recent interesting post from other side of pond
"
As many of you may have followed in the last few weeks, Thistle (John) has made some great headway in getting a better understanding of the R35 OEM Boost Control system. Tim Bailey and I had a chance to explore and test this new system on one of the Cobb GTRs this week.
I thought it would be helpful to put together a bit of technical data about that particular tune, as well as a more detailed look at the replacement boost control system. It should be noted that this first pass at replacing the OEM system is an alpha test, and in the coming weeks a more significant Beta test will occur so some refinements can be made. As John has mentioned, there are a lot of variables when changing the factory boost control so you maintain the full range of factory safety systems.
We did this test tune on a ‘Stage 2’ setup, which in this case was an R35 with a custom rear cat section, mid pipe, and 5Zigen exhaust. Before looking at the boost control in more detail, let’s start with a look at the Stage 2 tune itself.
Here you can see the resulting wheel power and torque as measured on our Mustang dyno, as well as the boost profile during the pull. This is using the new boost control algorithm, and it works very well. Boost builds quickly and is very smooth. During the rapid onset the damping is nearly perfect to allow fast build without overshoot. This is of course partly due to the well matched wastegate actuator spring weight. Having a wastegate actuator that is closely matched to your target boost is very helpful to keep boost control stable. This is a 4th gear pull with a resulting peak power of 513 whp and 525 lb-ft of torque. Boost peaked at 17.46psi, with a taper down to 13.4psi at redline.
Knowing this is getting close to the end of what the turbochargers and flow, it is worthwhile to look at where we are with the injector capacity. Most OEM manufactures size the injectors and turbocharger to be well matched, and this is certainly the case here. Take a look:
![]()
You can see from the above graph that the injector duty cycle peaks at 97.7%. In multiple run tests, duty cycle would occasionally hit 100%, and certainly anything over 85% is getting very close to the edge of the injector flow. You would obviously need larger injectors for any further modifications that increase airflow, especially changing the turbochargers. I also graphed measured wheel HP on this same chart as it illustrates that correlations between total airflow/fuel flow and resulting power. With a bit more math you can take the measured AFRs and correct for a single AFR to determine a basic BSFC (Brake Specific Fuel Consumption) None the less, unless something is done to fundamentally increase the engine BSFC, peak wheel HP will be limited by these injectors to something close to this.
Looking a bit deeper at this tune, we can inspect the timing and AFRs:
As is typical in a turbocharged car, timing dips at lower RPM and ramps rapidly towards redline. Given the boost falloff this is more pronounced as the reduced in cylinder pressure will decrease the combustion speed just a bit. AFRs in a good tune are pretty straightforward, and here AFRs are right around 11.3-11.5 in the majority of the run. Leaning out the entire mixture by .3-.5 AFRS didn’t make a significant difference in power, so this AFR works well to keep EGTs and knock under control.
As with the graph comparing injector duty cycle with HP, the ECUs representation of load is typically similar in shape to the torque curve (as one would expect as greater load is the result of more air and fuel in a single rotation, which results in greater torque output).
In the case of this ECU, the load is represented by a theoretical injector pulse width that would be needed to maintain a stoichiometric mixture given the air volume in the cylinder. This number is a scaled value of what most tuners might see as g/sec/cylinder. One can derive a similar number by taking the MAF g/s and dividing by the engine RPM. It is interesting that the load stays relatively flat in the midrange as the torque is raising then falling. I’m not sure if this is a result of some type on internal VE correction, or perhaps a real and significant change in engine torque produced with the same airflow. Some more investigation is in order.
![]()
Last but not least, here is the wastegate duty cycle as well as the boost. As you would expect the duty cycle is at 100% during the initial part of the spool, the quickly drops after boost onset slowly ramping to 100% at 6300rpm. With 100% duty cycle above 6300 rpm, it isn’t possible to get more boost without a change in the wastegate actuator, or by running a larger duty cycle down low and using some of the turbine inertial to carry a bit more boost up top. Either way, this is a pretty optimized map given the size of the stock turbos.
![]()
Here you can see the resulting power to the wheels in each gear. Having tuned many cars over the years, this is pretty much the most impressive multigear pull I have seen. Real power peaks in 4th gear, likely a result of a combination lower inertial loss as gears go up combined with less parasitic loss compared to 5th gear. Peak boost is slightly higher in 5th gear (16.9) compared to 4th, but less then what it was in the single 4th gear pull. This is due to the dyno having a bit less load in the multigear test. It is amazing how little the power drops between gear shifts. The power to the wheels is at approximately 500whp from 3rd gear all the way thru 5th. That is significantly more then most 500whp cars can muster.
![]()
This is an interesting overlay of boost during the multigear run, but overlayed based on in gear RPM. This allows a bit of a better comparison of gear based boost. In general, things are very close, and surprisingly close given the last of a gear comp system. Obviously the slightly higher 5th gear boost is a result of the greater engine load, and it also highlights the need to do a multi gear dyno run using a loading dyno ( as opposed to a non loading Dynojet).
Wastegate duty can be compared in a similar manner, and as expected the results are almost the same. Higher gears us a bit less duty cycle to produce a bit more boost. This effect might eventually lead to a minor addition to this boost control system (implement a gear based offset or comp). Obviously that isn’t something that is critical yet, as the gear to gear variance is minimal. More testing on that front to come.
Looking at the Injector Duty cycle and load in a per gear view during the run, we can see the load is similar in each gear (5th gear was not all the way to redline). Injector duty cycle is a pretty flat, as you would expect with flat hp.
Timing and AFRs during the run were right on from the map. You can see the AFRs are slightly leaner in 5th gear when compared to 4th, which is probably a combination of the slower rate of change in rpm as well as load not being the same.
Last, but certainly not least, here is the new boost map added to the ECU. This map provides the wastegate duty cycle based on current rpm and boost. You can see that the WGD dips in the midrange, as the turbocharges do not need as much duty cycle to maintain the target boost. As RPMs grow, duty cycle ramps up. Above 17psi you can see the falloff of duty cycle. This falloff provides the fundamental boost control.
Here is a graph of the duty cycle the car ran during the 4th gear pull plotted as height on the same scale.
I’ll post up some more details about boost control compensation ideas, plus some more results from the track testing. This GTR running this map ran at the Top Gear track day.
Special thanks to John (thistle), and of course Tim Bailey, Joe, and the rest of the crew at Cobb SLC. Exciting times ahead in the GTR world!
Jeff Sponaugle
Research Calibrator
Cobb Surgeline Tuning
