paddle shifters in drive mode

 

Let's go further

Electronic throttle body (DBW)

Purchased a mazda RX8 throttle (80mm) and made an adapter for mounting to the intake manifold



Evo vs Rx-8



First tests



Electronic pedal



The finalization of the ecu program and mod ECU, tune the idle for all states




and cold start

 

Final result with DBW and 272 cams

You Tube

 

Good day, ladies and gentlemen. Another portion technocracy and update software.
It was evening, there was nothing to do The idea to implement a wide-band lambda control of the mixture arose spontaneously.
Especially now it stands on almost every car and is used for logging.
It would be very convenient if the control unit itself adjusted the injection pulse width for the target AFR based on the WBO readings. In some small window, for example +-10%
I think many people understand that the stock ECU evo\airtrek\lancer can only work with a narrow-band lambda and adjust the mixture relative to the stoichiometry (14.7 AFR), by constantly switching richly\poor around this point. The whole thing only works at idle + cruise.

Who wants to know more about the narrow band lambda https://en.wikipedia.org/wiki/Oxygen_sensor

Therefore, at the first stage, the stock algorithm was first analyzed, and then adapted to the data of the WBO.
A small part of the description of the algorithm



It turned out to be a kind of narrow-band control with a target from the map.
orange - target AFR
yellow - wbo
green - correction
Looks of course so-so, you can see how the mixture is hanging out





The implementation has some semblance of a simple PID controller, but it is not very suitable for our purposes.
Some changes to the parameters made it better, but in any case, the algorithm needs to be rewritten.
Then I implemented my own PID with configurable parameters. After that: tuning, testing, tuning😁 for which many thanks to fill7878
And the first results
The first half of the log is idling at 14.7 afr(the mixture is stable, the correction works out as it should)
The second half — 3000 rpm 15.0 afr (you can see that there is an overshoot and you need to make some edits to the algorithm and coefficients)
orange - wbo
blue - target AFR
green - correction



We are on the right way
Making adjustments to the algorithm. The test is the same (Idle-14.7 afr, 3000 rpm-15.0 afr)



The cruise afr is mega stable, which is very pleasing 😂

Following test
I edit the VE map (+-10%) so that the mixture crawls away in some modes.
Trying on the boost ~ 1.4 bar. Four gears in the floor to 6000-7000 rpm.
green - rpm
orange - target afr
red - wbo
blue - correction



There are still some points to refine and configure. In particular, switching off the mixture correction in some transient modes (on decel fuel cut, accel enrichment delta TPS, etc.), etc. I'll add a long term correction for even greater accuracy.
But in General it turned out very good and accurate 😁
Stay tuned

 

Very good Mr Guru !!

 

Thanks buddy!

Some improvement for subi paddle shifters. Why not?🤪

 

I'm running 6 sec delay with my Evo X paddle shifts in drive mode. What time delay do you run ?

 

no delay

Try to check it out on this video, 8 second. If you look closely, you can see the switch being pressed and immediately switching gear

 

Video looks good! I use a 0-30sec delay relay to time out my gear after my set 6sec.
My shifts are instant. 6sec gives me engine braking 100 mtrs before a corner and through the
corner before timing out. It then shifts up automaticaly if I'm not heavy on the pedal. I can
push the shifter into tippy after I shift or hold my paddle if I want to override the 6sec.
My downshift is on my right. My setup is external and yours is prob ecu guru modded.