Dyno results S38B38: how to improve S38 max power, cruise, fuel economy

daytonaM5 Discussion starter · Jun 5, 2008

Cyrus · Jun 5, 2008

Excellent post.

I too have the same machine and have done many tests on the S38 including full live mapping on both ignition and fuelling. I have experimented with quite a few off the shelf chips and agree that these cars really need to be properly tuned. Off the shelf chips are only ever any good if they have been properly developed (not some crappy percentage increase of timing and fuelling accross all load points) combined with a car in similar health which was used to develop the chip at the time.

Just a few questions and comments about what I can see from the graphs:

Why are you running a tyre pressure of 32PSi may I ask? You really should be running atleast stock pressures or even higher. Dyno Dynamics clearly recommend this and there is good reason for it.

The RWHP reading is quite low on your car compared to mine and others I have tested which gives around 265-280bhp depending on what map I run.

I also notice your intake temps are very high. What kind of cooling have you got? Even while mapping under load for some time I cannot reach these temperatures.

Having actual intake tempertures as high as this will cause the ECU to retard ignition timing quite alot and hence lose you power. It will also cause over fuelling.

The huge dip in your BHP curve is also a bit extreme compared to the graphs I have seen from my own car and many others.

The fuelling is going as low as 10.6 AFR which isn't natural for these cars.

I would suggest you check your resonator flap is working and also try to see what the ECU is seeing with respect to oil temps, air intake temps and coolant temps while you run the car on the dyno.

The last thing I'd like to point out is that your engine is revving to 7700rpm. Is this correct? If it is then I'd suggest you make sure the stock rpm limiter is put in place.

I get the feeling you've calibrated the dyno incorrectly and this is easily done but errors like this should be avoided at all costs.

I like the way you've used the inlet manifold pressure to talk about intake efficiency. It's something I haven't yet done but will start.

jrobbo · Jun 6, 2008

Cyrus said:
The RWHP reading is quite low on your car compared to mine and others I have tested which gives around 265-280bhp depending on what map I run.

It's fairly well accepted that Dyno Jet dynos (seem to be widespread in the UK), read a fair bit higher than the Dyno Dynamics dynos like DaytoneM5 has. Roughly speaking, 236hp on a Dyno Dynamics dyno is about the same as 280hp on a Dyno Jet dyno. There a quite a few calculators around on the net to convert, here is one that I found:

http://www.to4r.com/calcs/pwtq.php

Cyrus · Jun 5, 2008

sorry....forgot to talk about the topic itself -

To get max power on an S38 is the same method as on any other NA car.

Advance ignition timing and fuel to give 13 AFR.

With a dyno that holds load and is as sensitive as the dyno dynamics using live mapping can give maximum power way before detonation could ever think of occuring. However, it's still not wise to try and extract maximum power from these engines as under very high altitudes problems can arise.

The mapping on the S38's is so conservative it's unreal. The timing is so under advanced compared to other M powered engines.

Once you tune a few you begin to understand why BMW did this and part of the reason is to do with noise rather than reliability.

The S38 really starts making alot of noise (good noise) once ignition timing is advanced. Where the engine may be almost silent under 30% load @ 3000rpm it will now be a little more intrusive and this from a manufactueres point of view may not be what all of it's customers want. You also have the pass by noise regulations coming into effect.
The problem is that the S38 has the potential to be one very very loud engine.
There is however a considerable gain in efficiency if the main load sites are mapped properly. Leaving part load fuelling alone and just advancing ignition timing gives a big jump in fuel economy.

Max Power and torque is wel improved with peak gains of around 10-20 bhp achieveable. Through the rpm range under full load similar if not more is available.

As an example my car started from 320bhp and now sits at 340bhp with a healthy 310 lb.ft torque with alot more power under part load.

There is loads more to come from an S38 in terms of driveability, part throttle power, smoothness. Little bit more in terms of peak power. In practical on road performance the S38 has the potential to be quite a bit quicker due to the smoothing out of the torque curve.
Better part throttle (cruising) fuel economy is just a massive bonus.

The biggest problem is actually finding someone to live map these cars properly. There's alot more to it than just sitting on a few load sites and dialling in changes in fuel and timing. It requires alot more attention to detail and in the ideal world the maps should be switchable just incase you do have to encounter very high altitudes or poor quality fuel. In those scenarios you can switch back to the standard map just to be on the safe side (it's unlikley you'll cause much damage unless the fuel quality is really bad).

I only know of a hand ful of people who can map properly and with responsibility. The others either don't do the job fully (leaving out loads of load sites) or don't keep an eye on the many variables whilst mapping. if this is done then that map is a waste of time and/or potentially dangerous for the engine.

Alan Archer · Jun 5, 2008

I always had the understanding that the S38 ignition timing is "conservative"
as the motor has no knock sensor..
IE> no retard control over detonation.
Ditto overfuelling (slightly) to the same end result...protect the motor at
all costs.

236hp corrected or not seems very low to me.

Roll on July 5th and we should have plently of graphs to add to the collection!

Regards,

Alan.

Seaneee · Jun 5, 2008

My 3.6 Results

<TABLE class=tborder id=post1290828 cellSpacing=0 cellPadding=6 width="100%" align=center border=0><TBODY><TR vAlign=top><TD class=alt2 width=175></TD><TD class=alt1 id=td_post_1290828>data
<HR style="COLOR: #ffffff" SIZE=1>Quote:
<TABLE cellSpacing=0 cellPadding=6 width="100%" border=0><TBODY><TR><TD class=alt2 style="BORDER-RIGHT: 1px inset; BORDER-TOP: 1px inset; BORDER-LEFT: 1px inset; BORDER-BOTTOM: 1px inset">Originally Posted by Nikosx

Great! Im excited to see how she did.
</TD></TR></TBODY></TABLE>

RPM Std Power Std Torque Mod Power Mod Torque

Sorry don't have graph electronically.

The comparison is with and without Conforti chip. Rear wheel , uncorrected, reputable independant dyno.

How much does a dyno dynamics chassis dynamometer cost , if you don't mind me asking. I looked at their website and they don't look like the kind of thing you would just have hanging around in the garage.
</TD></TR></TBODY></TABLE>

saturn5 · Jun 5, 2008

Here are the results that came with my 3.6 when I got it.

The valve clearances were WAY out. Exhaust side gaps were mostly 0.47 and overall I think 20 shims needed doing. I think it's a fare bit down on power.

What do you think? As I know nothing about tuning, maps, perameters etc.

Would be nice to get a comparison done when I finally get it running.

A.W.O.L · Jun 5, 2008

91 3.6 with Schrick 280 cams, D'sylva chip

Car was in limp mode as the battery died the day of the dyno and wasnt finished recalibrating, hence the jump in power over the first pull.

269whp with 247ft/lbs

I will redyno soon once the valves are adjusted and the car isnt in limp mode

GK M5 · Jun 6, 2008

very interesting post with details on what is good and what isn't. makes me think in the right hands a bit of tuning isn't perhaps so bad and would improve the car a fair bit.
however who are the ones who do a good job on the remap and any near belgium?

daytonaM5 Discussion starter · Jun 6, 2008

I use WINOLS to program oem sofware. What we do is take out the eeprom, read it, put it on WINOLS software, reprogram/make changes/load the new program to the eeprom, and put the eeprom back. See the changes.

This explanation wont be liked by tuners, as i am explaining the details, but this is how it works on S38B38. If it was a S54 (M3), you dont need to remove the eeprom. just load the program through the obd port.

The main issue is to know where the full throttle and part load fuel and ignition maps are located inside the oem software, and this is found with winols software.

BoSJo · Jun 6, 2008

Interesting

There is supposedly another remapping program under development :

http://forums.bimmerforums.com/forum/showthread.php?t=996261

But still you will need the dyno to get it right.

Where would i get my hands on a eprom reader/writer HW interface?
Thinking of playing a little with the 525 before i even touch the S38.

Do you know if it in anyway is possible to get live data out of the E34 brains?
I have a very skilled micro electrician in the family, but I don't wanna bother him if it's not possible in any way, and I don't even wanna think about how many hours it would take if he had to start from scratch!

Cyrus · Jun 6, 2008

I dynoed my 3.8 on the Dyno Dynamics today.

Made 332 bhp which I think is very realistic. Engine was breathing in around 25 degrees C with ambient temps of around 19.

I played around with the ECU a little and got this upto 338bhp but gained almost 25 bhp at around 5000-5500rpm.

Then came my special intake system.......final figure of 341bhp.

I'll post the graphs tommorrow.

Mapping on these cars MUST be done live in my opinion. I have found every single pre programmed chip to be over advanced in some area of the rev range or under advanced in others.

I should point out that my car never used to make this sort of power before I had the valve clearances properly setup.

farrell · Jun 7, 2008

I agree.
The mapping should be done live ideally on a bespoke basis.
Every vehicle is different, even M5 which is especially so after 15 yrs of use.
A base map for the main load sites can be achieved on the rolling road & can then be optimised on the road where ambient temp,
air flow & pressure is real world.
However, mapping with greater heat soaked air is no bad thing initially.
Alot of time can be saved by inital setting up on a rolling road that is operated
carefully & with operator skill.

Altitude & atmospherics are rogue elements that mean a conservative approach to the overal calibration.

RPM limits beyond 7,500 rpm with std crank, bolts & production balance !
Not recommended.

Ascertaining the true condition of the engine prior to any planned tuning work
is recommended.

Cheers

nebpor · Jun 7, 2008

farrell said:
Ascertaining the true condition of the engine prior to any planned tuning work
is recommended.
Cheers

There speaks the real words of wisdom :applause:

Anything even remotely past it's best gets a hideous amount of strain when it's being mapped on a dyno!

WMWNWMN · Jun 7, 2008

daytonaM5 said:
vacuum has increased to <?xml:namespace prefix = st1 ns = "urn:schemas-microsoft-com

</st1:metricconverter>1,0 inches of mercury from <st1:metricconverter w:st="on" ProductID="0,1 inches">0,1 inches</st1:metricconverter> of mercury. This shows that OEM M5 intake is restrictive.

think this would help?:thumbsup:

daytonaM5 Discussion starter · Jun 7, 2008

Oh my God, is this is a custom made air box? Who sells them?

Cyrus · Jun 7, 2008

daytonaM5 said:
Oh my God, is this is a custom made air box? Who sells them?

That carbon airbox will be so loud on an S38 engine that would think it would be pretty much impossible to run the car as a daily driver.

I have a carbon intake system on my car which just replaces the intake muffler and that is LOUD under load!

The standard plenum in my opinion is a master piece as far as all round performance and torque are concerned. The resonator flap is just pure engineering excellence.

jrobbo · Jun 7, 2008

The bottom line is that I never read too much into the absolute HP figures touted by various dynos, the simple fact that we have such an extraordinary range of outputs in this single thread is evidence enough that most of them are just plain wrong. Dynos are good as a relative reference, but not of mush use as an absolute reference, and i've said that many times here in the past.

FWIW, my last dyno run was 192kW, for a 3.6 with 100,000kms on the clock, S38B38 cam gears, EAT chip, and Eisenmann exhaust

John

Cyrus · Jun 7, 2008

There are some dyno places and operators who use these machines to great effect to give surprisingly very close to factory figures.

Operating a dyno is not just about strapping a car down and doing a power run. There is so much more to it.

The lack of attention to detail by 95% of dyno operators creates this problem of varying figures.

Sure, power figures will differ slightly between dynos of different makes but not by much unless your taking about a dynojet.

My dyno reads within 1 % of many other dyno's which are used properly and scientifcally as possible. These are different dyno makes too.

Here are some examples of stock cars I have dynoed:

E46 325i 192 bhp - dynoed at 193 bhp
E92 M3 414bhp - dynoed at 411 bhp
E39 528i 193 bhp - dynoed at 191 bhp
E46 M3 338bhp - dynoed at 335 bhp
E39 M5 400bhp - dynoed many and most between 395 and 410 bhp
E31 840i 286 bhp - dynoed at 280 bhp
E92 335d 286 bhp - dynoed at 290 bhp
..... etc etc .....the list goes on.

The setup and cooling are essential to get these figures. Not only this but intake temperatures have to be monitored from the ECU's point of view. Coolant temps and oil temps need to be analysed before a dyno run is attempted.

Tyre pressure and strap strength play a role also. Get these wrong and you'll get tyre slippage which will massively skew the figures and the shapes of the curves.

jrobbo · Jun 7, 2008

Cyrus said:
E46 325i 192 bhp - dynoed at 193 bhp
E92 M3 414bhp - dynoed at 411 bhp
E39 528i 193 bhp - dynoed at 191 bhp
E46 M3 338bhp - dynoed at 335 bhp
E39 M5 400bhp - dynoed many and most between 395 and 410 bhp
E31 840i 286 bhp - dynoed at 280 bhp
E92 335d 286 bhp - dynoed at 290 bhp
..... etc etc .....the list goes on.

So are the figures that you have been quoting actually measured at the rear wheels, or extrapolated crankshaft figures? From the table above, i'm assuming the latter, which would explain your high figures. The figures that I quoted for my car were was was measured at the rear wheels, and I'm assuming that the same thing is the case for DaytonaM5

John

Cyrus · Jun 8, 2008

Mine are estimated flywheel figures. I can quote wheel figures too but BMW do not quote these.

I am trying to show the correlation of what BMW quoted for these cars and what I got.

If you look at the variance it's only within a few % +/-.

My point is that dyno's can be very close even when using crankshaft figures if used properly.

I don't personally get to hung up on the final peak figures, I use the dyno more for comparing different maps, upgrades and live mapping. It's the difference the changes make which count and enable you to make a decision as to whether or not the changes are worthwhile.

daytonaM5 Discussion starter · Jun 8, 2008

is there a way to find out if the resonator flap works 100 %? anyway to diagnose it?

Rob Plummer · Jun 8, 2008

Great thread guys. Just replied to get me subscribed to it!cherrsagai

Lantz · Jun 8, 2008

Yes, the 3.6 does a self test on start
up.

For the 3.8, warm up the car normally, open bonnet and rev past 4k rpm, if the resonance flap does nothing it is not functioning properly.

Lantz

Ferrari-Bmw · Jun 8, 2008

In both cases you are missing the SPEED-SIGNAL,,
it was like that on my S38B38,,

AND,,,,, if the flaps is not working ((the speed-signal is missing )) you are loosing a lot of performance from the engine ,, :grrrr:

jrobbo · Jun 8, 2008

Mine are estimated flywheel figures. I can quote wheel figures too but BMW do not quote these.

I am trying to show the correlation of what BMW quoted for these cars and what I got.

If you look at the variance it's only within a few % +/-.

My point is that dyno's can be very close even when using crankshaft figures if used properly.

I don't personally get to hung up on the final peak figures, I use the dyno more for comparing different maps, upgrades and live mapping. It's the difference the changes make which count and enable you to make a decision as to whether or not the changes are worthwhile.

Yeah, that makes sense, that explains the discrepancy. There is no doubt that a dyno can be close, but my point is that without any defined and enforsed standards for calibrations, how do you know that the dyno is accurate? Here in Australia, all scales in Delis and Butchers need to be calibrated every 6 months, and then sealed with a lead seal. The technician who is calibrating them needs to be licenced as a Weights and Measures tecnician, and has to re-sit an exam every 3 years. The weights he uses need to be cabilrated every 3 months against standard weights that are kept under luck and key, and which are in turn subject to random checks from the authorities. All this ensures that when you are buying 250g of Mozzarella for your pizza, you are getting 250g. Same thing goes for petrol pumps etc.

But Dynos, at least here in Australia, are not treated the same way. What are the defined standards for calibration? What are the qualifications of the person doing the cailibration? How are the measures calibrated?

But you are 100% right, no point getting hung up about peak figures, dynos are far more useful as relative measures as you've pointed out

Cheers

john

daytonaM5 Discussion starter · Jun 9, 2008

Cyrus,

can you start a totally new thread with photos of your intake system?

Cyrus · Jun 10, 2008

My Flywheel Figures:

Made 283bhp @ Wheels with the chip and Carbon Intake.

The dyno RPM calibration was not 100% (you never get it right when it's your own car being dynoed!!). Car actually revved to 7200rpm stock with deCAT and around 7350 with the intake and chip.
This effects the torque figure / graph a little by reducing it because I'm showing to rev a little further than I actually am.

jrobbo · Jun 10, 2008

Did you post the wrong graph there Cyrus? That one reads Flywheel power.

Here was my latest graph for my 3.6 with S38B38 cam gears, EAT chip and Eisenmann rear box, 192rwkW (257rwhp):

Cyrus · Jun 10, 2008

jrobbo said:
Did you post the wrong graph there Cyrus? That one reads Flywheel power.

Here was my latest graph for my 3.6 with S38B38 cam gears, EAT chip and Eisenmann rear box, 192rwkW (257rwhp):

No, that's the right graph. I can quite easily post the correlating wheel power and wheel torque graphs also. They will show the exact same gains.

Just for reference my car started with 274 rwhp and ended up with 283 rwhp.

To get 257 rwhp from a 3.6 is pretty damn good! That's approximately 310-320bhp at the crank.

More importantly the shape of your graph is very good. A nice full power delivery. I would expect your car to be pretty quick on the road with a nice delivery under full load.

nebpor · Jun 10, 2008

Cyrus, did you map the chip, THEN add the airbox, or was the chip mapped with the airbox on the car?

I know you know this, asking for the benefit of others - if you've substancially changed the induction I'd expect you to be able to map with quite a bit more fuel once it was on

Cyrus · Jun 10, 2008

nebpor said:
Cyrus, did you map the chip, THEN add the airbox, or was the chip mapped with the airbox on the car?

I know you know this, asking for the benefit of others - if you've substancially changed the induction I'd expect you to be able to map with quite a bit more fuel once it was on

The chip is just something I had and decided to fit it. No mapping has been done to suit yet.

The chips ignition map is probably one of the best out there but still way out from optimal for most load sites.

I am just waiting for my Dastek Uni-Q to arrive and then I will live remap the car properly for fuel and timing for approx 260-300 load sites. This will take me around 8 hours. I'll post the dyno results but this will only give the results for 28 of those load sites! To really appreciate how the car drives after tha changes you'll need to come and drive it. A properly live mapped car is very impressive.....especially S38's, S14's and S50 engines.

Cyrus · Jun 11, 2008

Nebpor,

I have a good amount of experience with after market ECU's and have a car with MBE management (which is similar if not better than DTA).

On a car with a fairly good ECU like the S38 there is really no gain to be had with after market management. The standard ECU fuels for 14.7 AFR on part load which is pretty much what I would have fuelled for with after market management.

Anyone can bolt on an DTA or MBE, that's easy. Anyone can create a half decent base map. However, mapping the ECU properly to the same level as BMW is nigh on impossible. On the last car I mapped on after market management I found myself doing exactly what a manufacturer would have done. I had to wait for different conditions to arise and then see if the car still ran smoothly. Cold starts and smooth running under different temperatures is not easy to do properly. Then, different altitudes is the hardest part. You have to physically go to different places to see if any of the compensation maps need adjusting.

You finally end up with a car that basically does not run as consistently as one with a standard ECU.

The standard ECU has more than enough resolution so you can't really gain anything there either.

Aftermarket allows you to have different engine setup strategies such as running without MAF but my experimentation clearly showed me there is no gain from losing the MAF on it's own.

After market management is used in two scenarios generally:

1) Where the tuner cannot tune the standard ECU
2) The engines is so vastly modified that the standard ECU just is not good enough (rarely!)

The Uni-Q is the new generation of Unichip made by Dastek. It is completely different to the old stuff. I had a chance to get the old one but was never overly impressed. The new one however caught my eye so I decided to take it on and it's the best decision I have ever made.

Cars can retain their standard ECU and along with it all of massive amount of development given to the mapping and be live mapped.

You can convert a MAF based car to run pure AlphaN or MAP sensor so FI conversions are not a problem. There is almost more resolution than most standard ECU's so that's not a problem either.

Best thing....you can have upto 5 switchable maps. You can use that for almost anything! You can have a map for every gear if you wire in switches to the gear shifter (this is massive over kill), you can have a standard map, tuned map, maps with different boost levels and loads of other differently configured maps. You can have a wife map too which limits rpms to 3000rpm and take out all of the power!:7:

I have already done trials on E46 M3's and E39 M5's and it's excellent.

Dyno results S38B38: how to improve S38 max power, cruise, fuel economy

Attachments

Our Vendors

Top Contributors this Month

Recommended Communities