Mazda3 Forums banner

Self-tuning adventure!

21K views 39 replies 4 participants last post by  222speed3 
#1 · (Edited)
Goal:
Self tune the car geared towards my driving style.
So with all the power mods done, it's time to tune the car appropriately.

The biggest problem is having an adequate place to take my logs., mainly the WOT logs as safety and police are a concern. Anyhow this is how I'll break down the tunning steps. (I'm sure these can come in handy for people with turbo kits once I actually get results and more info posted).

1) MAF cal. Most important first step in order to get stable AFR values.
2) Adjust the boost to my liking
- Subsequently WGDC might have to be adjusted based on the changes and other values.
3) Once boost and the AFRs are dialed in, we can go ahead and look at timing advance to try and squeeze out more power.

* This will be a hybrid tune, meaning load will also have to be taken into account and adjusted (extra precaution due to the fact that temperature varies so much where I live. Ex: less boost is required in colder temps to produce the same load). I'll have to look a little more into this in order to determine when to start adjusting these, or if it's even necessary to adjust the Cobb OTS values.

4) Other parameters (throttle and so on if needed)




*** DISCLAIMER ***


I am NOT responsible for anything that happens to you or your car if you use info from this thread. Do your research and modify at your own risks. Happy reading :)

 
See less See more
#2 · (Edited)
Useful links

I'll reserve this post where I'll be posting all the useful links I have used to help me tune.

(*** If some links don't work, REMOVE the "s" from the beginning of the URL: https --> http ***)

MAF cal

Using Excel To Dial In Your MAF Curve


Dialing in Boost

Ignition Timing and General Timing

Smoothing Timing Tables

Calculating WGDC

More WGDC

KR Explained

Understanding Load Tuning (post ~26)
More Load Tuning (Summer vs Winter Tuning; post ~4)

Hybrid Tuning Manual
Hybrid Tuning (Some questions being addressed)

Throttle Tuning
- APP Tables

Hybrid tuning

Self Tuning 1

Self Tuning 2

Self Tuning 3 (APP table info)

Throttle Table Fix

 
#3 · (Edited)
MAF calibration

MAF cal is simple enough as seen in the link above. Just need to find a place for a WOT log to complete the top end tune of the calibration. I'll edit this post with parts of my logs and add a bit more info regarding possible issues/concerns/solutions.

Edit:
Closed Loop MAF cal done. While it seems 4th gear logs are preffered for accurate WOT logs, 3rd hear logs will be good enough. I simply need to enter the right compensation parameters. That being said, somewhat relieved I can tune in 3rd gear given that the speed limits all around here are too damn low, and not enough empty space (generally) to take safe 4th gear pulls.

WOT MAF cal up next...
 
#4 · (Edited)
On to boost!

I remember how much fun it was driving the car on the stock map, and when I switched to the Cobb map (stage 1 at first ) it felt like when I'd put my foot down, nothing would happen. Well finally got around to checking out the base maps boost curve and here it is:


We can see that boost comes on quite early but dies after about 5500 rpm. So good for low end crap for top end.

Here is the Cobb stage 2 map compatible with my current power mods:



You can see that here boost doesn't really come on until the accel pedal is at around 50%. This means that if I'm trying to pass somebody and accelerate slightly nothing really happens. Boost is really on/off so nothing happens then you get punched with everything.

My idea is to blend the OTS and stock map with the help of Excel, then interpolate the values in order to get a smoother curve and a driving experience more to my liking.

Here is gonna be the area of the stock map and OTS map that I will want to blend respectively (highlighted in white):







Next step would be to adjust the WGDC (and AFR?) table accordingly. I'll use the same method on excel as the boost tables in order to obtain some intermediate values to facilitate the transition. Minor tweaking will be done if needed.

Here are going to be the highlighted values of the Stock then OTS maps that I will try and blend:

WGDC:








I'll be posting up the excel sheets to give an idea of what I did and get input if it even made sense.

This will take a while so some updates might not happen for long periods of time. Anyhow, here we go!
 
#7 · (Edited)
Well I got bored in class so decided to quickly combine the boost targets.

Can't post up the excel sheet yet but basically here is all I did:

1) Selected the values between 18.75% and 56.25% throttle position all the way to 7k RPM.
2) Found the difference between the Stock table and the Stage 2 OTS table.
3) Took half the values of the difference and added it to the respective values in the OTS table.
4) Some negative values will appear which mean a reduction of boost.... NO F THAT! (Last two lines between 5.5k-7k RPM)
5) I increased those values and tried to smooth out the end so they are constant values. Despite being higher values then the OTS map, they still remain lower then the Stock values.

Here is a preview!



That little jump at about 56% throttle kinda of bothers me. I'll try and smooth off so it's more linear. Then again the only time i'll need full boost is when i floor it. Regardless, what do you guys think! Should i smooth it out?

WGDC adjustments might need to follow.

*Can't attach Excel files so here are a few screenshots*

Here are the stock boost targets followed by the OTS boost targets. What's in yellow is what I tried to "blend":





And here is the result of both of them together.




I was worried that too much boost in the low-end would be dangerous. But if we compare it to the Stock map, the ECU targets less boost then that. I'm assuming Mazda engineers did their research with regards to reliability, so it seems safe to me to target boost values that are higher then Cobb but under the Stock ones.
 
#9 · (Edited)
Contacted "Enki" on MSF. This guy is a frkn walking (or should i say driving :p) encyclopedia.

Here's what he told me:

Set your boost targets, update the boost throttle close and fuel cut tables (minimum of +1 for throttle close, +2 for fuel cut), then just do pulls until it's hitting your targets; make sure you're in the same gear for each pull, as dropping to a lower gear will make the adjustment more aggressive vs less aggressive in higher gears (due to there being more time to build boost).

AFRs should not need to be touched, though you will likely have to update WGDC to hit your commanded boost levels, since the normal maximum undercompensation adjustment is 20% (if wgdc is 50%, and you need 80% to hit target, you will only be able to hit 70% by default).

If it takes more than 3-4 pulls to start meeting boost targets, you can hit the WG Duty Boost Error Comp table under wastegate dynamics and double the values (highlight all, hit m, enter 2, ok) and that will make your under and overrun boost compensation twice as aggressive.
I wasn't entirely sure what other parameters would have to be tweaked following my boost increase. Being in the mid-range it seems i won't need to directly change the WGDC rather the compensation (Nor the fuel cut or throttle close tables as calculated by an excel file posted by Cobb). I'll take it for a spin and do a few pulls to see how it feels.

I'll finish with my MAF cal before changing any other parameters.
 
#10 · (Edited)
So I decided to review my values on my "custom" boost map. Well I should of payed more attention before because a lot of them were crappy. Tapering off towards higher rpm where it shouldn't of and in a couple places some random dips in values.

Here is the edited version with the excel values as well if you ever get curious to compare to the "previous" map.





If you guys notice anything in these values or have a suggestion let me know!

-edit: the row for 500 RPMs has been interpolated so that the values rise at a consistent rate rather than having a random dip in the middle.

-edit 2: Made another revision. Seems close to if not the final version. Here it is.



Here is the excel sheet with the values in red indicating and increase and the values in yellow indicating a decrease compared to the original/initial blend map. (post 7 1st graph and last excel sheet)

 
#11 ·
Finally got around to going for a proper drive, mainly highway but some city as well. Didn't log anything but did monitor my parameters as well as display the max and min values that were hit for boost and AFR.

Initial Impressions:
Car accelerates with less effort, meaning for the same throttle input compared to the OTS map, the turbo is in a smaller vacuum (negative PSI). So the acceleration/throttle is more responsive. Passing is also easier since with a slight increase in pedal position, some boost comes on. All in all it feels like more linear acceleration.

Concerns/remarks:
Experienced fuel cut in 5th gear going WOT (I believe I went WOT at around 3K RPM but let off soon). The max registered value was 20.5 PSI which is where the fuel cut table is set. Retried it again by resetting my max values and didn't go above 19.1 PSI which is in range of the 18.5.

*I might increase the targeted values to 19 but try to hit it within a much smaller degree error. So lets say +/- 0.5 rather then 18.5 and +/- 1.5

I'm running slighlty rich. Of course that's because my MAF hasn't yet been calibrated. I'm assuming that can be part of the reason there was a boost spike (or the cold weather? :) )

Just need to drive the car for another 60 km for the car to learn some fuel trims before doing some closed loop logs for the maf cal.

All in all I'm quite happy with the initial changes!
 
#12 ·
Alright well did a bit more reading (post 2 updated) and it seems as though i'll probably have to tune my throttle tables in order to achieve the desired throttle response. I ended up tuning my boost tables.
It might still work but I PMd snailD on MSF who posted Throttle Tables to ensure what's the best way of proceeding.

Also with regards to the WGDC tables. The stock turbo (K04) can hit 7-10 psi on 0% WGDC. This means i can go ahead and zero out the respective cells for where my boost targets are under ~10 psi. Assuming i am not hitting the target boost, the compensation table should have enough liberty to adjust the WGDC value in order to achieve the desired boost.
 
#13 ·
Update time!

Started reading the whole throttle tuning table thread of MSF (Check post #2 for a link).

Had a few questions so got into contact with the OP (snailD). Here's is the convo we had. Hopefully this can answer some questions other people might have.

Throttle Tuning
Hey man, I just read your whole thread about the throttle table fix and I had a few questions that i was hoping you could answer:

1) You mentioned that you altered your boost targets so they are more suitable for mid range values. Before this thread, I had thought about doing the same mainly because i wasn't happy with the throttle response or how boost came on. I ended up blending the stage 2 OTS boost values with the stock values so the boost curve looks something like this
Here are the values (ignore the fact that they're highlighted)
The car feels more responsive since i'm able to get into boost with smaller throttle input (vs stock ots). Did i do this backwards. Should i just leave the boost targets like the OTS map and simply adjust throttle? Or i a slight increase in boost and in the throttle tables ideal for maintaining mid range boost?

Last question, with regards to the APP tables. A smaller value means less input is needed for greater acceleration or does a larger value mean faster acceleration? (I believe it's the first case but I'd just like to double check).

Thank you for your time and help!

personally on the boost tables, i lowered max boost until 87.5 on my car. then blended it all the way to the negative. this means 50% throttle is more like 50% boost, instead of 50% throttle being 100% boost, which makes driving canyon roads harder to control.

just think of app as your foot actually pressing the pedal down. if you have a linear app table where every number in the graph represents the number above, what you give is what you get. if your numbers are higher, you get more throttle with the same input with your foot. if the numbers are lower you get less throttle for the same input with your foot.

there is a delicate line between the app tables and the DBW tables. what i posted is what i found to work the best after weeks of testing shit out. the first map i did wasn't even driveable it was insane lol

So in terms of how boost comes on that has more to do with the throttle tables rather then the actual boost table?

So assuming I want more acceleration at say 25% throttle I should be adjusting the throttle tables rather then increasing boost in that area? Or does that end up being similarly effective?

how boost comes on is really all in the boost tables. but your app tables may tell the car you want 50% throttle when you are really at 25% depending on how u have them set up. the car will then look at what 50% throttle is requesting in the boost tables and try to run that boost
Last question was is it more effective to alter the APP tables rather then the boost values or if one or the other was better, here's what he said:

personally if you still don't completely understand it, i would leave the app tables and adjust the boost tables. its just easier to understand that way
So basically with the current way the boost table is set up, I adapted the boost table to the Stage 2 OTS values in the throttle tables rather than increasing the values in the APP and DBW tables.
I'll create a map with stock boost values and his suggest throttle tables then drive my version and his back to back to see what i prefer.
 
#14 ·
So I made a new map with basically the OTS boost values but smoothed out. Made a few more iterations of the map with different variables changed : DBW table as posted by snailD in the throttle table thread. Then another one with his APP table and a last one with my "custom APP" table. I'll see what I like best then adjust accordingly.

Before that however I need to complete my MAF cal
 
#15 ·
Spoke to a few e-tuners on MSF with regards to taking appropriate WOT logs.

Main main concern was:
- 4th gear logs might not be the safest (ticket wise) to perform depending on where you live
- Would 3rd gear logs be adequate? And if so, is there any way you'd compensate?

Answers:
- 4th gear logs put a higher load on the engine, yielding more accurate results. However a 3rd gear WOT log would be sufficient.

- Quoted from ms3blackmica on MSF with regards to "compensating"
3rd gear will be sufficient enough so long as you understand it will not load the engine as much as 4th would so you may desire to underboost a little bit from target to account for the fact that the turbo will create more boost in 4th
- He also added that towards the end of the tuning process, you would still need to perform 1-2 4th gear pulls just to make sure everything checks out.

So the general consensus is 3rd gear is adequate enough as long as you compensate appropiratley or adjust some other parameters. 4th gear should be done at the end to confirm the data.
 
#17 ·
Warm temps means I'm back! I'm actually super excited for this update since it means tuning is not far away! :D

Alrighty so I don't think I updated what my custom boost curve looked like. Well here it is:





Most recent changes were changing the DBW and APP tables (throttle tables). Also had to swap to a modified version of the Cobb OTS map for it to be more linear, in line with the throttle tables. Link to throttle tables in post # 2

Here is the new boost curve



Now i'll post side by side pictures of both the DBW tables and APP tables (first old then updated)

DBW




APP




More updates to come later on in the summer. Full hybrid tune and so on. I'll go step by step and explain.

If there's anything i wrote before hand that i disagree with or that now sounds stupid i'll modify it as well :p
 
#18 · (Edited)
Lol while i was looking at the OTS boost table i realized it was an ugly thing. So i made it prettier.

this:



time to reflash

It's basically totally linear without any humps or weird spots.


** Also noticed i hadn't corrected the boost limiter tables (fuel cut and throttle close). These NEED to be adjusted accordingly every time you modify your boost curve.
 
#19 ·
Keep on buildin on my man!

Great stuff.

One day goin to trade the HT for a cobb AP.

This will come in very handy. Awesome work man.

You are having lots of fun :)

Looking forward to playing with the AP one day.

So I would say any performance mods are on hold for now? You look like your pertty much done except for the meth kit?
 
#20 ·
Lol I haven't really done much at this point. The links in the second post are of more useful.

I'm hoping I can at least keep posting info for people like you.

Performance mods I'm basically maxed out. As for a meth kit, it's probably better for me to put that towards suspension rather than more power. It's got enough and it's fwd so adding more would only go towards street cred and roll racing (which I don't care or do)
 
#21 ·
Alrighty. More info for yall thanks to a very knowledgeable mister.

Performing logs:

- Do 3 consecutive logs. That way you use the "majority rule" for any data that happens to vary. Assuming you get the same value in 2 logs, and a different value in the third log, you would use the value that appeared twice.

Knock Decay Table:

Here's a quote

I found this table basically dictates how long the ecu attacks a knock encounter.

The ots and stock values carry on and on. Reducing timing.

Most times when it's not even needed after knock has been squashed

This is good for and "umbrella" timing table that will work for many cars in many regions.

Since you customizing your timing you can take advantage of a more aggressive setting here.

I like to use a setting of 150.

This makes the ecu come in, pull timing and richen the afr which squashes the knock. then the get the fuck outta the way and go back to flagging the appropriate timing table.

That's one reason it's good to use fast decay.

The other reason is, if knock comes right back then you know you've got a real issue in that region of load and rpm.
Then I asked, if I see the knock coming back in that rpm range, does that mean that is my max value for timing?

A: Yes. At that point timing as increased too much.

That means you pull it back a degree or so.

- Temperature (for timing tuning)

The temp won't affect timing, however the grade of gas will.

-----------------

Towards the end of April I should be gathering quite a little more info, and I'll try to do somewhat of a "step by step" guide as to what to do/what to tune first, some key things to look into, and then what to do and so on.
 
#22 ·
Small update. Was experiencing quite a bit of knock at part throttle, and after a new refill, at WOT.
Refilled at Shell instead of at Esso and the knock basically disappeared. Thankfully it was fuel related.

Spark Plugs

Quick note. The stock k04 turbo prefers stock heat range plugs. There really isn't a need for one step colder 99% of the time on the k04. Once you upgrade to a bigger snail, that's a different story.

Since I'm so smart, I bought new plugs before doing ALL the research and making sure. Oh well
 
#26 ·
Alright a little more info for you guys!

Spoke with one of the tuners, I was asking him regarding leaning out AFRs

As far as gaining power from leaner AFRs at WOT, he said he couldn't quantify any gains in terms of power based on the data (logs) he had. However, it "felt" as if the turbo spooled faster. Then again, going from 11.4 to 11.6, it might just be in your head.

Our DI engines tend to make more power at higher timing levels, so it may be worth simply enriching towards redline where timing begins to induce knock, rather than pulling timing at a lean mixture.

If the injector duty cycle gets high, then we can lean out the mixture a bit.

With that being said, I should be beginning to make progress sometime next week.
 
#27 ·
Interesting stuff. all new to me though since I have never done a custom tune on a FI. Never done a custom tune at all lol.

I am curious with leaning AFR's what numbers you will be changing to try and keep timing up and enriching at the same time on the way to redline while keeping knock down and increasing power?

Did I get that right? Still trying to grasp the leaning of AFR's and your plans.

Is it worth going to redline or when the turbo stops? Thought I read our turbo's stop spooling before we redline?

Good stuff man
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top