S50 Blown Headgasket

[M3PO]

I am pretty sure that my 117K mile S50 has a blown headgasket. I've put a few thousand miles on it and never overheated it. I've got a new E36 rad, Stewart racing water pump, Spal fan in place of the aux fan, and a 80/88C switch. I have an aftermarket temp gauge to keep an eye on temps. It started to steam a lot a few weeks ago on cold starts and now the oil cap is looking milky. The coolant level is going down fast but I also have a leaky heater core. I read the Pelican DIY and now I have a bunch of questions.

1) Which gasket? Is OEM the only trustworthy one?

2) Head bolts? It won't see boost.

3) It's my only car right now so I don't want to send the head to the shop (weekend job) unless I have to. Is this irresponsible?

4) I have a liffter that sticks for a few seconds on start up sometimes. Clean and polish them? Replace that one? Replace them all?

5) Special tools. Best/cheapest way to get a cam holding tool? What about torx tool for head bolts?

6) I've heard the S50 has weak valve retainers. This engine will see a lot of high RPM track time soon. Should I replace them and with what?

7) "While I'm in there", I want to do some cams. Which ones? I want something that makes proven power but isn't outrageously expensive, ie the best "value".

8) I have VANOS rattle so I will be rebuilding it with the Beisan systems kit. Any tips?

[M3PO]

Start ups have been rough recently and the heater core likely blew around the same time. I am assuming that the coolant system is being pressurized by the exhaust gases and that's what blew the heater core. Rough startups are probably caused by coolant in the cylinders.

[exhaust]

Ah that sucks man. I've been down this headgasket road a few times and it really isn't as bad as it sounds. To answer your questions:

1. Elring or Victor Reinz is fine. I've redone three headgaskets using an Elring once and a VR twice. Once on a S50, once on a hard tracked M50 and once on a daily. The VR gaskets were using on the S50 and track car. VR makes gaskets that are a little thicker if you get your head decked as well.

2. Just use normal head bolts. You don't need to spend the money on ARP head studs or anything. VR head bolts are fine.

3. I always recommend sending the head out to be checked out. You don't want to be doing this again. Call a few shops and price shop around. Typically I've found that it's about $150 for a pressure test, cleaning and re-deck all said and done. I've brought the head to them with the cams and everything still assembled as well. Remove head, bring to shop, done.

4. Not sure about the lifters, I haven't dealt with this.

5. Go on eBay, buy the cam tool kit. Comes with the flywheel pin, cam block, sprocket wrench and tensioner. It's <$150 and has served me well.

6. I don't know the answer to this.

7. I hear great things about Schrick cams but they are expensive. Any cams won't be cheap.

8. The Beisan kit is great. Get the rattle repair and the seal kit and take care of it once and for all. Takes about 20 minutes once it's out of the car. Follow their instructions carefully and you will be good to go. As for tips:
-DO NOT over torque the hex bolt. It will snap easily.
-When getting it apart for the first time, use a torch on that shallow head. It will round itself easily and the heat makes life a lot easier
- Buy their shallow socket and the vice adapter. You may not think you need them, but you will really thank yourself. You can always resell them and they are not that expensive.

PM me if you need any clarification, I would be happy to help.

[berlow94]

I would only ever use a genuine OEM BMW gasket. Not worth saving a couple bucks there...

If you want 100% bombproof then bring your head and block to the machine shop, get them both decked, install stock thickness MLS gasket with ARP studs, and never ever ever worry about doing a head gasket again...

OEM head bolts are fine.

Not the end of the world to swap out 1 or 2 sticking lifters, but would be ideal to do them all at once.

I would definitely bring the head to the machine shop to get checked and decked. Have them look at your exhaust side valve stem seals as well.

I'm very happy with my Schrick 256/264 cams. The only cams aside from stock i would recommend for these motors.

I've found it's usually always worth it to buy the tools. You'll end up using it once or twice but your friends will thank you!

[M3PO]

Are injectors mandatory with cams? I feel like the duty cycle can't be increased THAT much.

[berlow94]

My motor has a little more than just cams. But once i plugged in my Vipec ecu and could accurately log my injector duty cycle, i was surprised to find that my 24# injectors were hitting 100% duty cycle. I am now using 32# injectors and i haven't seen higher than 85%.

Basically, you want the biggest injector you can put in there and still be able to idle.

[M3PO]

Next question lol!

Thicker head gasket? I am looking at the big boy VAC Dominator cams (286/272) so I am obviously concerned about piston to valve clearance. The 1.0 mm gasket (0.3 mm thicker than stock) seems like over kill for a light skim. I'd rather increase compression than decrease it obviously.

[berlow94]

Don't buy those cams... There is absolutely no reason. Spend your money elsewhere. Im making within 15 hp and torque of S54's. If anything use a stock thickness head gasket and limit or remove vanos. Happy tuning...

[M3PO]

I appreciate the info. These cams are part of a package that is basically a VAC Stage 5 kit and then some so it's a sweet deal. Is there a reason NOT to run these VAC cams?

[328ijunkie]

My motor has a little more than just cams. But once i plugged in my Vipec ecu and could accurately log my injector duty cycle, i was surprised to find that my 24# injectors were hitting 100% duty cycle. I am now using 32# injectors and i haven't seen higher than 85%.

Basically, you want the biggest injector you can put in there and still be able to idle.

Unless youre over 300whp there is no way youre maxing 24s. Im not sure how vipec logs IPW but how is it caculating IDC? IDC is engine/RPM dependent more than anything else. A S52 For instance at 7K RPM has 17.14ms of injection time available. So if the ecu is requesting that or more at that RPM then youre at or above 100% injector duty cycle.

But at 3500RPM you have 34.28ms of possible injection time...

Don't buy those cams... There is absolutely no reason. Spend your money elsewhere. Im making within 15 hp and torque of S54's. If anything use a stock thickness head gasket and limit or remove vanos. Happy tuning...

within 15hp of S54? Dyno?

[berlow94]

Unless youre over 300whp there is no way youre maxing 24s. Im not sure how vipec logs IPW but how is it caculating IDC? IDC is engine/RPM dependent more than anything else. A S52 For instance at 7K RPM has 17.14ms of injection time available. So if the ecu is requesting that or more at that RPM then youre at or above 100% injector duty cycle.

But at 3500RPM you have 34.28ms of possible injection time...




within 15hp of S54? Dyno?

Happy to share with you my ECU logs...
Also happy to put you in touch with my tuner.

As far as flow is concerned, im using a 75mm billet throttle body (with the same diameter piping all the way to the air filter. No MAF), a M50 manifold that was bored to work with that throttle body, 1mm oversize valves, supertech valve springs, Schrick 256/264 cams, REM polished vanos internals, and a 5 angle valve grind. Stock thickness MLS gasket with ARP studs.

My car was dynoed at Granite Sate Dyno in NH. That same day a bone stock E46 M3 was also dynoed on the same dyno. The M3 was in perfect working order and did not seem to be lacking any power. The difference that car put out was within 15 hp of mine to the wheels. I also have an REM polished tranny and diff so i'll admit i might have less drivetrain loss than the M3, but regardless, my only point for mentioning that is that you don't need gnarly cams to make big power with these motors. The engineers designed vanos for a reason. Why mitigate their efforts?

[berlow94]

S50 Blown Headgasket

Here is short list of logged values from a track day at NHMS on pump gas.
Max duty cycle with 32# injectors reads 77.5%.

Even if 24# injectors were reading >90% for someone, wouldn't you agree that you are getting more efficiency with the bigger injectors as long as the car will still idle?




Sent from my iPhone using Tapatalk

[328ijunkie]

You ideally never want to be above ~85% as the injectors are firing at the back of closed intake valves. what WHP are you at? I again have no idea how VIPEC caculates Max IDC but that doesnt seem accurate at all...

M50 Manifold S52, did ~240whp.

its at 15.293 IPW at 7K and 17.07 at 5900RPM. This is ~85% IDC on 21.5s
The result if i recall was around ~12.5AFRs which was a little too rich. Maybe your tune is targeting <12 or something, i can see you pegging injectors pretty quick targeting that rich. (these motors dont like sub 12...)

[berlow94]

I'm making around 290 to the wheels on a dynojet.


Sent from my iPhone using Tapatalk

[npavlo]

Injector duty explained

To keep things fair, I will answer and inform unbiased.

Duty cycle in general is defined as the ratio of on time vs off time during a period of time. In this case for injection, this would be the amount of time the injector is driven low vs RPM.

As the RPM increases, the number of cycles increases regardless of injector pulse width.

At 1000rpm, it takes the engine 60ms (0.060s) to turn once. So consider a full cycle x2 = 120ms.

At 10000rpm, it takes the engine 6ms (0.006) to turn once. So consider a full cycle x2 = 12ms.

If I just use 10ms of injection time in both situations,
the duty cycle at 1000rpm is 8.3%,
the duty cycle at 10000rpm is 83%

Hope that puts into perspective the dynamic change of an engine through the RPM range!

For fun, just picture at high RPM, and high duty, how long the injector is spraying at a closed intake valve waiting for it to be opened :) Consider the intake valve is only open for a 1/4 of the cycle (it's actually less) - at 10k rpm - that's 3ms!

The injector offset factor has not been mentioned. This is the lag time of the injector, also called deadtime. Injector duty doesn't factor lag time, but because of lag time, anything over 85-90% duty should be considered the injector actually never closes between cycles because of this. Average high impedance injector deadtime at 14v is probably somewhere around 0.800ms for example. A number of factors effect this like battery voltage, fuel pressure, fuel pressure differential (boosted cars), and more.

Running very high duty cycles is bad because :

The injector overheats (it is on all the time)

Control of the injector is not accurate. Most injector flow rates cap out at a certain duty cycle - it is basically open all the time

If the injector timing not adjustable, the mixture per cycle to be uncontrollable . Typically the mixture actually becomes uncontrollably rich, just before it becomes uncontrollably lean. (I can write 2 more pages on this)

Oh ya, and I almost forgot to mention you are at the maximum flow of the fuel system. This means if you have any need to compensate for temps/altitude/etc - it won't work.

Installing a larger injector allows much more accurate control of the injector combating all of the above issues.

So, you guys probably wonder why they just don't install a huge injector in every car?

This is because in order to improve control resolution of the fuel injection system, the OE will ALWAYS choose the smallest injector possible for the application.

For idling and cruising, the more range the ECU has to open the injector up, the more control it has regarding flow.

1ms of variance is much more sensitive on a huge injector vs a small injector. You are still dealing with a mechanical part the has to be powered up, pintle has to move, etc. Factor in the lag time of the injector and this becomes very important!

--

The best option is running the "smallest" injector possible with proper margin at high rpms.

--

This is similar to the OE concept, and why for performance upgrades they normally offer upgrades in small increments (19-21-24-28-32-etc)

The other reason they upgrade in small increments is because with the factory ECU it is sometimes difficult to remap the fuel system regarding a large flow change of an injector.

That's where you get those cars that run like shit at idle with a medium size injector typically. It is the control system that is the issue in that case, not necessarily the injector.

To throw in a curve ball, if your control system is good (proper ECU system, no negative effects), there are some tricks with running an oversized fuel injector for the application. Not being limited on injector timing due to high duty cycles allows the calibrator more options to experiment with. I have done a few cars prepped for E85 where on 93 octane (E10) the duty cycle was less than 50% at redline, but the cars picked up power with some tricks. This was also comparing to previous fuel systems running at 85+% duty though.

328ijunkie,

You should be able to calculate actual injector duty easily if you have your PW logged. It is simple math. If you are already at 85-90% duty, you will find you will eventually be tapped if you keep doing stuff to make more power. In your log, you can also see the injection time drop off on the top end since the VE of the engine is dropping off. That is helping the situation, and typical with M50 series engine.

Can you explain what you mean by these engines not liking sub 12 mixtures? Typically on normally aspirated engines, that could be true, but not necessarily the case always. For example, a richer mixture can control knock at high RPM, which is an issue for these engines around 6000rpm.

I am actually the one who tuned berlow94s Vipec, and while I am not looking at his calibration, I imagine the mixture target matches the ignition advance appropriated for making maximum power safely. He uses his car for road racing, and we actually looked at his detailed ECU logs (that the Vipec has on board) after his dyno sessions to totally validate mixture, knock, and more.
Regarding the power levels he is at, upgrading to the larger injector allowed for much more accurate control of the fuel mixture, especially at high RPM.

berlow94,

There are lots of cars that can idle with 1000+cc injectors at 14.7:1 depending on the control system, the ECUs ability to compensate for flow, etc. I hope reading the above puts that into more perspective for you as well!