A Brief Study Analyzing the Effects that Throttle Mapping have on the MS3 Engine Output

[Christian.]

A Brief Study Analyzing the Effects that Throttle Mapping have on the MS3 Engine Output

This analysis has been put performed to help us and the Mazdaspeed3 (MS3) community understand the varying effects that the throttle angle, throttle mapping, and throttle control have on the power output of the USDM MS3 engine. We are hoping to be able to prove various points and answer the following questions by analyzing the test results:

Does Cobb Tuning have full control of the throttle system on the MS3?

Does running the throttle system on full duty cycle all the way to redline increase the power output of the MS3 engine?

Is the programmed throttle closing a detriment to the power output of the MS3 engine?

Is the closing of the throttle prior to the engine redline the main reason for the drop-off in engine torque production?

For an introduction, we feel it would be best to review how the DBW (Drive-By-Wire) system operates. This information has been taken from COBB Tuning's EFI 102 class material. We also offer a FREE online EFI 101 class that can be found on this link.



The Throttle By-Wire (TBW) aka Drive-By-Wire (DBW) systems appears to be more complex than they actually are due the the redundancy which is built into the systems for safety reasons. Most TBW systems contain four major components; (1) the Throttle Actuator Control Relay, (2) the Drive Circuit for the throttle actuator which is located inside the ECU, (3) the Accelerator Pedal Position (APP) assembly, and (4) the Electric Throttle Body Assembly. With the implementation of the TBW system, manufacturers were able to eliminate the Idle Air Control (IAC) system because they are able to use the electronic throttle body to control the engine's deceleration and idle. The above diagram will demonstrate the basic TBW system operation, please notice that each sensory component of the TBW system has 6 wires connected to it; we will later explain what each of these wires are for.



The first part of this system is the Throttle Actuator Control Relay which powers and protects the TBW circuity. The second part of this system is the Accelerator Pedal Position (APP) assembly which transmits two pedal position inputs to the ECU. The gas pedal (APP assembly) is effectively two independent pedal position sensors which each contain a +5V reference, a ground reference, and a signal which is sent to the TBW circuitry located inside the ECU. One signal is considered the “Main” APP signal and the secondary signal is considered the “Sub" APP signal, both of which tell the ECU what the driver's foot is doing as far as requesting torque from the motor (how much the driver is pushing down on the gas pedal).

The next part of this system is the Electric Throttle Body Assembly which consists of the throttle actuator motor, throttle valve, and 2 throttle position sensors. Like the APP assembly this part contains 6 wires; two of which are the shared +5V reference and shared ground reference wires which power the two TPS sensors, The next two wires are the redundant TPS sensor signals (Main and Sub) which transmit the position of the throttle blade to the ECU. The last two wires are for the throttle actuator control signal, these lines control the opening and closing of the throttle blade.

The last part of this system is the TBW Circuitry located inside the ECU. This circuitry receives the Main and Sub APP signals so it knows what the driver's foot is doing (as far as requesting torque) then it sends a square wave, 12 volt signal across the two throttle actuator control signal lines to step the throttle actuator motor with a duty cycle; 100% duty cycle would open the throttle blade to WOT & 0% duty cycle would keep the throttle blade at the fully closed position. Based on the feedback received from the two TPS sensor signals located in the Electric Throttle Body Assembly, the TBW system is constantly receiving inputs (APP & TPS) and controlling outputs (throttle actuator motor) in order to control the throttle blade movement on the vehicle.



What this means to you:
...most modern engines will never show a 0% TPS since the throttle blade is always kept open (by 2-7%) to start the engine and to maintain idle.
...some modern engines will never report a 100% TPS while at WOT since their TBW system may have been programmed to only open up to a predetermined point.
...you can change the Rev Limits on an engine by tricking the ECU about the reported RPM but you may not be able to get the motor to accelerate past a certain RPM since the throttle body has been programmed to close by a predetermined RPM.
...porting of most modern throttle bodies is not a good idea unless you have the ability to fully re-calibrate that system.
...most modern engines will not allow the motor to start immediately after the battery has been disconnected then re-connected or if the ECU has been re-set because the systems needs approximately 5 seconds to fully calibrate the TBW system. If the battery has been disconnected then re-connected or if the ECU has been re-set, please allow the key to stay in the ON position for a minimum of 10 seconds to allow the TBW system to calibrate before you try to start the engine.
...providing a constant +12V to the throttle actuator motor will fully open the throttle blade, providing a constant -12V (reversing the polarity) to the throttle actuator motor will fully close the throttle blade.
...most modern EMS logic has been calibrated to control the throttle blade opening and closing speed. We have seen where the throttle blade actually opens slower if you slap down the gas pedal versus gradually depressing the gas pedal.
...most modern EMS programming was established to lessen the emissions output of the motors, so throttle opening and closing delays have been implemented to lessen the emission output of the motors...not to piss you off.
...most modern vehicles have some sort of Traction Control System which use the Electric Throttle Bodies to control engine torque output.
...most modern vehicles have some artificial throttle response programmed into their control. For instance, if the driver is pushing down on the APP 20% the TPS sensor may report a 50% reading. The ECU is creating artificial throttle response by opening the throttle valve much more that the driver is requesting.



The above matrix is an example of an ECU table used to control the electronic throttle body system (this is NOT the MS3 throttle control table). As you can see the Y-axis breakpoints are engine RPM and the X-axis breakpoints are APP sensor signal inputs. The ECU logic is going to look up what duty cycle it wants to step to the electronic throttle body motor based on these two inputs, APP and RPM. It will then use the logic programmed in the on-board PID system to control the mechanical DBW components. Most TBW systems operate in an closed-loop condition where feedback is given to this sub-system and this sub-system will use programmed corrective measures to properly control the throttle blade. Notice how the ECU has been calibrated to shut the throttle duty cycles down to 25% by 7000 RPM, this is done to further protect the motor from an overrev or overrun event. The calibration of this table can dramatically change how a vehicle drives.

Now that we reviewed the DBW system would feel it is best that we demonstrate that we have full control of this system on the Mazdaspeed3. In order to do this we have created four different calibrations that were sequentially run on our Mustang Chassis Dynamometer. These calibrations were programmed as follows:

FTCvA = contains stock throttle mapping
FTCvB = was programmed to cycle between full throttle duty cycle and 25% of full duty cycle (75% less duty cycle) starting at 3000 RPM
FTCvC = forces the throttle system to maintain full throttle duty cycle all the way to redline
FTCvD = maintains a throttle duty cycle that is 50% of full duty cycle (in other words the throttle was 50% closed)

Each chassis dyno graph will be accompanied by a full datalog from the Ford/Mazda factory IDS datalogging system. This is the exact same datalogging/diagnosis systems used by every Mazda dealer in the U.S.A. We have also plotted various dyno graphs on top of each other so that we can more clearly demonstrate the exact power differences measured between these various calibrations.

FTCvA calibrations were tested to allow us to get a baseline of how the stock torque control system constantly manipulates the throttle position in order to control the torque output of the MS3 engine.

IDS Datalog of FTCvA

Many of you are familiar with what the datalogs of what the factory throttle control system does in order to maintain a certain amount of torque. This datalog is from a stock calibration and clearly demonstrates what many of you know. We have mentioned how we try to make our calibrations safer, you will want to notice the lengthy report of Knock Retard being removed from the engine during this dyno run. You can know use this information as a baseline that we can compare the remaining datalogs to.

IDS Datalog of FTCvB

This calibrations was programmed to cycle between full throttle duty cycle and 25% of full duty cycle (75% less duty cycle) starting at 3000 RPM. As you can clearly see the on the top data line, the Accelerator Pedal Position (APP), or gas pedal is clearly depressed and held at full throttle during the entire run. The TP REL, or throttle blade cycles between full duty cycle and 25% of full duty cycle. The dynograph of that dyno run is below and you can clearly see the corresponding effects of this calibration.

Dyno Graph of FTCvB

You can clearly see how the reduction of throttle duty cycle throughout the power band shows up as dips in torque across the entire RPM range. Please feel free to view the short video that we filmed while completing this testing = Link to come as soon as we get it uploaded.

Dyno Graph of FTCvC

This calibration forces the throttle system to maintain full throttle duty cycle all the way to redline. We have composed a short video demonstrating these testing procedures. This video can be seem from this link.

IDS Datalog of FTCvC

This calibration was programmed to hold full duty cycle all the way to engine redline. You can clearly see the TP REL signal mimics the APP signal with full duty cycle and holds the throttle wide open during the entire run. The only time the TP REL signal decreases is when the APP signal decreases.

IDS Datalog of FTCvD

This calibration was programmed to hold 50% of full duty cycle all the way to engine redline. You can clearly see the TP REL signal mimics the APP signal and holds the throttle wide open during the entire run, although the duty cycle run to the DBW throttle body is 50% of full throttle duty cycle. The only time the TP REL signal decreases is when the APP signal decreases. Next is a dyno graph comparing WOT power between the 100% full throttle duty cycles and 50% full throttle duty cycles. The resulta are very interesting.

Dyno Graph Comparing FTCvC (Full TDC) of FTCvD (50% of Full TDC)

In conclusion, we hope this information clearly demonstrates that Cobb Tuning has full control of the throttle mapping for the MS3. We too found it interesting that running the DBW assembly at 50% of full throttle duty cycles made a nominal power difference when compared to a calibration that runs 100% of full throttle duty cycles all the way to redline. This should benefit the community by several means, no we can focus our resources on providing the community with what it wants, and throughout this process we should be able to educate ourselves and the MS3 community about the varying improvement that can be made to these vehicle so that they can be enjoyed for time to come.

The vast majority of MS3 owners are not likely to invest thousands of dollars on top of the price of the vehicle in order to gain additional performance. We want to make sure that we do what we can to make these vehicles more enjoyable. For those that choose to modify their MS3 beyond this point, we have clearly proven that we have full control of the stock MS3 Engine Management System and we will be happy to help you achieve your goals. Please feel free to communicate your desires with us and we will do what can to make these vehicles more enjoyable as long as the dependability of these vehicles is not compromised.

Q: Does Cobb Tuning have full control of the throttle system on the MS3?
A: Yes.

Q: Does running the throttle system on full throttle duty cycle all the way to redline increase the power output of the MS3 engine?
A: Our test results so us that holding full throttle duty cycles all the way to redline does not increase the power output of the USDM MS3 engine. This does not mean that holding the throttle at full duty cycle with vehicles that have drastically different turbos, intake manifolds, heads, camshafts, exhaust manifolds will not be of benefit.

Q: Is the programmed throttle closing a detriment to the power output of the MS3 engine?
A: Our test results demonstrate that the peak power output of the MS3 is not significantly different even when a 50% of full throttle duty cycle is driving the DBW throttle assembly.

Q: Is the closing of the throttle prior to the engine redline the main reason for the drop-off in engine torque production?
A: Our testing has concluded that this behavior is not the main reason why engine torque production drops off significantly as the engine RPM approaches redline. The ability for this engine to breath well in the higher RPM ranges is most likely limited to:
Turbocharger Mass Flow Capabilities
Camshaft Design
Exhaust Manifold Design
Head Mass Flow Capabilities

Take care,
Christian.

[Haltech]

Wow, some good reading right there. Thanks alot Christian!

[LBV]

Very impressive ... sounds like you guys really have done your homework here and aren't rushing anything to market just for the sake of being the first ... that's the way to do it IMO.

[NYpest]

you win post of the year award

[palerider]

Christian posted this at the other forum today before here... and i made a few posts there. I spoke with him at length today and Im starting to feel a certain way about things with our cars.... these are the posts I made at the other forum. Im not gonna get in the habit of of "pasting" here. But I wanted my impressions discussed amongst ourselves here as well. I think the dialogue will result in a better understanding.

First post

Ive got a standback, loads of other guys do as well. Xedes are out there. Weve had some successes but at the end of the day I think we still wanted to know how much we were really accomplishing with our "tunes". Ive always said the standback "nudged" the ecu, Im beginning to think its possible this car doesnt need tuning for our boltons anyway. Ive had every bolton available added slowly over the last 6 months.... then the turbo, manifold and inlet went on..... I proceeded straight to the dyno and.... guess what? The car ran in the 10's a/f..... adding timing didnt really help power because the damn ecu was so adaptive that by adding it slowly it just corrected itself anyway.

geez...in some ways this all frustrates me, but in others it actually is encouraging. I know of two guys right now running big turbo setups and their a/fs are "perfect". One of them is running a gt35 at 20psi on the STOCK ecu.... guess what. Everything looks "great". When we talk about engines that blew we get guys with aggressive procede tunes blowing on the way home, guys with big turbos who slap it on and rip it around the block with no clue as to the wastegate setting thereby not giving the ecu any time to adapt..... and then...

we get how many damn guys who say their engines blow stock or whatever...... with perfect a/fs or just driving around. i have yet to see a single mod fuck up the a/f on this car. thats unbelieveable. the car tunes itself if you give it time. maybe there are design flaws in the engine, coldweather gives it fits, fuel pressure can drop and trigger fuel cuts.

Fuel cuts are another issue..... i talked to Christian today for over an hour and they have never seen a fuel cut on their car. He even advised AGAINST the cdfp upgrade... I think that this is off... but at this point im sitting here wondering what the hell to do next.

If ecus are designed to tune themselves against every parameter that might arise, maybe it was only a matter of time before ecus would become so sophisticated that they would be able to withstand any possible alteration to the system and still correct themselves. I dont think this ecu is there completely, but based on my conversation today, and what i can say my experience has been so far.... Im beginning to believe I should just build my motor and buy new parts... the ecu isnt gonna be the problem. It IS my tuner.


2nd post

Ive got all the damn mods in my sig, INCLUDING a atp inlet, stuck into the cobb sri silicone used a reducer.... to a mscai. the mscai being the king of lean.

and you know what?
today, after rotating my tires and getting a new inspection. I drove the hell out of it on the way home with a/fs in the 10's on WOT.
You know what else?
I got fucking fuel "pulled" with the standback. How about them apples? I may run 11's in the cold but this SOB stays in the 10's no matter what I throw at it.

My theory?

A cdfp is the most important mod on this car. if you get it it will solve cold weather fuel cut for the most part..... and it will supply all the fuel this ecu would ever need to keep close to its commanded a/fs with all the mods. I dont understand any other difference.

Whoosh on the other forums I have spoken to and he is running 11.5-12.0 a/fs with turboback exhaust, 3 inch maf, homemade inlet, ported intake manifold, tubular exhaust manifold, fmic, you name it...all with a GT35 at 20 psi.... all on the stock ecu!!!!!!! But what does he also have?..... a mrlilguy cdfp.

Our last conversation was when i sent him, my Cobb maf housing, to replace his 3inch maf in order to get the a/fs down from 11.5....lol This is crazy but its true.

[Laloosh]

Help me understand this. It is my understand this is one of the most adaptive if not the most adaptive ecu you guys have ever seen. Something like 37 different parameters trigger fuel cut in this car. We have seen guys go full bolt on running gt30/35s and have a/f ratios in the 11s. Meaning the ecu is controlling this somehow right?

My question is...take a simple map induced fuel cut. If the map senses 17-18 psi, boom fuel cut. Solution, put a map clamp on and trick the map....fuel cut gone. If this ecu is so smart that 37 different parameters are intertwinned, a simple 15 dollar map clamp would not work cause something else would pick this up, and it is clearly not.

Example 2; This throttle. You guys said it yourself, an engine is a pump, you pump air in you pump air out. We all agree on this. How do u get more air? Obviously force it in or make the current pump more efficient right? I know im boring you with this basic info but stick with me. The turbo the manifold the cams the intercooler the throttle...are all forms of restictions. Upgrade all those and what is the only thing left? The throttle. Based on your maf volt readings you guys are reading high 3s low 4s....thats alot less than the 4.5-6 that bolt on car are reporting. Maybe the throttle is not doing squat on a intake car, but are you seeing the same exact thing on a fully bolted ms3?

From what many of us have gathered, this ecu does a damn good job at adapting itself to run a/f and timing no matter what, this is proven by BT guys. My question here, and dont take this the wrong way, are you not changing what the ecu commands because you simple don't know where to start or because if you release your ap with parameters that make it hold 17psi, make it hold thorttle, make it traget a lower a/f ratio, you simple wont make shit on your cobb protunner/race version of the ap. Because afterall, the stock tune works pretty much every god damn thing we throw on this car, why wouldnt a slightly altered ap version of that do the same?

[palerider]

good questions chris!!! im going to bed and I'll read christians responses in the morning.

im an addict right now....lol

[Ziggo]

I just ran some rough numbers on the stock fuel pump, and by the numbers it supplies almost twice the fuel required to run an afr of 12 with a volumetric efficiency of .95 and 18lbs of boost. IAT of 140ºF. There can be losses elsewhere, but I don't think they would result in a reduction of half.

Makes me think if you have control of the solenoid you can supply plenty of fuel.
My opinion (and it is just that, an opinion) having a higher flowing pump raises the fuel pressure much quicker than the stock once the solenoid gets the signal and in effect is tricking the stock ECU into keeping the fuel pressure nice and high. If you could control the solenoid directly and have it open fully for less throttle position you would get the same result.

I guess my question is, is this what you have done? Are you controlling the fuel pump solenoid in any way?