RomRaider

Hey everyone,

the following shows what i did to a M52B20 from 1994. Some may find it intersting to see.
Backstory: In 2009 i wrecked my E36 320i which was my first car. I took the engine, subframes etc. out and put it in storage. In between i had a clapped E30 325i cabrio and a E30 318is (which i still own) before i got another E36 (also a 320i).
Because the B20 was pretty slow compared to the rebuilt M42 i started looking at some options.
The obvious one were to boring/expensive imo...and there is another B20 just sitting around collecting dust.

The route i went with is stroking the B20 by changing the crankshaft. Having the B20 block/head designation is also kind of useful here...
Using the M51D25 crankshaft (which is way cheaper than all the other M52 cranks) gives 82,8 mm stroke instead of 66mm. This results in 2497cm³. With B28/B30 135 mm rods and standard HG the CR is 11,1:1.
What changes significantly is the squeeze effect at TDC compared to the B20.
Also using the smaller 80 mm Head is reducing top end power compared to a stock M52B25.

I started the disassembly of my old engine. It had 194.000 km but looked very good inside. The cylinders showed minimal wear. The bearings were nearly perfect. Pistons and all journals measured ok.

From random pictures from the internet i measured out the crank stub of the M51 to check if the valvetrain can be alligned with the M51 keyway. The M51 has only one keyway in slightly a different position. I wasnt to shure but bought the crank anyway for 100€ in perfect state.
A modification to the sprocket was neccesary to have some meat on the keyway for allignment before the crank-bolt is torqued.

6x 135 mm rods were 95€ but they came from 3 different engines so the wheigt classes were all over the place. I had to get quite aggressive on grinding to get them within 1g.
The rest of the assembly was done with minimal financial effort: New rod bolts, bearings, headgasket and a lot of rtv was used.

To swap the engines i dropped the complete subframe to avoid messing with the steering rack/allignment to keep cost down.

Of course it would not start at first. Two problems occured:
The injectors went bad but one from beeing stored for a long time.
The insulation of the crank angle sensors wires was gone on both units.

The car finally started with the B20 tune and was even driveable. This is the point where i have to say thank you to the developers of the MS41 tuning! Otherwise this endeavor would not have been undertaken.

The car pulled strong and drove fine. The added torque is real nice with the short gearing.
For some reason i got crazy knock between 1000-2000 rpm at 250 load even after reducing the ignition map excactly by the logged amount of knock.

After some time i changed to the bigger MAF-setup of the B25/B28.

Right now i have two main problems:
1. knock between 1000-2000 rpm at 250 load occurs no matter how much i reduce the advance in that area (less than 0° of advance)
2. At wot from 4300 to 5500 my log shows lower injector pw and higher timing, car loses a lot of power there, i cant figure out why. Vanos is turned off for now and ect/iat compensation shouldnt be doing stuff like this.

Looking forward to you guys feedback/tips


Sorry that the pictures are in reversed order...

You may have knock problems because the piston quench area is now further away from the head than it was previously. This can make the engine lot more likely to knock than it was before. This is also the reason why thicker head gasket is the worst way to lower compression ratio because the engine can be even more prone to knock with lower CR that it was with higher.

Also have you tried what happens if the vanos is working normally. meaning it's on at 1000 to 4000ish RPM.

yes that is a possibility. Compared to stock the quench height is 1,6mm bigger.
Piston to Head surface:
Stock B20: 2,04 mm
my stroker: 3,64 mm
Stock B25: 1,54 mm
Stock B28: 2,92 mm

Also the combustion chamber design is worse than the 84 mm engines. The B20 Pistons are domed versus most others which are flat.

I tested with Vanos stock before. same issue. I may have to live with it...

But this is not the biggest issue right now i think. If you take a look at the log attached. You can see the crazyness going on from 4300 to 5500 rpm. I cant figure out what makes the ecu think it suddenly need less fuel and more timing. It is for shure not following ignition and fuel maps.
Is there a hidden max load Value somehere which i cant go over with the b20 software. Maybe it thinks the maf is s*** cause of the high load. It happens in the area iam logging the highest load.

That looks same as the engine load overflow in turbo engines. But at much lower Load. Do you have log with MAF airflow and voltage? You should try to use b28 full flash and change the maps to work on your engine.

Thanks for your suggestion! Changing to the B28 Software may have solved BOTH Problems!
I would say right around 600-610 load the M52B20 Software has its limitation.

Knock adaptions were quite different now as well. I still need to do a longer drive but so far i have gotten way less knock than before.

...my stroke of 82,8 mm is now also way closer to the B28s stroke, this may have been problematic for the knock detection of the B20 Software

I can confirm that the software change to the B28 oftware is the way to go. Knock reading seems to work good. I get only sporadic 1-3° on some cylinders.

Rest of the tuning went also ok. Looks like the car is completely driveable now =D

Vanos helps only from 2600-4000 rpm.

The power is definately increased compared to the B20. Even on higher rpm i dont have the feeling its falling on its face. Torque down low is really nice now. 3000-5000 rpm feels really good!

Thanks everyone for making this possible!

Hi everyone,

i am exploring options regarding machining M5X blocks.

Did some measurements on a M50B20 non-vanos block (casting 1735813). There is a step in the cylinders thickness just some millimeters below the deck. It has a cylinder wall thickness of around 5,3-5,5 on the thinner part.

For comparison the S52 Block shown below has a similar step where the cylinder wall gets thinner below the deck. Iam guessing that M50B25s are very similar/same as the S52.

Can anyone provide some information on the M50B25 or S52s cylinder wall thickness? Has anybody a spare block to take some quick mesurements? It would be much appreciated!

Looking forward to your replys =)

Quick update on the stroker.
Limiter @7000, no issues so far with the M51 crank.
Its now running since 16.000km mostly street and 5 trackdays:
https://www.youtube.com/watch?v=4k6yNThXgzI

What is the configuration of the motor today? Fuel injector, camshafts, fuel pressure regulator 3.0 or 3.5? Can you please download the current partial 24kb here?

Absolutely everything except crank und rods is from the M52B20

Nice slide and catch towards the end there
https://youtu.be/4k6yNThXgzI?t=226

Did you try to fit the b25 or b28 head on b20? With this stroker the perfect gain is the bigger valves. What you mean about that?
Congratulations car runs fine!
Can you send posted your e-mail for ask you some thinks?

84mm head on 80mm Block is impossible because the valves would hit and flow would not be increased that much if you get closer to the circumference of the bore. Increasing valve size on a constant bore+valve axis position gets less efficient the more extreme you go. But of course there is some potential.

According to the logged load the stroker is not that choked even on high rpm. I rev to 7000.
Also when looking at the m54b22 (same Valves as b20) the valve size is good for 170 hp with a generally better breathing (Variable-Intake, longer duration cams, dual exhaust) engine.

A next step for the stroker engine could be cams and dual exhausts all the way to the back.

im gonna open my m52b20 and im thinking of doing same thing here
so i need:
- m51 crank
- 135mm rods
what about the MAF ??? can i keep using the b20 one ?? an m50b20 intake manifold will be a nice add ??

You should do the opposite Using B20 head on B25/B28 engine gives you higher CR...