These are my results and findings from 2015, when I was trying to find the safe and conservative limits of a bone stock setup, save for an HPFP upgrade. I had just gotten a new engine, and had been tuning for months after the break-in period. I wanted to see exactly how much the platform can do without any hardware power mods (the HPFP is a reliability mod in my mind). Here goes:
First off some quick notes: I have been tuning a stock MS3 platform for about 6 months now, as I believe that fairly significant gains can be had without any modifications beyond the safety net of an HPFP upgrade. I keep seeing folks throwing parts on their car before even buying a COBB AP or any tuning platform, and it just seems like a waste of money and a non-scientific approach to making power.
So, here is the premise: We have a ‘12 genPu, bone stock, 190 psi compression across all 4 cylinders, .028 gap on stock heat range NGK plugs, 20K miles on the engine (stock engine from Mazda), no aftermarket parts on it save for the HPFP, 93 Octane (possibly winter blend already). Having previous non-DISI tuning experience and having tuned this platform with various mods (3”intake, full catless and full catted DPs, and an FMIC) this particular tune took 22 very incremental maps to get to where I believe it is nearly final. The only issues I can see are a) very minor (< 1) transition KR around 2,500 rpm, and b) the 1.31KR under WOT described below.
The following dyno runs were made, with an approx. 60 deg F ambient temp on a DynoJet. I have AP logs of dynos 2) and 3) but not 1), as the COBB AP was not married to the car during the first run:
1) (Blue) - Bone stock PU, HPFP upgrade, no COBB AP installed. Rounded 217 whp / 239 wtq.
Video:
Ignore the weird plateau table thing in the AFR section, the dyno software had some glitch. The real AFR line for this is the bottom blue line.
2) (Red) - Bone stock PU, HPFP upgrade, my tune, version B.1.7.5 (I have a method for labeling each consecutive map). Rounded 257 whp / 294 wtq.
Video:
At the end of the pull you can see the black smoke caused by a fuel-cut created KR which leads to a fuel dump by the ECU, a phenomena discussed here: Gen 2 Cobb ATR discussion
Main tune points: Target AFR: 11.3, Target boost: 17psi tapering off to 15 (more like 14 on the dyno), and fairly conservative timing (12.5 deg max logged). Log is attached.
3) (Green) – Bone stock PU, HPFP upgrade, COBB OTS Stage 0. This was dynoed to show my long-standing suspicion that the stock Mazda map is NOT exactly what COBB describes as their Stage 0, and the difference is obvious – 16whp / 10wtq increase in Stage 0 vs. the stock Mazda tune. I don’t know what is different about the Stage 0 vs. the OEM Mazda tune, but aside from very slightly colder intake temps, there were no environmental differences. A few degrees should not make that huge of a difference, hence it’s COBBs tune. Rounded 232 whp / 249 wtq. Video: phone wasn’t cooperating/rebooted. Identical to run 1).
Dyno result notes: I believe run 1) was slightly under performing because the car sat on the dyno for a bit, heat soaking while we were getting ready. Since I don’t believe the COBB Stage 0 tune to be THAT far off from Mazdas’ stock tune, the real stock numbers are probably half way between the two, so around 224 whp / 244 wtq, which puts the car around a fairly average 13% drivetrain loss relative to the official stock 260hp/280tq crank figures. So the real tune gains are ~33whp / 50wtq, if we’re being conservative, or 40 whp / 55wtq if we go right by the difference between 1) and 2). Regardless, I have yet to see anyone post a bone stock tune with dyno results anywhere near those gains. Typically these are achieved with an intake/TIP or both, a DP and E blend. My philosophy is that this doesn’t need to be the case. I dislike the extra noise when it is not directly related to performance (and even then it’s a trade off for a daily). Having all stock parts and making this kind of power on 93 is how I believe this car should have been sold by Mazda. I’m sure there are some reliability concerns, but none that overshadow those present with a modded (with parts) platform. As an aside: Virtual Dyno representation of the logs for pulls 2) and 3) is showing power numbers about 40% higher than noted. Sort of useless.
Tune notes: I tried to be as conservative as I could, most of the timing is 1-1.5 degrees under any KR values of .5 or above that I’ve encountered during my logs. I’ve been very incremental about increasing timing, so I never really hit KR over 1 (except for the 1.31 on the dyno). The 1.31KR seen in the log is responsible for the power dip around 5k rpm. This wasn’t present during my street logs, as the loads hit on the dyno were lower than on the street and those cells will need to have 1-1.5 deg of timing taken out, just to be safe. I only got to do 3 pulls, so I didn’t have the opportunity to log another run with my tune to see if it was a fluke.
Overall notes: I hope this map and dyno help the community, as I’ve gotten help from the multiple guides (and put one together myself in the FAQ), threads, and other people’s expertise to get to this point. My map is free for anyone to copy as they wish, I just ask that you don’t do it for commercial purposes – I’m posting it up here so that folks can have a much better base for their tunes than what COBB provides us with, and potentially decide that this is more than plenty for a reliable stock (w/HPFP) daily. Some slight alterations may be necessary as no one car is exactly the same, but this gets you 95% there. I did learn a crap load while working on this tune, so keep that in mind before you simply copy/paste and never touch it again. One of my biggest goals with the maps was to create a very smooth map, which I’ve succeeded at if you look at the power curve compared to the stock and Stage 0 maps. I’ve applied this to most tables, making them as smooth and non-peaky as possible, while maintaining values needed to run the mpg/power figures I was looking for.
I may stop here (after fixing the transition/tip-in KR and the 1.31 KR spot), and perhaps wait ‘til the stock turbo craps out and throw the hybrid K04 on there down the road, which would warrant a re-tune. The car drives well, slightly touchy after you pass the 1.0 load threshold, but that is a preference that can be tuned out, if so inclined.
The dyno sheet, my logs and my map are attached. Comments and questions are welcome.
I’d like to thank everyone who has been putting up with my bugging them about random stuff over the past few years and their posts/threads which I’ve perused over and over, especially @Enki; @ms3blackmica; @phate; @rfinkle2; @Lex; @Rokusek; @Captain KR; @silvapain; @timjs; @JeffroMS3; @dereeek; @mikesms3; whoever banned/unbanned me from shout at any given point, and the entire NATOR community for being awesome.
Bonus: Corner weighed stock genPu MS3, full tank of gas, TRMotorsports C3 17x9 wheels, Hankook V12 255/40-17 tires, and a small 8” ZoomBox-cased sub in the driver side trunk cavity + small 1ch amp in place of stock spare-mounted sub (both weigh 25lbs at best):
Without driver (3311 lbs): https://lh3.googleusercontent.com/-...z_x-rqhA/s1440-Ic42/IMG_20151024_175351-1.jpg
With 194lbs fat ass driver (3505 lbs): https://lh3.googleusercontent.com/-...995QtkaKRI/s1280-Ic42/IMG_20151024_175836.jpg
DISCLAIMER for map downloads and usage: Downloading and using these maps are AT YOUR OWN RISK. Neither MSO nor I will be held accountable if anything were to happen to your car after using/mimicking/copying/installing one of these maps.
Update: Uploaded map vB.1.7.9 - Minor tweaks to Closed Loop, Fuel and ignition tables, made sure CL is entered early and the newly available Closed Loop Max Load Limit tables are the same as the Closed Loop Max Load tables. Finally fixed the overly touchy throttle between load 1 and 1.5, now the acceleration is much smoother and linear. SWAS disabled in 1st/2nd gears, using the new COBB update.
This is the final stock map. All future maps will be supporting mods. Having driven on it, transition, part throttle, and WOT KR are all 0.5 or less, which I believe is the true mark of a proper tune.
Update 2: Uploaded my old ATR version that I used to create the maps. It will only be useful to open the uploaded maps so you can copy the tables over to your own ATR maps for your car.
UPDATE 2021 09 25: Uploaded a zip of the ATR that opens my maps. Just unzip to your E:\Program Files (x86)\Accessport\ folder so that its in "E:\Program Files (x86)\Accessport\Accesstuner - FORD USDM 2011-2013 MAZDASPEED3\" and run ATR from there. It should open the maps. Let me know if you run into any issues.
First off some quick notes: I have been tuning a stock MS3 platform for about 6 months now, as I believe that fairly significant gains can be had without any modifications beyond the safety net of an HPFP upgrade. I keep seeing folks throwing parts on their car before even buying a COBB AP or any tuning platform, and it just seems like a waste of money and a non-scientific approach to making power.
So, here is the premise: We have a ‘12 genPu, bone stock, 190 psi compression across all 4 cylinders, .028 gap on stock heat range NGK plugs, 20K miles on the engine (stock engine from Mazda), no aftermarket parts on it save for the HPFP, 93 Octane (possibly winter blend already). Having previous non-DISI tuning experience and having tuned this platform with various mods (3”intake, full catless and full catted DPs, and an FMIC) this particular tune took 22 very incremental maps to get to where I believe it is nearly final. The only issues I can see are a) very minor (< 1) transition KR around 2,500 rpm, and b) the 1.31KR under WOT described below.
The following dyno runs were made, with an approx. 60 deg F ambient temp on a DynoJet. I have AP logs of dynos 2) and 3) but not 1), as the COBB AP was not married to the car during the first run:
1) (Blue) - Bone stock PU, HPFP upgrade, no COBB AP installed. Rounded 217 whp / 239 wtq.
Video:
Ignore the weird plateau table thing in the AFR section, the dyno software had some glitch. The real AFR line for this is the bottom blue line.
2) (Red) - Bone stock PU, HPFP upgrade, my tune, version B.1.7.5 (I have a method for labeling each consecutive map). Rounded 257 whp / 294 wtq.
Video:
At the end of the pull you can see the black smoke caused by a fuel-cut created KR which leads to a fuel dump by the ECU, a phenomena discussed here: Gen 2 Cobb ATR discussion
Main tune points: Target AFR: 11.3, Target boost: 17psi tapering off to 15 (more like 14 on the dyno), and fairly conservative timing (12.5 deg max logged). Log is attached.
3) (Green) – Bone stock PU, HPFP upgrade, COBB OTS Stage 0. This was dynoed to show my long-standing suspicion that the stock Mazda map is NOT exactly what COBB describes as their Stage 0, and the difference is obvious – 16whp / 10wtq increase in Stage 0 vs. the stock Mazda tune. I don’t know what is different about the Stage 0 vs. the OEM Mazda tune, but aside from very slightly colder intake temps, there were no environmental differences. A few degrees should not make that huge of a difference, hence it’s COBBs tune. Rounded 232 whp / 249 wtq. Video: phone wasn’t cooperating/rebooted. Identical to run 1).
Dyno result notes: I believe run 1) was slightly under performing because the car sat on the dyno for a bit, heat soaking while we were getting ready. Since I don’t believe the COBB Stage 0 tune to be THAT far off from Mazdas’ stock tune, the real stock numbers are probably half way between the two, so around 224 whp / 244 wtq, which puts the car around a fairly average 13% drivetrain loss relative to the official stock 260hp/280tq crank figures. So the real tune gains are ~33whp / 50wtq, if we’re being conservative, or 40 whp / 55wtq if we go right by the difference between 1) and 2). Regardless, I have yet to see anyone post a bone stock tune with dyno results anywhere near those gains. Typically these are achieved with an intake/TIP or both, a DP and E blend. My philosophy is that this doesn’t need to be the case. I dislike the extra noise when it is not directly related to performance (and even then it’s a trade off for a daily). Having all stock parts and making this kind of power on 93 is how I believe this car should have been sold by Mazda. I’m sure there are some reliability concerns, but none that overshadow those present with a modded (with parts) platform. As an aside: Virtual Dyno representation of the logs for pulls 2) and 3) is showing power numbers about 40% higher than noted. Sort of useless.
Tune notes: I tried to be as conservative as I could, most of the timing is 1-1.5 degrees under any KR values of .5 or above that I’ve encountered during my logs. I’ve been very incremental about increasing timing, so I never really hit KR over 1 (except for the 1.31 on the dyno). The 1.31KR seen in the log is responsible for the power dip around 5k rpm. This wasn’t present during my street logs, as the loads hit on the dyno were lower than on the street and those cells will need to have 1-1.5 deg of timing taken out, just to be safe. I only got to do 3 pulls, so I didn’t have the opportunity to log another run with my tune to see if it was a fluke.
Overall notes: I hope this map and dyno help the community, as I’ve gotten help from the multiple guides (and put one together myself in the FAQ), threads, and other people’s expertise to get to this point. My map is free for anyone to copy as they wish, I just ask that you don’t do it for commercial purposes – I’m posting it up here so that folks can have a much better base for their tunes than what COBB provides us with, and potentially decide that this is more than plenty for a reliable stock (w/HPFP) daily. Some slight alterations may be necessary as no one car is exactly the same, but this gets you 95% there. I did learn a crap load while working on this tune, so keep that in mind before you simply copy/paste and never touch it again. One of my biggest goals with the maps was to create a very smooth map, which I’ve succeeded at if you look at the power curve compared to the stock and Stage 0 maps. I’ve applied this to most tables, making them as smooth and non-peaky as possible, while maintaining values needed to run the mpg/power figures I was looking for.
I may stop here (after fixing the transition/tip-in KR and the 1.31 KR spot), and perhaps wait ‘til the stock turbo craps out and throw the hybrid K04 on there down the road, which would warrant a re-tune. The car drives well, slightly touchy after you pass the 1.0 load threshold, but that is a preference that can be tuned out, if so inclined.
The dyno sheet, my logs and my map are attached. Comments and questions are welcome.
I’d like to thank everyone who has been putting up with my bugging them about random stuff over the past few years and their posts/threads which I’ve perused over and over, especially @Enki; @ms3blackmica; @phate; @rfinkle2; @Lex; @Rokusek; @Captain KR; @silvapain; @timjs; @JeffroMS3; @dereeek; @mikesms3; whoever banned/unbanned me from shout at any given point, and the entire NATOR community for being awesome.
Bonus: Corner weighed stock genPu MS3, full tank of gas, TRMotorsports C3 17x9 wheels, Hankook V12 255/40-17 tires, and a small 8” ZoomBox-cased sub in the driver side trunk cavity + small 1ch amp in place of stock spare-mounted sub (both weigh 25lbs at best):
Without driver (3311 lbs): https://lh3.googleusercontent.com/-...z_x-rqhA/s1440-Ic42/IMG_20151024_175351-1.jpg
With 194lbs fat ass driver (3505 lbs): https://lh3.googleusercontent.com/-...995QtkaKRI/s1280-Ic42/IMG_20151024_175836.jpg
DISCLAIMER for map downloads and usage: Downloading and using these maps are AT YOUR OWN RISK. Neither MSO nor I will be held accountable if anything were to happen to your car after using/mimicking/copying/installing one of these maps.
Update: Uploaded map vB.1.7.9 - Minor tweaks to Closed Loop, Fuel and ignition tables, made sure CL is entered early and the newly available Closed Loop Max Load Limit tables are the same as the Closed Loop Max Load tables. Finally fixed the overly touchy throttle between load 1 and 1.5, now the acceleration is much smoother and linear. SWAS disabled in 1st/2nd gears, using the new COBB update.
This is the final stock map. All future maps will be supporting mods. Having driven on it, transition, part throttle, and WOT KR are all 0.5 or less, which I believe is the true mark of a proper tune.
Update 2: Uploaded my old ATR version that I used to create the maps. It will only be useful to open the uploaded maps so you can copy the tables over to your own ATR maps for your car.
UPDATE 2021 09 25: Uploaded a zip of the ATR that opens my maps. Just unzip to your E:\Program Files (x86)\Accessport\ folder so that its in "E:\Program Files (x86)\Accessport\Accesstuner - FORD USDM 2011-2013 MAZDASPEED3\" and run ATR from there. It should open the maps. Let me know if you run into any issues.
Last edited:
