M50 Manifold on S52 – TESTED (With TMS Stage 2 & 3)

Tested: S52 vs. M50 Manifold and the TMS Stage 2 & 3 Packages for the BMW E36 M3

by Billy Johnson

M3’s have always been fantastic handling cars but let’s face it; the US-spec E36 M3 is the only generation M3 that did not receive the same engine as its European counterpart. The E36 M3 came to the US in 1995 but did not come with the high-revving individual throttle bodied S50B30, but instead we got a 3.0L S50 motor which is essentially a stroked 325i motor with a revised aluminum head, single throttle body, and a single (intake-only) variable valve timing VANOS system that put out a mediocre 240hp.

In 1996, the motor was stroked even further to 3.2L (S52) but was choked down by a restrictive intake manifold in attempts to bump low end and midrange torque for us Americans. While these changes did result in more torque and mid-range power, peak power remained unchanged from the 3.0L’s 240hp. This also made the S52 feel like it was running out of breath in the last 2,000rpm of its trip to a 6,800rpm redline.

When compared to the 3.2L European S50B32 which pumped out 321hp, had individual throttle bodies, Dual VANOS, and a 7,600rpm redline, the US S52 does not do favorably. It doesn’t matter whether the import cars are from Japan or Germany, here in the States, we always seem to get the short end of the stick.


M50 vs S52 intake manifoldThe M50 intake manifold (right) from the E36 325i and 3.0L M3 is a popular and inexpensive way to improve the high RPM breathing and power of the later 1996-1999 cars. The narrow intake runners on the 3.2L S52 (left) improve mid-range torque at the expense of choking the motor at high RPMs.
The Turner Motorsport Stage 3 system (which includes a larger 3.5” HFM, AFE cold air intake, 24lb Injectors, Shark Injector tune, and Schrick Cams) further frees up the intake of our S52 and provides more fuel to take advantage of the added air to make power closer to what we should have had in the first place.

Rumors and dyno charts fill the internet on the positives and negatives of this swap with claims of 5-20hp gains, huge losses in mid-range torque, and many subjective opinions on the difference in feel. To dispel the myths we did back-to-back testing of the stock 3.2L manifold to the M50 manifold as well as the Turner Motorsport Stage 2 (which includes a 3.5” HFM, 24lb injectors, an AFE cold-air intake, and a tuned Shark Injector) all in the same day, on the same dyno for a true test of this commonly misunderstood modification. For kicks we went a step further and installed a set of Schrick cams with a new Shark tune (Turner Stage 3) in an attempt to come even closer to the European S50B32’s numbers.

E36 BMW M3Our test car is my personal 1999 BMW M3 which has seen everything from track time, to LA stop and go traffic, to a coast to coast drive to end up in North Carolina. The car has a bone stock drivetrain, and 199,000 hard miles under its belt. Prior to this test the only modifications were a fan delete (uses a radiator mounted electric SPAL fan) and a K&N Drop-in air filter.
Page 1 of 5 Next Page
Bookmark and Share
Thursday, January 09, 2014 6:33 AM
Exhaust is going to open this thing up! I bet you see 25hp with tune.
This makes me want to get an E36. They are the best looking cars.
Dusty Duster
Dusty Dusterlink
Thursday, January 09, 2014 6:48 AM
I love me some E36.
Pablo Mazlumian
Pablo Mazlumianlink
Thursday, January 09, 2014 8:22 AM
Nice article, test and car! Have always been a big fan of the E36 M3 and tested many. One thing to note, and what a lot of people don't realize, it just how much faster an M50-swapped 3.2-liter with tune ECU is over stock in a straight line...even moreso than these dynos show.

I've tested the manifold both ways, with and without software. Sure, there's that momentary loss at around 3500rpm but between 6500-7000rpm there is up to a 32whp difference! It's hard to picture looking at your 2nd graph here, because it's with stock software to 6500rpm. But if you can picture the M50 graph continuing straight to 7000rpm without dropping, and the baseline dropping to around 175whp by 7000rpm with the M52 manifold (yes it's like a 45 deg line going down--looks like a top end torque curve)--that's where the M52 mani is super choked.

And that's where the real difference is: not only in the 30+whp in the top end but the ability to keep it in gear longer, to boot. It's a night-and-day difference in a drag race at least, where the midrange does not matter as much. Best bang for the buck!

BTW, here's one of many examples of what an ecu'd M52 manifold looks like from 6500-7000 rpm. With one of those cars you'd still want to shift at 6500rpm in 2nd or 3rd gear-on (as opposed to one with an M50 manifold): http://turnermotorsport.typepad.com/.a/6a0105370140f1970b014e881095d4970d-800wi
Pablo Mazlumian
Pablo Mazlumianlink
Thursday, January 09, 2014 9:01 AM
Also, I assume the car is running top-notch drivetrain fluids (Lubry Moly, Redline, Royal Purple, Amsoil, etc) in engine and also diff/transmission. If not, that is also your best bang for buck, along with iridium NGK plugs (or other good ones). have tested this on 4 different M3s alone with favorable results each time, netting up to 5-10whp peak gains. Of course the gains depend on what was in there previously.

lastly, dyno testing these cars is easy to get consistent, even if you can't do it on the same day (assuming same tank of gas, oil, etc, and you're in similar humidity), so long as the IAT and coolant temps are the same (which therefore net not only the same 02 going into motor but ignition timing). You can monitor IGN, IAT and coolant with a cheap OBD2 scan tool, which has worked great for me. But, while repeatable, these M3s are indeed very finicky if these temps aren't the same). Best of luck!
Mike Kojima
Mike Kojimalink
Thursday, January 09, 2014 11:41 AM
Awesome bolt on test!
Thursday, January 09, 2014 3:17 PM
Nice article and nice gains!

I've been looking for exactly this info (manifold swap).

Monday, January 13, 2014 8:31 AM
The car had recently changed fluids - Redline in Trans and Diff and Shell Rotella T6 5W40.
Friday, March 21, 2014 12:00 AM
I'll venture a well founded estimate that the Conforti Stage 2 software with the M50 manifold and stock cams is costing you 5-10 rwhp and a few MPG, because it isn't written for the M50 manifold and runs too rich when you use it like that.
Here's why I say this:

What are your air/fuel ratios and MAF flow numbers? Did the dyno record these or did you synch up a wideband O2 system like an Innovate while you did the test?
On my '97 OBDII w/M50 manifold, same Porsche MAF sensor, 24# injectors, still stock cams, similar to your setup with Conforti Stage 2 s/w, my engine runs way too rich. I get about 11.5:1 typical, on some days 12:1 at best. Only differences are I added a 68.5mm throttle with Turbohoses intake boot, made my own 3.5" inner diameter MAF tube, adjustable fuel pressure regulator, and have tested with several CAI and a gutted out Euro BMW airbox. The gutted Euro airbox flows best. Best CAI was 32 lb/min. Gutted out Euro airbox with no internal cones or snouts flows 37+ lb.min.
Previously I had Dinan Stage 1 s/w with the Siemens sensor, running at 4.6bar fuel pressure. With fuel pressure adjustments, I could run as lean or rich as I wanted.
Conforti does run clean in normal street driving. With 198k on the car, this passes California smog with a Supersprint exhaust and original stock cats.
With the previous Dinan stage 1, my motor measured 251rwhp and 233 torque, at ~13.1 air/fuel ratio at the RPM where peak hp occured. (my stock baseline on the same dyno a few years earlier was 205 rwhp) Old data shows that when the air/fuel was around 12:1, the dyno showed about 243, so another 8rwhp by getting the air/fuel dialed in.
After switching to the Porsche MAF sensor and Conforti s/w, the air/fuel ratio stays too rich regardless of fuel pressure unless I lean it out so much that the engine runs around 17:1 during warmup in open loop mode. Also I get an intermittent 1188/1189 CE code during warmup.
I get measured and repeatable 5.1 zero to sixty (with 3.38 rear, but needing an upshift to third to reach 60mph), so the car is silly fast for a CA legal street car and run even or better with E46 M3 and 1/335i, but the uncontrollably rich air/fuel ratio and constant negative fuel correction values during highway cruising costs me a few MPG and at least 5+hp at the rear wheels.
If you haven't yet installed a wideband O2 sensor system, then if you are going to play with this level of intake mods and severe s/w remapping, you _do definitely need to.

Feel free to contact me privately regarding the tuning, my experiences may save you lots of time and $$.
Sunday, October 04, 2015 1:55 AM
mpathic, would Turner's Stage 3+ Conforti software correct for this? That package includes the M50 manifold so I'd assume this tune would correct for richness.
Sunday, January 03, 2016 1:36 AM
Not sure what led me back to this article....
@Andalau, Yes, after sending Turner some logged OBDII data, they said that my setup was breathing in as much as as they would expect a car with cams to breathe in, and switched me to the Stage 3 tune. That solved the CE and mixture range problems. Stage 3 tune for cams brought the air/fuel at open loop during warmup back down to no leaner than ~15.5:1, with ~12.5 air/fuel at WOT on good days. No more touchy CE's when I used a Pipercross Viper airbox. The factory euro airbox still has some turbulence. Solved that problem enough to pass the visual part of CA smog by inserting two apple slicers as an air guide into my 3.5" inner diameter MAF tube. Car is more tractable overall with the Pipercross. Maybe has more top end with the factory airbox but would need a dyno to measure. Sold the car last year. Never got back to a dyno again. Did switch to a 3.91 rear with 6 speed OD trans. Unreal sweet. Car reached >152mph and still pulling on the I15 toward Las Vegas.
Another local street prepared E36 M3 with a similar setup dyno'ed 254 at the rear wheels, in a less street legal motor.
Tarik Laaraj
Tarik Laarajlink
Wednesday, May 04, 2016 9:55 PM
Wouldn't a e46 M3 motor swap be about the same price?
Micah McMahan
Micah McMahanlink
Thursday, May 05, 2016 1:21 PM
Mooresville caught my eye...then again I'm typing (on break) from where I designed on those engines you race, Billy ;)

I ought to pop on over to Draco and check it out. Haven't had much time for my toy cars/local scene with the work load.
Thursday, May 05, 2016 10:42 PM
Thursday, May 05, 2016 11:52 PM
An E46 swap done _properly would could thousands more in custom wiring and integration into the cars electronics. The UUC S54 E36M3 conversion reportedly cost around $40k. Lots of people can stuff an S54 into a car and get it to run, but running flat out and being a tractable street engine too is a different level of expertise.
Even with stock cams, my '97 M3 California smog passer with 251rwhp@6300 and 234 torque at 4300, was more powerful up to ~ 6500RPM than a stock S54. With cams, this would beat an S54 to ~7000.
Making power is about how much air you can stuff into an engine. Stuffing in fuel is relatively easy. The S52 with M50 manifold, the right cams, and a 69mm throttle body sucks enough air for the same hp as a stock S54 at up to the same rpms. Problem with an S52 is beefing it up to handle the revs needed to rev as high as an S54.
Post Comment Login or register to post a comment.

MotoIQ Proudly Presents Our Partners:

© 2018 MotoIQ.com