TephraMODs for the X
yeah well it would be load and knocksum based... really you would only be interested in 180+load - which you probably wouldn't be hitting mid corner...
but you would also be able to control the output % as well, maybe ill make it a factor of your current TPS, so 50% of whatever it was meant to be - rather than just 0%
but you would also be able to control the output % as well, maybe ill make it a factor of your current TPS, so 50% of whatever it was meant to be - rather than just 0%
Thread Starter
Joined: Feb 2007
Posts: 9,486
Likes: 67
From: Melbourne, Australia
Thread Starter
Joined: Feb 2007
Posts: 9,486
Likes: 67
From: Melbourne, Australia
ok
just a bit of an update.
I've revised the knockcel code.
Previously as soon as the load and knocksum threshold was reached KnockCEL started flashing (slow or fast depending on which KnockSum threshold was crossed)
Now say your load threshold was set to 150 and knocksum threshold at 4.
This meant that if the knocksum was at 10 at load 100, and was on it's way down when you passed 150load then it would flash... Which is not very accurate...
So what I have done is only enable the flashing when the knocksum this loop through is GREATER than last time... this means it wont flash in the above scenario.
I have also done the TPSCut based on High KnockSum - let me know if you would like another set of tunable knobs for the TPSCut, rather than just reusing the KnockCEL ones..
So basically:
Load threshold = minimum load for knockcel + knocktpscut to work
Low Knocksum threshold = minimum knocksum for SLOW flashing
High Knocksum threshold = minimum knocksum for FAST flashing AND TPSCut
There is a TPSCut attirbute, which is a percentage applied to the current TPS %
ie, if TPSCut detects a High KnockSum event then it will multiply the current TPS % by this
therefor to disable TPSCut you could set this to:
* 100% - ie leave the current TPS alone
* 50% - ie use 50% of the current TPS signal
* 25% - ie use 25% of the current TPS signal
* anywhere in between
I will be testing this tonight/tommorow and then I will do up a 52680022 ROM for the testers
just a bit of an update.
I've revised the knockcel code.
Previously as soon as the load and knocksum threshold was reached KnockCEL started flashing (slow or fast depending on which KnockSum threshold was crossed)
Now say your load threshold was set to 150 and knocksum threshold at 4.
This meant that if the knocksum was at 10 at load 100, and was on it's way down when you passed 150load then it would flash... Which is not very accurate...
So what I have done is only enable the flashing when the knocksum this loop through is GREATER than last time... this means it wont flash in the above scenario.
I have also done the TPSCut based on High KnockSum - let me know if you would like another set of tunable knobs for the TPSCut, rather than just reusing the KnockCEL ones..
So basically:
Load threshold = minimum load for knockcel + knocktpscut to work
Low Knocksum threshold = minimum knocksum for SLOW flashing
High Knocksum threshold = minimum knocksum for FAST flashing AND TPSCut
There is a TPSCut attirbute, which is a percentage applied to the current TPS %
ie, if TPSCut detects a High KnockSum event then it will multiply the current TPS % by this
therefor to disable TPSCut you could set this to:
* 100% - ie leave the current TPS alone
* 50% - ie use 50% of the current TPS signal
* 25% - ie use 25% of the current TPS signal
* anywhere in between
I will be testing this tonight/tommorow and then I will do up a 52680022 ROM for the testers
ok
just a bit of an update.
I've revised the knockcel code.
Previously as soon as the load and knocksum threshold was reached KnockCEL started flashing (slow or fast depending on which KnockSum threshold was crossed)
Now say your load threshold was set to 150 and knocksum threshold at 4.
This meant that if the knocksum was at 10 at load 100, and was on it's way down when you passed 150load then it would flash... Which is not very accurate...
So what I have done is only enable the flashing when the knocksum this loop through is GREATER than last time... this means it wont flash in the above scenario.
I have also done the TPSCut based on High KnockSum - let me know if you would like another set of tunable knobs for the TPSCut, rather than just reusing the KnockCEL ones..
So basically:
Load threshold = minimum load for knockcel + knocktpscut to work
Low Knocksum threshold = minimum knocksum for SLOW flashing
High Knocksum threshold = minimum knocksum for FAST flashing AND TPSCut
There is a TPSCut attirbute, which is a percentage applied to the current TPS %
ie, if TPSCut detects a High KnockSum event then it will multiply the current TPS % by this
therefor to disable TPSCut you could set this to:
* 100% - ie leave the current TPS alone
* 50% - ie use 50% of the current TPS signal
* 25% - ie use 25% of the current TPS signal
* anywhere in between
I will be testing this tonight/tommorow and then I will do up a 52680022 ROM for the testers
just a bit of an update.
I've revised the knockcel code.
Previously as soon as the load and knocksum threshold was reached KnockCEL started flashing (slow or fast depending on which KnockSum threshold was crossed)
Now say your load threshold was set to 150 and knocksum threshold at 4.
This meant that if the knocksum was at 10 at load 100, and was on it's way down when you passed 150load then it would flash... Which is not very accurate...
So what I have done is only enable the flashing when the knocksum this loop through is GREATER than last time... this means it wont flash in the above scenario.
I have also done the TPSCut based on High KnockSum - let me know if you would like another set of tunable knobs for the TPSCut, rather than just reusing the KnockCEL ones..
So basically:
Load threshold = minimum load for knockcel + knocktpscut to work
Low Knocksum threshold = minimum knocksum for SLOW flashing
High Knocksum threshold = minimum knocksum for FAST flashing AND TPSCut
There is a TPSCut attirbute, which is a percentage applied to the current TPS %
ie, if TPSCut detects a High KnockSum event then it will multiply the current TPS % by this
therefor to disable TPSCut you could set this to:
* 100% - ie leave the current TPS alone
* 50% - ie use 50% of the current TPS signal
* 25% - ie use 25% of the current TPS signal
* anywhere in between
I will be testing this tonight/tommorow and then I will do up a 52680022 ROM for the testers
What code does the 52680022 inherit? 52680015?
Had to create a XML for 52680022. I suggest you open the 52680020 xml. Save as 52680022.XML under TYPE: ALL FILES. Then open that XML up and change the 52680020 in that file to 52680022. Then save the XML file. Restart ECUFLASH.
Last edited by Clipse3GT; May 13, 2010 at 12:35 AM.
I'm getting all tingly 
Off topic warning....
Tephra, out of curiosity, I was told by an Australian that Foster's beer is considered one of the worst beer's in Australia, is that correct?
They sure do allot of advertising here to make us over here think there top notch.

Off topic warning....
Tephra, out of curiosity, I was told by an Australian that Foster's beer is considered one of the worst beer's in Australia, is that correct?
They sure do allot of advertising here to make us over here think there top notch.
Thread Starter
Joined: Feb 2007
Posts: 9,486
Likes: 67
From: Melbourne, Australia
just loaded my latest build on my car and it didn't brick my ECU
yay :P
re fosters:
put it this way, I can't think of where to buy it... ive never seen it in a bottle-shop/bar or pub
yay :Pre fosters:
put it this way, I can't think of where to buy it... ive never seen it in a bottle-shop/bar or pub
Just a suggestion before this gets released... do you think it would be better to call it 52680120 like we did with the CT9A stuff? The only problem I think we might run into is that (unlikely as it is) Mitsu may release another update that would then share the same ID.
Yay!
Just a suggestion before this gets released... do you think it would be better to call it 52680120 like we did with the CT9A stuff? The only problem I think we might run into is that (unlikely as it is) Mitsu may release another update that would then share the same ID.
Just a suggestion before this gets released... do you think it would be better to call it 52680120 like we did with the CT9A stuff? The only problem I think we might run into is that (unlikely as it is) Mitsu may release another update that would then share the same ID.
The new KnockCEL setup sounds good tephra. Looking forward to it. Also, was that sufficient data for FuelTemp? I can get more if you'd like more samples.



