Knock fix with Australian Fuel 98/100 RON

[steve99]

Originally Posted by Turdinator

I have been daily driving my car on Shiv's stage 1 tune and my IAM has been rock steady at 1 on Shell V-power. I plan to take some logs this weekend and see where things are at more specifically as per your first post. Am i just lucky that it has stayed at 1 or is it more likely the cool weather?

P.S. @steve99 thank you for keeping the first post up to date

Yeah probably cooler weather and maybe driving style. You usually only get the iam drop when your in 2000-3500 and stomp on the throttle in high gear. and it recovers pretty quick usually, i worry more about the fbck and the flck especially at upper rpm which does not drop the iam. logging is the only way its too quick to see real time and you cann't hear it at upper rpm. The flck logging is still a bit sus on oft till they fix it so best to log kc learned as well and look for knock as in first post

[Turdinator]

This is my first log. I did it on my commute home tonight so its all over the place I'm sorry. That saqid i would love to get peoples thoughts on it.

http://www.datazap.me/u/turdinator/1...=0&data=1-9-11

[steve99]

Originally Posted by Turdinator

This is my first log. I did it on my commute home tonight so its all over the place I'm sorry. That saqid i would love to get peoples thoughts on it.

http://www.datazap.me/u/turdinator/1...=0&data=1-9-11

Looks ok knock wise 5000-7500 carn't see any sharp dips in KC Learned, the FLKC logging looks like its not working property yet for G series.

your AFR at higher rpm is pretty rich 11.3 which is probably why their is no knock top end. shiv target 11.6-11.8. Commanded AFR is 12.5

That's just higher end MAF scaling out a bit. I think they do it on purpose.

The LTFT is pretty high -6 to so low end. again MAF scaling.

No great problems, however if you do lean in out top end to 11.8 you will probably get a bit of knock and you will need to pull timing.

If you run united 100 octane which has 10 % ethanol it will lean out AFR a bit and has better knock resistance. Think shiv tunes done on their fuel which has 10%E so on straight petrol they tend to run a bit rich.

Their is a bit of FBCK knock arround the usual 2000-3000 area just when you jump on throttle max is about -1.4 you just don't do it often enough to get the IAM to drop.

http://www.datazap.me/u/turdinator/1...5870-4948-4523

[Turdinator]

Thank you @steve99. I saw my LTFT was high as I finished my drive home. I plan to scale the MAF before worrying about anything else at this point. Although I suspect it will be a slow process.

Thanks again

[steve99]

Originally Posted by Turdinator

Thank you @steve99. I saw my LTFT was high as I finished my drive home. I plan to scale the MAF before worrying about anything else at this point. Although I suspect it will be a slow process.

Thanks again

Just watch if you use the automated maf scaling programs as they will adjust to OL maf scaling to achieve 12.5 AFR at high RPM as this is what is in the OL Fueling table. This may be more aggressive than you want, as I said shiv targets 11.8 high rpm on petrol.

[Turdinator]

Originally Posted by steve99

Just watch if you use the automated maf scaling programs as they will adjust to OL maf scaling to achieve 12.5 AFR at high RPM as this is what is in the OL Fueling table. This may be more aggressive than you want, as I said shiv targets 11.8 high rpm on petrol.

Would it not be more beneficial to have the MAF scaled as close as you can get it and to then adjust the fuel targets to get the 11.8?

[steve99]

Originally Posted by Turdinator

Would it not be more beneficial to have the MAF scaled as close as you can get it and to then adjust the fuel targets to get the 11.8?

As I am not a tuner I don't like to answer those kind of questions. It might be better to ask shiv or maybe jamesm.

I believe both will end up with same outcome, and I would assume its better to have MAF scaled correctly and adjust fuel tables to get the OL AFR correct.
@jamesm just posted some good comments in the maf scaling thread re the MAF being used for load calcs as well as fueling.

I just don't understand all the ramifications of each method.

[jamesm]

You'll always want to make sure that you're hitting fueling targets in open loop. There has been some debate in the community over how important hitting targets actually is, since you can always just change the targets to get what you want in real life. I am of the opinion that it is extremely important because the goal of calibrating the input to any estimation algorithm (which load calculation in an ecu is an example of) is to best approximate reality, not just to achieve whatever side effect you intend to achieve. As i stated in the other thread, this isn't just part of a fueling calculation it's part of a load calculation. Your goal when calibrating it should be to optimize the accuracy of the load calculation, not just to hit an arbitrary AFR target.

Besides, it can be done perfectly so there is no reason that it shouldn't be. Anything less is just half-assing it. If it can be done correctly (and being on target is most certainly 'correct' in this sense), then it should be.

TL;DR - Fueling error in open loop is pretty much the best objective assessment you can make as to the accuracy of your airflow model, so it should be as correct as possible.

[AL008]

Hrmm.. Not too sure why my IAM was pulled but I experienced no knock...

http://datazap.me/u/toy-sooby/casual...ata=1-3-5-9-11

[jamesm]

Originally Posted by AL008

Hrmm.. Not too sure why my IAM was pulled but I experienced no knock...

http://datazap.me/u/toy-sooby/casual...ata=1-3-5-9-11

there was tons of knock... just not FBKC. you were pulling timing from FLKC, which isn't loggable via OFT on the australian roms. you can see from watching the applied advance value, where it's down to zero in places.

[jamesm]

Originally Posted by vgi

@jamesm IAM is pulled when both, flkc and fbkc knock are detected or just flkc. I had fbkc going to -5 yet iam stayed at 1, curious if I got correct address for IAM

i've seen IAM drop without a logged instance of FBKC many, many times (in fact it's the common case). i'm not sure if this is because it's not happening, or because the sampling rate of the log just doesn't grab it, but it definitely happens.

[Kodename47]

Originally Posted by jamesm

i've seen IAM drop without a logged instance of FBKC many, many times (in fact it's the common case). i'm not sure if this is because it's not happening, or because the sampling rate of the log just doesn't grab it, but it definitely happens.

I took the literal definition of something you described before and I believe it to be the way it works. You can only have 1 knock control operation active at any time:
- FBKC
- FLKC
- Reduction of IAM


If the knock was extreme enough to instantly reduce IAM without FLKC being reduced then this can happen. Also, check the Knock learn. It certainly looks like it's detected knock in that area.

[jamesm]

Originally Posted by Kodename47

I took the literal definition of something you described before and I believe it to be the way it works. You can only have 1 knock control operation active at any time:
- FBKC
- FLKC
- Reduction of IAM


If the knock was extreme enough to instantly reduce IAM without FLKC being reduced then this can happen. Also, check the Knock learn. It certainly looks like it's detected knock in that area.

The common knowledge is that only one is ever active at a given time (it runs on cycles, like an infinite loop with conditionals determining whether to switch to a different 'mode'), but everything documented on this is based on wrx ecus, so who knows how different the 86 logic is, if at all.

All I know for sure is that the iam can drop without catching any fbkc in the log. I see it happen all the time, and this log is an example of that. It's clearly pulling from flkc, as the applied advance is zero, which can't happen otherwise (given a positive value in the cell) unless iam is zero.

[jamesm]

Originally Posted by vgi

for me it's the other way around, FBKC drops to -5 and IAM stays 1

I've seen that as well. It's f'n voodoo, so I just make sure there's never any knock at all, period.