OpenFlash Tablet off-the-shelf E85 Datalogs on stock car

[Shiv@Openflash]

Originally Posted by BobbyBlue86

Here is my log. Car is stock aside from a K&N drop in...
@shiv@vishnu The car feels great, but that VVT exhaust hump between 3800-4500 is a little unnerving... makes quite a sound!

Let me know what you think!

Running 1.42 stage 2 EL E85

http://datazap.me/u/khallin/log-1-10...data=1-2-5-6-9

Looks really nice. AFR is pretty much perfect. Judging by the negative fuel trim, your ethanol content isn't very high (probably around 60%). That explains advance multiplier dropped down slightly to 0.88. Still completely safe and sound. And the change in intake noise at 38000rpm is normal with the EL tune. Just a side-effect of the VVT mapping that is necessary to reduce the torque dip in that RPM range. If you want to get rid of it (and the dip), get a UEL header and run the UEL E85 tune. It will make considerably more power and sound better

[BobbyBlue86]

Originally Posted by shiv@vishnu

Looks really nice. AFR is pretty much perfect. Judging by the negative fuel trim, your ethanol content isn't very high (probably around 60%). That explains advance multiplier dropped down slightly to 0.88. Still completely safe and sound. And the change in intake noise at 38000rpm is normal with the EL tune. Just a side-effect of the VVT mapping that is necessary to reduce the torque dip in that RPM range. If you want to get rid of it (and the dip), get a UEL header and run the UEL E85 tune. It will make considerably more power and sound better

Well Im waiting for my tax return so I can pick up your OFH

Seriously good job though. The car really feels like it has way more life in it and its not so rough at higher revs. I am really excited with the direction you guys are moving. I just hope that one day you guys will have some tunes for the Innovate SC... as that is my goal further down the road.

The fuel, however, is probably because I had roughly 2-4 gallons of 91 in my tank on fill up.

[steve99]

Originally Posted by shiv@vishnu

Looks really nice. AFR is pretty much perfect. Judging by the negative fuel trim, your ethanol content isn't very high (probably around 60%). That explains advance multiplier dropped down slightly to 0.88. Still completely safe and sound. And the change in intake noise at 38000rpm is normal with the EL tune. Just a side-effect of the VVT mapping that is necessary to reduce the torque dip in that RPM range. If you want to get rid of it (and the dip), get a UEL header and run the UEL E85 tune. It will make considerably more power and sound better

Hi Shiv,

Interested in running this tune in Australia. The E85 at the pump is rated at 105 RON (petrol rated at 98 RON ,think i'st equivalent to your 93 in US)

Will E85 rated at 105 RON (Australian) be sufficient for this tune on standard car ??

[jflogerzi]

Originally Posted by shiv@vishnu

Looks really nice. AFR is pretty much perfect. Judging by the negative fuel trim, your ethanol content isn't very high (probably around 60%). That explains advance multiplier dropped down slightly to 0.88. Still completely safe and sound. And the change in intake noise at 38000rpm is normal with the EL tune. Just a side-effect of the VVT mapping that is necessary to reduce the torque dip in that RPM range. If you want to get rid of it (and the dip), get a UEL header and run the UEL E85 tune. It will make considerably more power and sound better

OMG I can't wait!!!

[Fizz]

I was thinking of using the E85 tune as well but there's only 1 station in my area, not very close to my home anyway. Would be simply too inconvenient for a daily driver.....

Hoping E85 become available more easily in future

[Shiv@Openflash]

Originally Posted by steve99

Hi Shiv,

Interested in running this tune in Australia. The E85 at the pump is rated at 105 RON (petrol rated at 98 RON ,think i'st equivalent to your 93 in US)

Will E85 rated at 105 RON (Australian) be sufficient for this tune on standard car ??

Unfortunately, I have no experience with your E85 but I suspect it will be fine. But it's a good idea to test carefully when first running it with the E85 maps. Just for peace of mind

[steve99]

Originally Posted by shiv@vishnu

Unfortunately, I have no experience with your E85 but I suspect it will be fine. But it's a good idea to test carefully when first running it with the E85 maps. Just for peace of mind

H'mm might wait till logging update for IAM monitoring I think

[Shiv@Openflash]

Originally Posted by steve99

H'mm might wait till logging update for IAM monitoring I think

You don't need iam logging. You can just look at kc learned numbers. That's actually a better approach when first testing to knock since iam usually takes more time to drop when knock occurs.

[steve99]

Originally Posted by shiv@vishnu

You don't need iam logging. You can just look at kc learned numbers. That's actually a better approach when first testing to knock since iam usually takes more time to drop when knock occurs.

Ok what KC learned values should I be looking for

Value above XX
Stays above XX for certain amount of time

Currently don't see above 6 and only short duration spikes few seconds most often on lower RPM ranges 1500-4000rpm this is on 98 RON petrol

What is dangerous ??


Won't have to worry about cold start here if it gets below 50F everyone would be panicking thing the next ice age is coming

Thanks

[steve99]

Originally Posted by shiv@vishnu

You don't need iam logging. You can just look at kc learned numbers. That's actually a better approach when first testing to knock since iam usually takes more time to drop when knock occurs.

Thanks shiv and Jamesm

Think I have this "KC Leaned" parameter sorted now. spent last hour or so comparing Engine Load Vs RPM Ignition advance table values against KC Learned value's from my logs. Can now see why your tunes pulled advance in 1500-3000 Rpm area against stock, evidence of knock when I was on stock ROM (2 or 4 degrees discrepancy between KC Learned and advance table values) but very little if any on 1.42 or 1.53 tunes

The IAM dropping is and indication you need to look deeper at "KC Learned" and see where your knock occurrs to pull timing I now realise

Hopefully I am on the right track

[jamesm]

Originally Posted by steve99

Thanks shiv and Jamesm

Think I have this "KC Leaned" parameter sorted now. spent last hour or so comparing Engine Load Vs RPM Ignition advance table values against KC Learned value's from my logs. Can now see why your tunes pulled advance in 1500-3000 Rpm area against stock, evidence of knock when I was on stock ROM (2 or 4 degrees discrepancy between KC Learned and advance table values) but very little if any on 1.42 or 1.53 tunes

The IAM dropping is and indication you need to look deeper at "KC Learned" and see where your knock occurrs to pull timing I now realise

Hopefully I am on the right track

to expand on that... it is important to understand that there are three knock control strategies implemented in the stock logic. fbkc, flkc, and coarse correction. only one is ever active at any given time.

what this means is that just because you don't see knock in one place doesn't mean it wasn't there. you have to see the whole picture. knock correction learned value will show you the result (but not the actual value of) flkc and coarse correction. coarse correction is your IAM btw.

so, if you have a targeted advance value of 5deg.. and your knock correction learn value is 2.5, it means either your iam is 0.5, or that cell in flkc is pulling 2.5deg (which isn't a valid value, they're incremented by 0.35 in the stock rom, but whatever lol), or some combination of the two. so, kclv essentially shows you what's happening in two of the modes. fbkc is just instantaneous correction.. dumb knock correction basically.

it's also important to note that the ecu has no notion of knock severity. the knock flag is a boolean value.. i.e. true or false. every x number of cycles the ecu checks this value, and if it's true, it decrements something, if it's false it increments something or does nothing depending on what mode it is currently in. remember we're talking fractions of a millisecond here, which is why it seems like they're all working at once, though they arent.

so it works something like this (the short version): by default you're running around in fbkc. when significant knock is detected by the ecu (greater than 3.1 deg of correction iirc), it'll write a value into flkc. if the value written into flkc is large enough (again i believe 3.1deg), it will decrement the IAM by the value set in the rom. after a while of not seeing knock in a load/rpm range, it will increment flkc in that range. this means that it's only possible for flkc to be positive if the iam is less than one.

that's just a very basic overview of how things work, but if anyone would like a more detailed explanation feel free to pm me and i'll do the best i can.

[steve99]

Originally Posted by jamesm

to expand on that... it is important to understand that there are three knock control strategies implemented in the stock logic. fbkc, flkc, and coarse correction. only one is ever active at any given time.

what this means is that just because you don't see knock in one place doesn't mean it wasn't there. you have to see the whole picture. knock correction learned value will show you the result (but not the actual value of) flkc and coarse correction. coarse correction is your IAM btw.

so, if you have a targeted advance value of 5deg.. and your knock correction learn value is 2.5, it means either your iam is 0.5, or that cell in flkc is pulling 2.5deg (which isn't a valid value, they're incremented by 0.35 in the stock rom, but whatever lol), or some combination of the two. so, kclv essentially shows you what's happening in two of the modes. fbkc is just instantaneous correction.. dumb knock correction basically.

it's also important to note that the ecu has no notion of knock severity. the knock flag is a boolean value.. i.e. true or false. every x number of cycles the ecu checks this value, and if it's true, it decrements something, if it's false it increments something or does nothing depending on what mode it is currently in. remember we're talking fractions of a millisecond here, which is why it seems like they're all working at once, though they arent.

so it works something like this (the short version): by default you're running around in fbkc. when significant knock is detected by the ecu (greater than 3.1 deg of correction iirc), it'll write a value into flkc. if the value written into flkc is large enough (again i believe 3.1deg), it will decrement the IAM by the value set in the rom. after a while of not seeing knock in a load/rpm range, it will increment flkc in that range. this means that it's only possible for flkc to be positive if the iam is less than one.

that's just a very basic overview of how things work, but if anyone would like a more detailed explanation feel free to pm me and i'll do the best i can.

Thanks jamesm,

Every day i learn something new is a good day

will digest info above , might take you up on your offer once i have my head around this.

[s2d4]

Originally Posted by jamesm

so it works something like this (the short version): by default you're running around in fbkc. when significant knock is detected by the ecu (greater than 3.1 deg of correction iirc), it'll write a value into flkc. if the value written into flkc is large enough (again i believe 3.1deg), it will decrement the IAM by the value set in the rom. after a while of not seeing knock in a load/rpm range, it will increment flkc in that range. this means that it's only possible for flkc to be positive if the iam is less than one.

Doesn't it go into fbkc when there is a rapid change in engine load?
so default would be flkc? until driven spiritedly, WOT etc?

[jamesm]

Originally Posted by s2d4

Doesn't it go into fbkc when there is a rapid change in engine load?
so default would be flkc? until driven spiritedly, WOT etc?

nope... fbkc is what you 'roll around in' but that's probably a terrible way to put it given that we're talking about extremely short time intervals here lol. flkc is only written to after the ecu sees knock activity or when the engine is in a load/rpm range that the value is currently non-zero, as i understand it (could be wrong though, or things may have changed since the 16/32bit wrx ecus most of the available information is based on). fbkc is the default 'reactive' knock control mode. i guess 'default' might not be the right word. think of it as 'the one you're in when nothing else is happening' lol.

there's talk of a notion of pre-emptive knock control that is based on the load delta, but it's not well defined and i've only been able to find a little bit of (old) information on it.