"In this entry, we hit the dyno and tune the S85-EVO. I establish fundamental dyno knowledge, set our hypotheses and provide answers based on our experiments and in-depth analysis."
- Matt 

Understanding dynos. 

Since the start of this Series, I've had conversations with various owners of S85s regarding the Sxx-EVO programs. I sought to gauge reactions and questions to create the right content.

These conversations were evidently focused on dyno numbers, yet the overall misunderstandings on what dyno numbers actually mean was striking. I've lost count on the claims of factory freaks and unicorn'esque 500whp with bolt ons. 

To set the stage, an introduction to dynamometers is required.

Fundamentally, dynos are devices engineered to measure torque relative to rotational speed. In our use case, they are used as a tuning tool: we're looking for a before/after comparaison in a controlled environment to optimize engine management parameters.  

Chassis dyno types.

Most common to the automotive aftermarket are the chassis dynos, often referred to as "rolling roads". There are 2x primary sub types: inertia and load based. Dynojet is the common brand for inertia, while MustangDyne is the common brand for load based. 

Dynojet (DJ).

The inertia type is straight forward: engine torque is derived from the car's ability to move the rollers to speed. Speed is converted to RPM and the graph is plotted.

Operators will use the gear closest to 1:1 ratio to eliminate the transmission's torque multiplication benefits. In an S85 SMG3, this would be 6th gear with a 1.00 ratio. 

As there are no operator inputs, DynoJets tend to provide the most accurate comparative tool between dyno when numbers are corrected. 

Dynojets will tend to read the highest across chassis dynos with an approximate 15% drivetrain loss for front engine, rear wheel drive and manual transmission cars. 

MustangDyne (MD).

Load based dynos offer additional flexibility as they can account for variations in load to simulate various driving conditions.

The MustangDyne's load cell adjusts load per gear. Consequently, the gearing's torque multiplication is nullified. The gear selected will solely influence the length of the run. If you're interested to learn more, MustangDyne has a 431 page manual here.

In short, they're not good for marketing: MustangDynes will read lower by 5 to 10% compared to Dynojets, and a total drivetrain loss of approximately 20%.

This is the dyno we selected for the development of tuning maps for the Sxx-EVO program. My choice was driven by operator experience and dyno accessibility. 

Notes on correction factors. 

Understanding correction factors is critical to properly assess dyno charts. 

Correction factors are determined by the Society of Automotive Engineers (SAE) using standardized weather parameters such as pressure, temperature and humidity to provide comparable figures for accurate comparaisons. 

In North America, we use SAE and STD correction factors. These standards will be selected by the dyno operators. Marketing savvy operators know numbers sell: DynoJets and STD correction factors are the steroid placebos of the dyno world. 

The increase in power will vary from to 2 to 3% from SAE to STD corrected numbers.

	SAE	STD
Pressure	29.60 InHg (100.2 kPa)	29.92 InHg (103.3 kPa)
Temperature	77°F (25°C)	60 F (15.5°C)
Humidity	36%	0.00%

As an example, the dashboard below showcases a Weather Correction Factor of 0.962. It is automatically provided by the dyno's weather station using sensors to continuously update the coefficient. The below 1.00 factor indicates our uncorrected numbers would read higher than average.

The uncorrected numbers were useful to demonstrate the MSS65's ability to adapt to colder temperatures. With the +/- 10C ambient temperatures, we experienced a +/- 5% power increase.

All of our MustangDyne dyno charts posted in this entry show weather corrected numbers using SAE standards. DynoJet charts will indicate the correction factor on the top right of the chart.

Notes on stock baseline. 

I have never dyno'ed my S85 in stock trim, neither have any other stock S85 been on this dyno. As such, the stock numbers provided in the table summaries will be a calculated figure from drivetrain losses.  

Nonetheless, upon scouring forums and social media, you'll find approximate drivetrain losses per dyno type will line up with theoretically calculated numbers. As an example, this thread offers realistic Dynojet numbers showcasing differences in correction factors for what is otherwise an original M5. 

Stock M5s will dyno +/- 400whp on a MustangDyne and +/- 430WHP on a dynojet with SAE weather corrected numbers. These are the numbers I used as the stock baseline. 

Notes on FBO baseline. 

To set the baseline to a tuned S85, we used InnovAuto's 1 series drift car fitted with an S85B50.  

Their V10 is considered Full Bolt-On (FBO): it is equipped with catless headers and exhaust system. The original plenum was installed without side intakes nor any filters for optimal performance (not reliability). 

The car uses the MSS65 DME as a stand alone. It was performance tuned on the dyno for 91 octane pump fuel with a hard rev limiter set at 8,500 RPM.  

Most importantly. this car had been tuned on the same dyno with the same operator - providing a reliable comparative baseline to evaluate performance gains.

The run-of-the-mill FBO upgrade package provides on average a 10 to 12% power increase over stock, translating to +/- 50whp over stock on most dynos. 

This was translated on the dyno chart below.

Peak horsepower of 450whp was reached at 8,000 RPM. The relatively flat torque curve has the common dip in the mid range that leads into peak torque of 340wtq at 6,100 RPM. 

The original S85's camshaft durations run of out of breath by 8,000 RPM. There is no additional horsepower gain beyond. Nonetheless, increasing the rev limiter to 8,500 is useful to benefit from gearing advantages and additional power under the curve.

Hypothesis 001: Can the MSS65 manage the high duration camshafts on its own?

S85-EVO baseline. 

Message boards of yesteryears have sporadic threads with comments about camshafts with claims that they do not provide any gains on the S65 and S85. 

To test this hypothesis, I uploaded an AlphaN tune with stock ignition and VANOS timing tables along with the factory 8,250 RPM rev limiter. The car was driven to allow adaptations to settle. 

As a reminder, AlphaN is an engine mapping protocol and does not provide power on its own. For detailed understanding, read the entry "Alpha V10".

S85-EVO untuned chart. 

The chart below shows our best run with the AlphaN mapping with standard parameters.

Peak torque of 345 wtq was achieved at 5,750 RPM and immediately dipped and trended downwards unevenly. The torque dip between 3,500 to 4,500 of a standard S85 was accentuated along with the torque ramp up. 

The rough torque curve generated an uneven power curve that cut off at the soft rev limiter of 7,900, approximately 300 RPM before fuel cut off.  It managed a respectable 448whp. 

This is more than likely where previous attempts at fitting cams to an S85 ended.  

Results 001: No.

The chart below compares the untuned S85-EVO with the tuned FBO S85. This comparaison was made to highlights the challenge of tuning ignition timing for high duration camshafts. 

It did not yield meaningful gains - and lost a significant amount of mid range torque. 

The MSS65 cannot meaningfully adapt to new duration parameters on cams as the entire valve timing is mechanically different.

Instead, it will adjust valve timing based on knock, combustion quality, intake temperatures and other factors in an attempt to manage the high duration camshafts within the original parameters.

This chart is our short answer: no, cams did not provide gains.

Hypothesis 002: Can the MSS65 be tuned for high duration camshafts?

Addressing misinformation. 

Prior to diving into the tuning, I will address various comments I have read online about MSS65 and MSS60 VANOS timing tables being encrypted or simply unavailable. 

I cannot comment on where that idea comes from. My observation is they tend to come from individuals without significant experience with these engines and DMEs.

Ultimately, this is misinformation.

A good tuner will have those tables decoded and labeled to understand and adjust each parameters in relation to one another. As an example, here is my M3 V8 and its MSS60 being tuned for its S65-EVO program.  

Theoretical tuning. 

Prior to the dyno session, we developed a base map for the S85-EVO using the technical data provided by Schrick on valve lift and duration versus the original camshafts. 

Tuning for high duration camshaft is focused on valve overlap: the moment when the intake and exhaust valves are opened simultaneously. In a nutshell, we are looking to optimize airflow coming into the combustion chamber and the scavenging effects for exhaust gases to exit the chamber post combustion.

The higher duration camshafts give us the potential to maximize those parameters for high RPM use, unlocking top end power. To make use of the higher durations, we increased rev limiter to 8,500 RPM - the usual rev limiter for conventional off the shelf maps.

Results 002: Yes.

Preliminary tuning gains. 

Prior to the first runs, we proceed with warming up the car to its operating temperatures and begun testing for partial throttle complete runs. This allowed us to adjust timing parameters before going wide open throttle.

The gains over the original map were staggering: 493.8 whp at 8,130 RPM and 362.7 wtq at 5878 RPM. 

Most importantly, the torque curve was smoothened out, making for a steadily increasing horsepower curve. 

A keen eye will notice a unique feature of the power curve: the rate of horsepower increase didn't falter as we hit the soft rev limiter at 8,150 RPM. This was far different to original cams that peak at 7,800. 

We concluded more power could be found higher up in the rev range.

Hypothesis 003: Will we continue making power at higher revs?

Notes on soft rev limiters.

The MSS60/65 DMEs integrate soft revolution limiters that pull ignition 200 to 300 RPM prior to complete fuel cut off. The exact RPM will vary based on various data inputs such as water and oil temperatures, gear selection, etc. 

The soft limiter's purpose is to protect the internals from abrupt, rapid changes at high revolutions.

This is visualized on the charts with the steep downward slopes on the horsepower and torque curves. While we increased rev limiter was increased to 8,500 - horsepower peaked at 8,150 at the soft rev limiter.

To better understand the soft limiter and fuel cut off dynamic, I've uploaded the video below upon testing the 8,800 RPM limiter. By the mid way point between 8,000 and 9,000, you can hear the DME pulling power and letting it rev out to 8,800.

It appear as if we are revving to 9,000 - we were not. 

Results 003: Yes. 

Our next test was to evaluate whether the high duration camshafts would continue making power at higher revs. We increased the fuel cut off from 8,500 to 8,800 RPM. This would allow us to test for an effective 8,500 RPM soft rev limiter. 

The increased of revs correlated to a steady increase in power: we crossed the 500whp mark with weather corrected numbers, hitting 522whp uncorrecte at 8,420 RPM. 

To put the numbers in perspective: 500whp on a DynoJet is an extremely strong S85 with high likelihood of fudged correction standards. From my research, 500whp SAE corrected on a MustangDyno is unheard of for an S85. 

Sustained rate of increases. 

The charts below compare the final S85-EVO tuned run versus our S85-EVO baseline with the original engine map on AlphaN protocols. 

Previously, we gained +/- 30whp from 7,000 to 8000 RPM with a 3whp per 100 revs. This time around, the extra 300 revs provided us with +/- 8 whp for a rate of 2.67whp per 100 revs. 

The very slight decrease is due to the heat building up into the S85's cooling system after 2 hours of intermittent full throttle runs.

Our tests indicate a strong likelihood more power could be achieved by revving higher. We chose not to for reliability purposes. This is where our testing ended on this round of dyno tuning.

Notes on higher revs & reliability. 

When evaluating reliability for higher revs, there are 3x fundamental variables: piston speeds and inertia.

From the numbers below, it can confidently be stated the S65/S85 have the underpinnings to sustain the slightly increased revs. It would be prudent to reduce maintenance intervals on oil changes, spark plugs and other engine consumables.

Valve springs. 

The original S65/S85 valve springs have been known to fail over time. The S85-EVO package includes uprated and lightened beehive valve springs & retainers to eliminate valve float and potential valve to piston contact.

Piston speeds.

I've calculated piston speeds for BMW M high revving naturally aspirated engines. The formula is driven by stroke length and RPM as determined by rev limiters. 

Engines	Piston Speed (m/s)	Redline
S85B50	20.60	8250
S62	20.80	7000
S85-EVO	22.00	8800
S54B32	24.30	8000

Inertia. 

As piston speed increases, inertia increases exponentially. Inertia is driven mostly by piston weight and stroke length. As a comparison, the S85-EVO's original BMW pistons are of similar weight to the S54 with lower piston speeds by 2.3 m/s. 

The S65/85 use lightweight aluminum cast piston of 92mm diameter traveling over a 75.2mm stroke with a weight of 488 grams.

The S54 uses aluminum cast pistons of 87mm diameter traveling over a 91mm stroke with a weight 485 grams - ouch. 

Hypothesis 004: Do we sacrifice mid range torque over an S85 FBO?

It is commonly accepted that higher duration camshafts tend to maximize high RPM torque gains and sacrifice low to mid range torque. This phenomenon is exacerbated on engines without variable valve timing.

The S85's dual VANOS system provides significant adjustment capabilities within the duration range of the camshafts as documented during the VANOS timing in "Get the dishes in.",

Specifically, the actuator on the intake side of maximum 25.25 mm corresponds to 30° camshaft angle (60° crankshaft angle). The actuator on the exhaust side has a maximum range of 14.25 mm corresponding to 18.5° camshaft angle (37° crankshaft angle).

The VANOS systems use high oil pressure to push the actuators into the gears, effectively turning the camshafts to adjust valve timing within the duration range.

Result 004: Yes. 

There is a torque dip between 3,500 to 4,500 RPM, resulting in a maximum loss of 26wtq and 22whp at 4,500 RPM. In the real world this matters little, I will explain why in the next entries discussing gearing. 

Hypothesis 005: Can we safely reach 125hp/liters?

Notes on crank power. 

In MustangDyne's chassis dynamometer software, engine crank power is a calculated value.

The overall gearing between the engine and the dyno's rolls has been calculated and used to scale the measured torque from a dynamometer roll shaft value to an engine crankshaft value. No correction factor are applied to account for drivetrain losses.

EngineTorque = (DynoShaft Torque * DynoShaft RPM) / EngineRPM

Result 005: Yes!

The calculated number is found on the dashboard below under Eng Power (HP). Per MustangDyne's crank horsepower calculations, this resulted in an engine peak horsepower figure of 627 hp.

We succeeded at hitting the legendary atmospheric threshold of 125 horsepower per liters.

Keep in mind we achieved this using 91 octane pump fuel, on original fuel injectors along with original secondary catalytic converters still installed. All original safety parameters such as knock control and misfire detection were retained. 

There is power left on the table. I will explore potential maximization strategies later on. 

Final dyno charts and tables. 

For clarity's sake, you'll find the final charts below. They all use SAE weather corrected numbers and 91 octane fuel. 

S85-EVO.

S85-EVO tuned v untuned.

S85-EVO v FBO. 

The S85-EVO in perspective. 

I've gathered the data we've compiled to tell various stories that put the S85-EVO in perspective.  

Gains per stages. 

The table below compares the various tuning stages of the S85 and the current state of naturally aspirated performance gains on the original 5.0 displacement. 

Stages	MD (SAE)	DJ (SAE)	Crank Power (hp)	Peak Power (RPM)	Torque (ft lbs)	Peak Torque (RPM)
Stock	400	425	500	7750	384	6100
Stage 1	415	441	519	7750	384	6100
FBO	450	478	563	7900	425	6200
EVO	500	531	625	8425	458	5900

$ per hp.

The F1 heritage is real.