love it! feel the pain!!!

 

blue-

if you have LS2 coils, the next thing you should do is try an get your dwell settings up to snuff. Would be a nice 'proof of concept' for the S2 guys who are looking to do that.

 

And we can certainly point you the right table, assuming you have the "Truck" coils!


Or the opposite... the point that the ECU considers the motor to no longer be in cranking mode... I don't think it cares too much about stalling

 

Indeed... so much to do, so little time

Sorry, stock coils here, and I have a 2nd set of stock coils on order. Is there anyway I can practically measure changes to dwell settings? I'm happy to try changing them to prove it can be done. There is just one table for that so it looks pretty easy, maybe this is proof enough :-)

 

I don't think there's much more to prove at this point... at least as far as what my eyes have seen. What you see is what you get... now go strap a turbo on that beast and drive the **** out of it!

 

lol shelldude what is this unicorn car you're referring to?

pics or it didn't happen!

 

At least in the piston world, commanded torque is used to calculate gear selection and, I believe (not totally sure), throttle position on drive-by-wire AT cars. It seems entirely logical to use it as a basis for throttle position in drive-by-wire manuals as well, but I'm not sure if this is done.

All this activity makes me really want to get into tuning my R3, but I'm a little worried if I pull the trigger on ME, VT will release and I'll regret it.

 

Well I'm now on my 2nd MAF calibration and iterating...

Anyone have any idea what "Closed loop fuel correction A" and "Closed loop fuel correction B" do? I don't think I need to change them, but am curious.

 

Just what it says ................. you can adjust closed loop fueling with those tables . Although I tried the other day and couldn't get the map to adjust the numbers the way i wanted . Something screwy with the way the number is derived by ME in each cel ................ I had no problem adjusting them in Protuner.

 

Thanks, I was thinking that, but was not sure, and was also uncertain as to what the numbers mean (%, offset, ...?). Anyway, I am not too interested in changing them, but it is interesting that mazda have some set to non-zero values implying they are running closed loop at something other than stoic. (or compensating for something to get to true stoic)

 

Just for fun, here is a graph showing the average of the fuel correction for each MAF voltage reading after my first edit to the MAF table where I shifted the entire table by a fixed % from stock.



Ignoring the data for the higher voltage readings (anything >2.8V) since there are too few data points in my logs to draw any conclusions there, you can see that the factory curve is pretty close to the real curve (within +/-3%) and that my initial change did a good job of centering the corrections around 0% (the average fuel correction across all samples from this set of logs was -0.57% and I had made a correction of 5.8% based on my previous logs).

Based on this data, I have now made some changes to the factory curve (by that I mean adjusting each MAF table cell by an individually calculated %).

Now I need to do some more logging, and work out how to get some more useful data for higher MAF voltages...

 

hmmmmm - wouldn't recommend this methodology . Unless your maf tube is significantly different from stock .

 

Any reason why not? I figured that this was a reasonable first step before adjusting individual parts of the table since this is what the LTFT is doing anyway

 

Put it this way .... I have f*cked around with the maf table like you wouldn't believe . 5 years later + multiple tunes to RX8s has taught me to only mess with the first part of the table that affects LTFT . UNLESS I suspect the maf tube design varies the voltage output relative to stock . EG - it is a different diameter or has a bend real close etc.

Remember when you alter the maf table it affects ALL load ranges - not just the one you logged .

Far better to mess with the actual fueling tables once you have LTFT under control . Also - if LTFT is high , don't just blindly alter the maf table without investigating for possible causes.

 

^good to know, thanks, I still have a long way to go in my learning here and some of this I will have to prove to myself just because that is the sort of person I am... So, several more experiments planned, but I am waiting for my mid-pipe before going crazy with the fueling tables and logging open loop behavior.

One thing I am wondering right now is how accurate the factory (semi)wideband is?

I know on my civic there is a correction factor you can manually apply to reported AFR readings (the ECU does not have a table for this, but you can convert your logs to get calibrated AFR values). Hopefully I'll be able to get to a dyno in around a month so I can do some testing there.

EDIT: there is a good, but old, discussion on the factory sensor for the s1 here: https://www.rx8club.com/series-i-aft...anyway-138952/

There is also an article here that implies sensor wear over time causes the calibration to drift: http://www.innovatemotorsports.com/r...ycalibrate.php

And the referenced document: http://www.promazdachampionship.com/...03mar21eng.pdf
(pages 8 and 9 are the interesting ones, note that the previous article cherry picked the data for extreme rich and lean conditions, the sensor is much more accurate at 14.7AFR/lambda=1 and while the test shows the accuracy gets 3x to 4x worse at the extreme ends of the measurement scale, it barely gets any worse at lambda=1)

 

Without reading any of that I can tell you that yes the factory wideband is accurate at 14.7 but tends to err on the rich side as it gets closer to it's limit (11.1) by as much as 0.8 . Have never seen one err to the lean side .
The unfortunate thing for FI owners is that we really want to run around that minimum level so safely tuning to achieve that is a difficult task.

 

Hmm, so I'm starting to suspect that at anything over about 3V from the MAF sensor, or 100g/s the ECU is NOT using the MAF table to calculate fuel. At least, not the MAF table we have access to...

 

Oh dear - you got problems ......

 

any updates??

 

Not yet from my side, I've been quite busy with work. I'm hoping my problem is just a tuning one rather than a mechanical one. I'll be at Thunderhill in less than 2 weeks and hopefully will be able to get some good open loop data and even a little dyno time while I am there.

 

Talked with Steve at VT last night (late too) and he was up working on the software. Wrapping up on the failsafe checks and such. Should have a beta in my hands soon.

 

Glad to hear VT is still pushing forward. Just got new exhaust put on this week.

 

here is an excerpt from the email

 

I am gonna wait for the VT version to release, I want to supercharge my 2009 R3... but will do this this next winter, so not that much in hurry.

Been waiting for few years for a tuning solution for this car, now its around the corner great.

 

I don't get it .............. ME is already available and proven (although it's not perfect) . Why wait for something that is in development and unproven.