Category Archives: Tuning

Still tuning the Turbo Concepts

I’m continuing to work on replicating the boost profile that the car operated at with the BorgWarner K04’s.  In this effort I am mainly concerned with the top end boost so that I can compare the pre-turbine backpressure produced by the Turbo Concepts Stage 1 turbochargers with that from the other logged turbo’s.  The current state of the boost profile is shown below:

Turbo Concepts Stage 1 Turbos Boost
Turbo Concepts Stage 1 Turbos Boost

This tune started from a FrankenTurbo F21 tune that I used with the same hardware, aside from the turbochargers.  The wastegates on the Turbo Concepts turbochargers are set about 1.5 psi less than what the FT’s had, so I was anticipating some adjustments needing to be made.  The first few revisions were not generating the results I was wanting, so I went back and logged the boost developed at a variety of fixed wastegate duty cycles.  A sample of some of these are shown below:

Turbo Concepts Fixed Wastegate Duty Cycle Boost Levels - 50, 60, 70
Turbo Concepts Fixed Wastegate Duty Cycle Boost Levels – 50, 60, 70.

I suspect one of the reasons this effort to modify the tune is proving a bit more of a challenge than I had anticipated is on account of the unsteady boost curve.  I’d have expected a fixed WGDC to produce a straighter line.   I was concerned with the wastegates when I was trying to set the pre-load and was having trouble, the variations shown above further my concern about the wastegates on these turbochargers.

WG Refinement

Matching 8 psi preload wastegates with the TTE550 sparked some discussion over the potential need to adjust the tune to accommodate this component pairing.

I’ve been satisfied with the responsiveness from the TTE550, but in the interest of finding out if “doing it the right way” would make any difference in the results I decided to undertake an effort to Calibrate the Tune’s KFLDRL table.  A detailed report on one persons attempt to do this is documented on the Nefmoto site and I’m using that as a starting point for my effort.

After setting CWMDAPP and KFLDRAPP appropriately I began by collecting 0% and 10% data.

 

0% N75DC
0% N75DC
10% N75DC
10% N75DC

 

An early observation is that I can likely refine where I am collecting data without adversely affecting the outcome.  Attempting to log boost data from 1000 to 6500 rpm in every gear is time consuming, risky, and not possible due to the speeds reached in 6th gear at 6500 rpm.  I’m not interested in conducting this work on a dyno so I am limited by conditions on public streets.

With the intent being to capture the maximum boost pressure achieved at a particular engine speed for each duty cycle setting, it looks as though I am able to use 6th gear to record from 1000-2500 rpm, 4th gear from 2750-4000, and 3rd gear from 4500-6500.  I believe this will produce ‘good enough’ results and reduce the amount of measurements I need to make.

Some additional measurements, 20% and 30%.  These were made at ambient temperatures about 20 degF greater than the other two, which raises the question, should the fixed wastegate duty cycle boost pressure be recorded at the same temperature?  I would think so, but then if it did matter what does that say for the KFLDRL results that are being tabulated and how they hold across different temperature ranges.

20% N75 DC
20% N75 DC
30% N75 DC
30% N75 DC

These latest results are prompting me to also drop taking measurements in 4th gear since I’m not seeing any value added from those readings.

I will be logging 3rd and 6th gears with only the <3000 rpm region of the 6th gear data being used.

Shown below are the composite 3rd gear results thus far.  Note: The 10% DC pull was very short.

0-30% Fixed WG Duty Cycle Boost Readings
0-30% Fixed WG Duty Cycle Boost Readings

End of the K03 Journey

Today I dyno’d my S4 with K03 turbochargers and a couple of different engine tunes capping a long process of collecting baseline performance data for the car on these turbochargers.

Along the way I received some significant support from Nye and Daz at the Nefmoto forum and Doug at FrankenTurbo.

The charts below show the final results on the dyno but are only a portion of the data collected on ET Tuning’s Dynojet dyno.  More results will be posted as I plot the data showing various comparisons.

Best numbers from the two stock tune pulls: 253 wtq / 237 whp

BorgWarner K03's With the stock B5 S4 tune
BorgWarner K03’s With the stock B5 S4 tune

 

Best numbers from the two Nef tune pulls: 362 wtq / 283 whp

BorgWarner K03's on the Nefmoto B5 S4 community tune
BorgWarner K03’s on the Nefmoto B5 S4 community tune

 

Stock pulls vs Nef tune pulls

Stock tune vs Nef community tune on K03's
Stock tune vs Nef community tune on K03’s

The ‘proof I was there’ shots.

Post activity on the dyno
Post activity on the dyno

k03_dyno_side

I also did a set of pulls using the NVR Stage 2++ tune that used to be available on the Netmoto site.  This tune came with a warning to ensure the K03’s were in good shape as it runs a lot of boost.  This is apparent through the mid-range where the torque remains strong compared to the Nef tune that immediately begins ramping down the boost pressure following the initial spike.

Best numbers from the two NVR tune pulls: 375 wtq / 303 whp

Chart of NVR Stg2++ vs Nef Community Tune
NVR Stg2++ vs Nef Community Tune

The DynoJet software has a function for reporting the area under the curve (AUC) for the torque and horsepower.  As a way to illustrate the performance difference from the stock to Nef to NVR tunes as a percentage value, rather than horsepower or ft-lbs, which can be contentious since dyno’s can read differently, I’ve tabulated the various AUC values and show the percentage increase as the tune is made more aggressive.

DynoJet reported Area Under the Curve Comparison
DynoJet reported Area Under the Curve Comparison

Now for the charts…

Boost comparison of all three tunes:

Manifold Air Pressure
Manifold Air Pressure

 

Mass airflow:

k03_dyno_maf

Here’s an approximation of where these three tunes land on the K03 compressor map.

k03_dyno_comp_map

Fuel Injector Duty Cycle (you want to play, you got to pay):

IDC
IDC

Timing:  This is a single sample of Timing from each tune.  If all six timing curve lines were shown  it would be hard to read due to the overlap in timing values.

Chart of K03 Dyno Timing Curves
K03 Dyno Timing Curves

The chart below illustrates the temperature rise that occurs in the turbocharger with the different boost levels that each tune operates at.  Temperature out of the turbocharger is recorded approximately 18 inches downstream of the outlet coupler, which allows for some temperature drop before the reading is taken.  This means that the temperature exiting the turbocharger is greater than shown below. The tunes were recorded in the following order: Nef, Stock, NVR.

Air temperature into vs out of the turbocharger
Air temperature into vs out of the turbocharger

The next chart shows the temperature of the exhaust gasses going into the turbine as recorded by the RS6 exhaust gas temperature sensors (upper lines).  The temperature of the gasses leaving the turbocharger turbine housing are recorded downstream at the bottom of the downpipe (lower lines), prior to the catalytic converter, by a K-type thermocouple.  It’s expected that the exhaust gasses will have cooled some by the time they reach the thermocouple.  The downpipes are ceramic coated with the Swaintech white lightning coating and then wrapped in DEI exhaust thermal wrap which may help reduce the heat loss through the downpipe.

Pre & Post Turbine Exhaust Gas Temperatures
Pre & Post Turbine Exhaust Gas Temperatures

Interestingly the significantly different manifold pressures developed by each tune do not appear to have an affect on the exhaust gas temperatures.

This next chart shows the intake air temperatures as recorded by the vehicles temperature sensor inside the intake manifold.  Even though these pulls were done on K03’s and with a little help from some water-methanol injection, I am still impressed by the performance of the SRM intercoolers.

Intake Air Temperatures
Intake Air Temperatures

I decided to supplement the intercoolers with WMI as a result of prior experiences on dyno’s having absurdly high intake temperatures.  None of the tunes were adapted for WMI so I wanted just enough liquid to help keep temperatures down over several pulls.  I went with a pair of Aquamist nozzles, one sized 0.5 mm and the other 0.6 mm.  Together with the TorqByte WMI controller I had the setup configured to flow no more than 300 mL/min.  By comparison, on a pair of 1 mm nozzles the system can flow up to 1000 mL/min.

One of the hopes I have with using the dyno is to be able to come back later and produce similar results.  To see how realistic this may be I dug up a set of DynoJet logs that I made with my car in 2005 just prior to having APR’s Stage 3 kit installed.  At the time the car was equipped with stock intercoolers, AWE-Tuning’s downpipes, and a Supersprint exhaust.

Fast forward to 2015 and a mostly stock version of my S4 is going back onto a DynoJet, this time with SRM intercoolers, Autospeed downpipes, and 034 Motorsport’s 3.5″ single exhaust, along with a WMI system.

Here’s how the results from two different dyno’s separated by about 900 miles and ten years panned out.   2005 it was about 70 degrees at the dyno, in 2015 it was around 60 degrees.

2005_vs_2015_djdyno

To help better break out the results I exported the data and reformatted it.

2005_vs_2015_dyno_readings

Lastly, the intake air temperature differences:

2005_vs_2015_iat

The DynoJet dyno results look fairly consistent.

Another question that can be asked is how accurate the DynoJet figures are at representing the true torque and horsepower produced by the car.

As a cross check I like to look at road calculated values and compare them with what the dyno is reporting.  The curves below are for each of the tunes on the street.

Road Calculated Horsepower and Torque Curves
Road Calculated Curves

Here are the peak values tabulated along with the peak MAF reading, which I have found correlates pretty well with the wheel horsepower figures when using a stock MAF housing.

Chart of Road Calculated Maximum Values
Road Calculated Maximum Values

 

Here are each of the tunes broken out with a Road versus DynoJet chart:

Stock

Chart of Stock Road vs DynoJet
Stock Road vs DynoJet

Nef

Chart of Nef Road vs DynoJet
Nef Road vs DynoJet

NVR

Chart of NVR Road vs DynoJet
NVR Road vs DynoJet

I believe the DynoJet reads higher than what the true values are, but the repeat-ability of the device makes it a good tool for what I am using it for.