnafluxing works...I'd consider careful inspection and likely replacing them all.

Not positive, but I suspect the valve lash may have been off...

[Martin in Bellevue]

Hmm, cylinder #1 might be a bit oil starved. That is curious; I'd throw some rags over everything & crank, checking for oil from the spray bar.

I guess the half length rocker shafts can't get past the radiator core support & condenser; pulling the head can be fun, I guess.

[John B.]

Before you replace them it would be nice to figure out why both rockers on the same cylinder let go. It seems unlikely that both rockers were faulty.

[DueyT]

Hmm, cylinder #1 might be a bit oil starved. That is curious; I'd throw some rags over everything & crank, checking for oil from the spray bar.

I guess the half length rocker shafts can't get past the radiator core support & condenser; pulling the head can be fun, I guess.
Yeah, too bad it wasn't #6...either the head or the rad comes out. Ouch!

BTW, I don't think that's oil...I think the lash was off. Looking at the metal fracture on the spring side of the rocker, right at the base of the arm at the maximum extent of the cut slot for the roller tip. IMHO, that's the metallurgical stress/strain game...no issue with oil there.

Duey

[Bill R.]

may want to contact bmw about it. You might at least get some rocker arms out of the deal...
Rocker Arms (http://www.bimmernut.com/~billr/images/m30rockerarms.pdf)

[George M]

agree with comments implied above...strong hunch oil starvation was at the root of your failure. Did you replace the rocker shaft end plugs when you rebuilt the head last time? The plugs maintain oil pressure to the rockers through the hollow rocker shafts.
There is a hole at each rocker through each rocker shaft for oil to penetrate rocker journals. My guess is either the banjo bolts, oil distribution bar was installed backwards or a rocker plug or three were left out of your last rebuild to precipitate your result...only a hunch.
If your cam is a bit scored, I would clean that up as well...possibly a regrind...Martin may provide some advice there for a touch hotter grind...as you can't remove the rocker shafts on the big six...in particular the rearward shafts without pulling the head.
Don't think I would replace the valve springs...stretch check the head bolts for reusage...replace all rockers and likely 4 shafts and fresh eccentrics.
Also consider purchasing Don Gale's banjo bolt safety wire kit since you need to rebuild your head.
HTH,
George

[e34bim]

only twice i have seen this and the cause was backfire and piece broken from piston jamming in the valves, other fault could be broken spark plug but the fault does sound like its inside the cylinder

[Bill R.]

was oil starvation, if it is , it will be easy to spot because the rocker arm will be galled to the rocker shaft.. I've seen it before on other cars. However if the rocker can rotate and isn't seized to the shaft then how could oil starvation cause it to break where it did... If the bulletin i posted doesn't apply or has already been done, then it looks to me like valve float, meaning that the rpm was actually a ways south of 5k and the valves floated, When this happens i've seen it break both rockers on the same cylinder quite frequently.. along with bending valves and touching pistons.... Now if its valve float then i would definitely replace the springs when you replace the rockers and whatever bent valves you have. All the springs. John you may not have seen this before cuz desmos can't float the valves :)

[Martin in Bellevue]

Ireland has the 284 (& hairier 292) regrind & recommends their valve caps or their oversized eccentrics for the 292. Based on the marginal loss of low end torque from their 284, I'd be scared of a stock m30 with the hairier 292 cam. The 284 regrind doesn't lose its breath as much as the stock cam did, to 6k rpm.

https://secure3.nexternal.com/shared/StoreFront/default.asp?CS=ireland&BusType=BtoC&Count1=578960597&Count2=496101021

It is Saturday. Pull that head off & take some pictures. Take care with those exhaust manifold studs to the downpipes. Definitely get Don Gale's banjo bolts. They should be stock on these things.

This is an opportunity to add a little something to the performance of yer 535,
Martin

[Dave535Phoenix]

may want to contact bmw about it. You might at least get some rocker arms out of the deal...
Rocker Arms (http://www.bimmernut.com/~billr/images/m30rockerarms.pdf)


My chassis number is 13070 - in the range given in the bulletin. There's no paint spot I can see on the oil pressure switch stub, and there's no dealer sticker on the B pillar.

BMW might be making this repair....thanks man. I'll let you know how it works out.

[John B.]

If the bulletin i posted doesn't apply or has already been done, then it looks to me like valve float, meaning that the rpm was actually a ways south of 5k and the valves floated, When this happens i've seen it break both rockers on the same cylinder quite frequently.. along with bending valves and touching pistons.... :)

Bill - What RPM would you have to hit to float the valves on the M30 engine? 7K? or more?

[Bill R.]

design criteria for the valve springs, rocker assy was probably well in excess of 7.5 k rpm however (here's the LBV part) if your familiar with valve float and valve spring design you probably already know this. The valve spring has a resonant frequency that occurs at x rpm.. meaning that when you compress and uncompress that valve spring at a certain speed it loses its spring tension and can't hold the valve shut. This is known info and the designers use alloys and spring tensions to make sure that this frequency doesn't occur in a rpm that the engine will be seeing normally. On a number of engines they use an inner and an outer spring on the same valve, since these springs are different diameters, spring pressures and or alloys they have their harmonic or resonant frequency occur at different rpms, so the outer spring may turn to silly putty and 7k but the inner will still be providing spring pressure at that rpm, the inners harmonic will be at a different rpm so they overlap and you have constant spring pressure to keep the valves closed throughout the engines designed rpm range. Now this is all fine and dandy and the single spring on the m30 works throughout the designed rpm range. But as springs age the metal fatigues and changes its characteristics. And typically the rpm at which they may float will change too... A quick definition of float is the rpm at which the springs aren't strong enough to keep the valve following the cam lobe and the valve can bounce open when this happens, IF the valve happens to be hangine open when the piston comes up, the piston smacks it HARD and slams the valve closed, the valve hits the rocker arm HARD and since the rocker arm is usually in the process of coming up towards the valve at the time , the weakest link gives and thats usually the rocker in this case.. At any rate when the valve springs get old the rpm at which they float can change and get unpredictable.. Thats why its a good idea to have a valve job done on a head any time you have it off for a head gasket, any machine shop worth their salt will check and set spring pressures when they do a valve job...

[John B.]

IF the valve happens to be hangine open when the piston comes up, the piston smacks it HARD and slams the valve closed, the valve hits the rocker arm HARD and since the rocker arm is usually in the process of coming up towards the valve at the time , the weakest link gives and thats usually the rocker in this case.. At any rate when the valve springs get old the rpm at which they float can change and get unpredictable.. .


This is what I was thinking when I saw the pics. It looked like the valves hit something hard (piston) & snapped the rockers. Popping the head off will answer that question.
So it might not be such a good idea to explore the extended RPM range of my EAT chip on a regular basis at 142K?

[George M]

Nice explanation Bill and you may indeed be correct if the rockers aren't frozen or bound to the rocker shafts. Or could have been weak rockers per the bulletin you provided. Dave do you ever recall over-reving the engine?
In any event all bases are covered as to possible root cause :-) Bill how do spring pressures get adjusted on a big six short of swapping out springs?..not aware of any different shims available to increase seat pre-load.
And Martin agree with ya...the grind you are running sounds perfect for a stock or lightly modded M-30 street motor and might add I like your philospohy as well. When something breaks always look upon it as an opportunity to make it better than it was...a great perspective.
George

[Bill R.]

Say at an average engine speed of 2k rpm the springs open and close
1000 times per minute
Now lets say that corresponds to 50 miles per hour at 2k rpm
Thats 60,000 times a hour and 120,000 time per hundred miles
or 1,200,000 times for every 1k miles..
or 12,000,000 every 10k miles
or 120 million times every 100k miles
So you've got to be hitting the 150 million times mark for squeezing those poor little bastard valve springs on your car about now..... Its a wonder they last this long...











QUOTE=John B.]This is what I was thinking when I saw the pics. It looked like the valves hit something hard (piston) & snapped the rockers. Popping the head off will answer that question.
So it might not be such a good idea to explore the extended RPM range of my EAT chip on a regular basis at 142K?[/QUOTE]

[Martin in Bellevue]

I forget which issue last fall, but Hot Rod had an intertesting article about how these springs were lighter & stronger, at least for the ubiquitous chevy small block. The springs required smaller retainers up top. They were shown to reduce valve float to much higher rpms. If a machinist could be persuaded to give 'em a try, they might be fun.

[Bill R.]

shims, the machine shops have standard shims in various thickness's and diameters for adjusting spring pressures... The thing that i have been wondering is that i can't find a spring pressure spec for the bmw anywhere.. I guess you have to take a new one or 10 and measure the installed height pressure to come up with an accurate number for what the spring pressure should be on them. The only numbers i have are for older cars 87 and older and that was 64lbs pressure at 1.48 height.












Nice explanation Bill and you may indeed be correct if the rockers aren't frozen or bound to the rocker shafts. Or could have been weak rockers per the bulletin you provided. Dave do you ever recall over-reving the engine?
In any event all bases are covered as to possible root cause :-) Bill how do spring pressures get adjusted on a big six short of swapping out springs?..not aware of any different shims available to increase seat pre-load.
And Martin agree with ya...the grind you are running sounds perfect for a stock or lightly modded M-30 street motor and might add I like your philospohy as well. When something breaks always look upon it as an opportunity to make it better than it was...a great perspective.
George

[George M]

just hadn't seen any shims on the ETK Bill. My bet is few shops really sweat this detail as important as it is.
Thanks,
George

[Dave535Phoenix]

Overreving? Er, um..well. OKAY! I ADMIT IT!!... I've hit the rev-limiter quite a few times over the last couple years. It just revs so nice with the EAT chip.

[Dave535Phoenix]

I'm sticking with Bill here. I knew all about the oil-spray bar and banjo bolt when I did the head gasket and made damn sure to get it in correctly and loctite the bejesus of the banjo bolts. They're nice and snug now.

Having the benifit of fondling the pieces, they show no signs of undue wear - they're actually in amazingly good shape if you consider there's 162k miles on 'em. Well, aside from the rockers being broken in half that is. There are no signs of oil starvation anywhere. Wear is consistent for all rockers and cam lobes.

I don't buy incorrect valve lash either. As time goes on, valve lash decreases as the valve/seat contact point wears. It won't ever increase. The only way to get too great of a clearance is to adjust them incorrectly, and they haven't needed adjustment since I've had the car (3 1/2 years now). If the valve clearance gets too small, it causes problems with the valve sealing, that's all. Nothing will break.

To break both rockers, the only thing that really makes sense is if they tried to push on a valve that didn't want to move. To stop valves from moving, they have to hit something - and aside from the unlikely disintegrating piston or broken plug, the only thing a valve can hit is the other valve.

Valve float fits. Knowing that the tired springs have been closing the valves for 160k bolsters the case.

To know for sure, I have to get the head off. If I see signs of valve-to-valve contact I'll know it was valve float. If I don't, it had to be weak rockers (but both at once?).

It'll be a week or two before I get the head off. I'll start another thread then.

I probably will breathe on the head a bit before it goes back on. I wasn't planning on it, so I'll have to look around and see what fits my budget and performance goals.

Thanks for all the responses guys.

Dave

[e34bim]

i hope your problems are minor and good luck

[George M]

hmmm....now that you say you hit the rev limiter often on a 162k mile motor with original valves springs...I'm with Bill too...LOL. Pretty much sums it up. To add insult, isn't the rev limiter raised with Mark's chip?
Good Luck with the rebuild,
George