What do we know about knock control?
Jun 4, 2008, 01:00 PM
#16
Account Disabled
iTrader: (3)
Join Date: Apr 2006
Location: USA
Posts: 1,029
Likes: 0
Received 0 Likes
on
0 Posts
I was just saying to watch knock voltage. Its funny and the voltage actually gets lower as the car enters boost. Its also interesting to watch the raw knock sums in evoscan. Again its just weird stuff.
If we can't find out HOW the stock ecu does it I don't think that mimicking how AEM does it as being a bad thing. I know its not ideal.
If we can't find out HOW the stock ecu does it I don't think that mimicking how AEM does it as being a bad thing. I know its not ideal.
Jun 4, 2008, 04:39 PM
#17
Evolved Member
Join Date: Apr 2008
Location: Sydney, Australia
Posts: 604
Likes: 0
Received 0 Likes
on
0 Posts
Can someone point me out as to where I can find the info to disable the knock control? My car is driving me nuts with "phantom" knock and its to a point where the car can become almost undrivable.
Now before I get bombarded with people telling me to check for exhaust rattles and loose brackets, I am a very experienced tuner and would normally (if it were another type of vehicle) have ditched the ECU for an aftermarket unit. It is not a rattle, it is the ECU thinking it is knocking.
I have 3 wide-band units at my shop, with the best an expensive NTK setup, along with an Autronic and an Innovate LC1 permanently in the car. Its not fuel. Its not timing either, as the knock control setup will pull timing back until it caps at 10ATDC (-10*) under conditions where it should be running near 20BTDC. You can watch the knock control pulling out huge amounts of timing when it does it, mostly at part throttle low-boost areas. Go WOT and no problems at all. Give it a 20km drive on the freeway and by the end of it the octane number is back to zero!
Now is the part most people would be more interested to know. I spent a full day yesterday playing with all of the so called knock filter maps. All of them! The result was...... NOTHING. No change at all under any conditions. I tried making small changes, I tried maxxing all of the maps out one by one, tried zeroing them, tried combinations and could not get it to change the way it behaves in even the smallest part. I was looking for more sensitivity under conditions where there was no previous knock retardation (eg. WOT) and still no change. So I could not force it to think it was knocking.
I think most people on here dont understand detonation. Most blindly look at at the knock sum values and take it as gospel that if it says something it must be knocking. Not always true my friends. I trust my tuning, and I am more than willing to run around without knock control. 99% of aftermarket ECU's out there in competition and street use dont use knock control and guess what, they dont blow motors (when tuned correctly I should add :-D)
Next question. Has anyone figured out at what octane value the car is 100% on the low octane map? I did some testing and it 100% uses the high octane map on an octane value of 100. When my car gets an octane value of zero, it does not use any of the ignition maps I have. I have setup my low octane map a few degrees lower in timing, but not 30 degrees lower. So there must be a point at which it crosses over 100% from high to low, and there also must be stored values of retard applied over a 3D map. The only way for me to return to normal is to reset the ECU.
So for now, I am plugging in a spare knock sensor and leaving it strapped to the side. Screw knock control! Unless we have full control over its conditions, playing in the dark is just a waste of time. And I tell you what, next time you check I will bet my motor is still running fine and didnt grenade the first time I drove it.
And of rant.
Now before I get bombarded with people telling me to check for exhaust rattles and loose brackets, I am a very experienced tuner and would normally (if it were another type of vehicle) have ditched the ECU for an aftermarket unit. It is not a rattle, it is the ECU thinking it is knocking.
I have 3 wide-band units at my shop, with the best an expensive NTK setup, along with an Autronic and an Innovate LC1 permanently in the car. Its not fuel. Its not timing either, as the knock control setup will pull timing back until it caps at 10ATDC (-10*) under conditions where it should be running near 20BTDC. You can watch the knock control pulling out huge amounts of timing when it does it, mostly at part throttle low-boost areas. Go WOT and no problems at all. Give it a 20km drive on the freeway and by the end of it the octane number is back to zero!
Now is the part most people would be more interested to know. I spent a full day yesterday playing with all of the so called knock filter maps. All of them! The result was...... NOTHING. No change at all under any conditions. I tried making small changes, I tried maxxing all of the maps out one by one, tried zeroing them, tried combinations and could not get it to change the way it behaves in even the smallest part. I was looking for more sensitivity under conditions where there was no previous knock retardation (eg. WOT) and still no change. So I could not force it to think it was knocking.
I think most people on here dont understand detonation. Most blindly look at at the knock sum values and take it as gospel that if it says something it must be knocking. Not always true my friends. I trust my tuning, and I am more than willing to run around without knock control. 99% of aftermarket ECU's out there in competition and street use dont use knock control and guess what, they dont blow motors (when tuned correctly I should add :-D)
Next question. Has anyone figured out at what octane value the car is 100% on the low octane map? I did some testing and it 100% uses the high octane map on an octane value of 100. When my car gets an octane value of zero, it does not use any of the ignition maps I have. I have setup my low octane map a few degrees lower in timing, but not 30 degrees lower. So there must be a point at which it crosses over 100% from high to low, and there also must be stored values of retard applied over a 3D map. The only way for me to return to normal is to reset the ECU.
So for now, I am plugging in a spare knock sensor and leaving it strapped to the side. Screw knock control! Unless we have full control over its conditions, playing in the dark is just a waste of time. And I tell you what, next time you check I will bet my motor is still running fine and didnt grenade the first time I drove it.
And of rant.
Jun 4, 2008, 05:01 PM
#18
Account Disabled
iTrader: (3)
Join Date: Apr 2006
Location: USA
Posts: 1,029
Likes: 0
Received 0 Likes
on
0 Posts
Thanks for the rant, this is not the place for it. Make another thread.
Last year I had a clogged fuel sock. I too thought I had "phantom knock" due to my built motor that I had. I say had because I disabled the knock sensor then this happened. http://s298.photobucket.com/albums/m...incylinder.jpg
Fixed the clogged fuel sock, threw a stock bottom end back in it, and it runs fine WITH the stock knock control. But you know more than me so go ahead..........
Last year I had a clogged fuel sock. I too thought I had "phantom knock" due to my built motor that I had. I say had because I disabled the knock sensor then this happened. http://s298.photobucket.com/albums/m...incylinder.jpg
Fixed the clogged fuel sock, threw a stock bottom end back in it, and it runs fine WITH the stock knock control. But you know more than me so go ahead..........
Last edited by dan l; Jun 4, 2008 at 05:25 PM.
Jun 4, 2008, 05:53 PM
#19
Next question. Has anyone figured out at what octane value the car is 100% on the low octane map? I did some testing and it 100% uses the high octane map on an octane value of 100. When my car gets an octane value of zero, it does not use any of the ignition maps I have. I have setup my low octane map a few degrees lower in timing, but not 30 degrees lower. So there must be a point at which it crosses over 100% from high to low, and there also must be stored values of retard applied over a 3D map. The only way for me to return to normal is to reset the ECU.
You are getting timing numbers lower than the octane map because the ECU's knock control is more than just the low and high octane maps. The low and high octane maps are more of a longer term knock trim, so to speak.
The immediate knock control pulls about 1* of timing per 3 knock counts. So, if you are getting 36 counts of knock, you are probably running 12* lower than the timing maps call. If your octane number is 0, then you should be 12* lower than the low octane map calls for.
I'm not sure if the ECU has additional retard if you are at an octane number of 0 and a knocksum of 36. Perhaps it will pull even more.
Eric
Last edited by l2r99gst; Jun 4, 2008 at 05:55 PM.
Jun 5, 2008, 12:30 AM
#20
If you truly have to have it, set the periphery 0 value (addy FAA) to 0x5658 and then tune using only the low octane maps (it kicks it over to those as a result).
The 3.45v was repeated noise, not just the odd harmonic. Having tuned AEM cars and having to watch what might be knock and being educated on that (as well as the old standby of plug reading) I know what you mean Dex, but I am trying to show that high voltage can be fine and low voltage (1.1v) can be 10 counts sometimes. There is little rhyme or reason to it it seems.
The 3.45v was repeated noise, not just the odd harmonic. Having tuned AEM cars and having to watch what might be knock and being educated on that (as well as the old standby of plug reading) I know what you mean Dex, but I am trying to show that high voltage can be fine and low voltage (1.1v) can be 10 counts sometimes. There is little rhyme or reason to it it seems.
Jun 5, 2008, 07:13 AM
#21
Evolved Member
Join Date: Apr 2008
Location: Sydney, Australia
Posts: 604
Likes: 0
Received 0 Likes
on
0 Posts
dan l , obviously I do know more than you if you needed a shovel to pick up your motor.
My post was about the effects of the supposed knock filter maps which is directly about the thread topic, but it seems you dont want to hear about any experimentation with them. My motor is not running lean and everything is in top stock standard condition. Put any timing values in the map and it still throws the same huge knock sums and knock retardation.
I dont think zero octane is the low octane map. It could be, but at the moment I am not sure about it. I noticed that unplugging the knock sensor (throwing a CEL) causes the ecu to report zero octane and use a massively retarded timing map, none of which lines up with the programmed maps. I believe there is a backgound map that says how much max timing can be pulled out in any given area of the map and when it goes limp it pulls it all out. From what I can gather from my experimentation, one of the filter maps (number varies from model to model) is a map telling what areas of load vs rpm are checked for knock, and in these areas applies it to the octane number. Knock in that area drops the octance number over time, no knock in that area raises it back up and rather quickly I might add. Managed to raise mine back from low 80's to 100 in less than 5 mins of driving. It also doesnt touch the off-boost areas of the map. These appear to remain as they are programmed in the low octane map.
I also tend to think that the knock sum is a given number of knock voltage spikes over a given number of engine cycles, and not directly related to the actual knock voltage. I have seen the same effect today with testing, noticing it will throw some high knock sum values even at low knock voltages, and sometimes high voltages show low knock sum. Obviously the word sum means it is adding things up, over which can only be time.
Part of todays testing was to plug in another knock sensor and leave it hanging so it does not pick up anything at all from the motor, and monitor intensely for knocking without any ecu influence of retarding timing. Got less than 10 seconds up the road and it throws a CEL for the sensor. Damn ECU seems smart enough to realise that no knock voltage at all is a fault even with the sensor plugged in! I tied it to a bracket off the rocker cover so it picks up some background noise and no more CEL..... and what do you know.... no more stupid knock retarding and a perfect running motor. 100% my problem is phantom and not real knock.
Next test is to plug back in the sensor in the block and bring the octane number down. I usually set a large area of light throttle parts of the ignition maps to a set number so I can see when it is using that map. 100 octane definitely uses 100% of the high octane map, so I will see how low the number needs to get to use 100% of the low octane map. I might also try varying the knock filter map I was talking about earlier and try to get the system to drop octane on free-rev rather than needing load to do it. If this works then at least we might learn what one of the maps truly does.
My post was about the effects of the supposed knock filter maps which is directly about the thread topic, but it seems you dont want to hear about any experimentation with them. My motor is not running lean and everything is in top stock standard condition. Put any timing values in the map and it still throws the same huge knock sums and knock retardation.
I dont think zero octane is the low octane map. It could be, but at the moment I am not sure about it. I noticed that unplugging the knock sensor (throwing a CEL) causes the ecu to report zero octane and use a massively retarded timing map, none of which lines up with the programmed maps. I believe there is a backgound map that says how much max timing can be pulled out in any given area of the map and when it goes limp it pulls it all out. From what I can gather from my experimentation, one of the filter maps (number varies from model to model) is a map telling what areas of load vs rpm are checked for knock, and in these areas applies it to the octane number. Knock in that area drops the octance number over time, no knock in that area raises it back up and rather quickly I might add. Managed to raise mine back from low 80's to 100 in less than 5 mins of driving. It also doesnt touch the off-boost areas of the map. These appear to remain as they are programmed in the low octane map.
I also tend to think that the knock sum is a given number of knock voltage spikes over a given number of engine cycles, and not directly related to the actual knock voltage. I have seen the same effect today with testing, noticing it will throw some high knock sum values even at low knock voltages, and sometimes high voltages show low knock sum. Obviously the word sum means it is adding things up, over which can only be time.
Part of todays testing was to plug in another knock sensor and leave it hanging so it does not pick up anything at all from the motor, and monitor intensely for knocking without any ecu influence of retarding timing. Got less than 10 seconds up the road and it throws a CEL for the sensor. Damn ECU seems smart enough to realise that no knock voltage at all is a fault even with the sensor plugged in! I tied it to a bracket off the rocker cover so it picks up some background noise and no more CEL..... and what do you know.... no more stupid knock retarding and a perfect running motor. 100% my problem is phantom and not real knock.
Next test is to plug back in the sensor in the block and bring the octane number down. I usually set a large area of light throttle parts of the ignition maps to a set number so I can see when it is using that map. 100 octane definitely uses 100% of the high octane map, so I will see how low the number needs to get to use 100% of the low octane map. I might also try varying the knock filter map I was talking about earlier and try to get the system to drop octane on free-rev rather than needing load to do it. If this works then at least we might learn what one of the maps truly does.
Jun 5, 2008, 07:28 AM
#22
Did anything of what I wrote make any sense?
I see you keep saying that you don't think 0 octane means 100% of the low octane map, but I think this has already been found to be true via disassembly.
There are other safeguards and algorithms which will still pull timing below the low octane map. Again, think of the low octane map more of a long term knock trim. Timing still gets pulled further dependent on knock counts. The octane number begins decreasing for a knock count value of 6 or more to try to alleviate the knock in the long term. The immediate term is pulling roughly 1* per 3 counts.
As far as unplugging the sensor, that is a totally different scenario. I'm sure the ECU has a limp mode in that case, so that you don't destroy the motor.
I see you keep saying that you don't think 0 octane means 100% of the low octane map, but I think this has already been found to be true via disassembly.
There are other safeguards and algorithms which will still pull timing below the low octane map. Again, think of the low octane map more of a long term knock trim. Timing still gets pulled further dependent on knock counts. The octane number begins decreasing for a knock count value of 6 or more to try to alleviate the knock in the long term. The immediate term is pulling roughly 1* per 3 counts.
As far as unplugging the sensor, that is a totally different scenario. I'm sure the ECU has a limp mode in that case, so that you don't destroy the motor.
Jun 5, 2008, 07:32 AM
#23
All of our stock ECU engine builds retain the balance shafts for this reason.
Why do you think we have not gone 10's since 2006?
Once the shafts were out, it was game over.
Why do you think we have not gone 10's since 2006?
Once the shafts were out, it was game over.
After doing a bunch of research, messing with forged motors and cars without balance shafts, I still dont have any real clear idea about what the "knock" maps we can see in ECUflash.
Balance shaft deletes can cause odd harmonics from 4,000 and down it seems. I saw 15 counts at 72%load and 33* timing cruising today.
Forged motors can (dont always) throw 8-10 counts in when accelerating and randomly through the rev range.
I am curious where we are at in the disassembly of this pretty sophisticated control system? I mean does it look probable that we will be able to have knock control like in DSMlink eventually or is it significantly more sophisticated than that?
Knock control parameters that would be cool:
Control of degrees per count
RPM range (like the subaru has, rpm correction ranges both low and high)
Load based control like the subaru as well?
Ignore knock voltage below a certain level
Ideas? Input, criticism, praise, etc.?
Balance shaft deletes can cause odd harmonics from 4,000 and down it seems. I saw 15 counts at 72%load and 33* timing cruising today.
Forged motors can (dont always) throw 8-10 counts in when accelerating and randomly through the rev range.
I am curious where we are at in the disassembly of this pretty sophisticated control system? I mean does it look probable that we will be able to have knock control like in DSMlink eventually or is it significantly more sophisticated than that?
Knock control parameters that would be cool:
Control of degrees per count
RPM range (like the subaru has, rpm correction ranges both low and high)
Load based control like the subaru as well?
Ignore knock voltage below a certain level
Ideas? Input, criticism, praise, etc.?
Jun 5, 2008, 07:36 AM
#24
dan l , obviously I do know more than you if you needed a shovel to pick up your motor.
My post was about the effects of the supposed knock filter maps which is directly about the thread topic, but it seems you dont want to hear about any experimentation with them. My motor is not running lean and everything is in top stock standard condition. Put any timing values in the map and it still throws the same huge knock sums and knock retardation.
I dont think zero octane is the low octane map. It could be, but at the moment I am not sure about it. I noticed that unplugging the knock sensor (throwing a CEL) causes the ecu to report zero octane and use a massively retarded timing map, none of which lines up with the programmed maps. I believe there is a backgound map that says how much max timing can be pulled out in any given area of the map and when it goes limp it pulls it all out. From what I can gather from my experimentation, one of the filter maps (number varies from model to model) is a map telling what areas of load vs rpm are checked for knock, and in these areas applies it to the octane number. Knock in that area drops the octance number over time, no knock in that area raises it back up and rather quickly I might add. Managed to raise mine back from low 80's to 100 in less than 5 mins of driving. It also doesnt touch the off-boost areas of the map. These appear to remain as they are programmed in the low octane map.
I also tend to think that the knock sum is a given number of knock voltage spikes over a given number of engine cycles, and not directly related to the actual knock voltage. I have seen the same effect today with testing, noticing it will throw some high knock sum values even at low knock voltages, and sometimes high voltages show low knock sum. Obviously the word sum means it is adding things up, over which can only be time.
Part of todays testing was to plug in another knock sensor and leave it hanging so it does not pick up anything at all from the motor, and monitor intensely for knocking without any ecu influence of retarding timing. Got less than 10 seconds up the road and it throws a CEL for the sensor. Damn ECU seems smart enough to realise that no knock voltage at all is a fault even with the sensor plugged in! I tied it to a bracket off the rocker cover so it picks up some background noise and no more CEL..... and what do you know.... no more stupid knock retarding and a perfect running motor. 100% my problem is phantom and not real knock.
Next test is to plug back in the sensor in the block and bring the octane number down. I usually set a large area of light throttle parts of the ignition maps to a set number so I can see when it is using that map. 100 octane definitely uses 100% of the high octane map, so I will see how low the number needs to get to use 100% of the low octane map. I might also try varying the knock filter map I was talking about earlier and try to get the system to drop octane on free-rev rather than needing load to do it. If this works then at least we might learn what one of the maps truly does.
My post was about the effects of the supposed knock filter maps which is directly about the thread topic, but it seems you dont want to hear about any experimentation with them. My motor is not running lean and everything is in top stock standard condition. Put any timing values in the map and it still throws the same huge knock sums and knock retardation.
I dont think zero octane is the low octane map. It could be, but at the moment I am not sure about it. I noticed that unplugging the knock sensor (throwing a CEL) causes the ecu to report zero octane and use a massively retarded timing map, none of which lines up with the programmed maps. I believe there is a backgound map that says how much max timing can be pulled out in any given area of the map and when it goes limp it pulls it all out. From what I can gather from my experimentation, one of the filter maps (number varies from model to model) is a map telling what areas of load vs rpm are checked for knock, and in these areas applies it to the octane number. Knock in that area drops the octance number over time, no knock in that area raises it back up and rather quickly I might add. Managed to raise mine back from low 80's to 100 in less than 5 mins of driving. It also doesnt touch the off-boost areas of the map. These appear to remain as they are programmed in the low octane map.
I also tend to think that the knock sum is a given number of knock voltage spikes over a given number of engine cycles, and not directly related to the actual knock voltage. I have seen the same effect today with testing, noticing it will throw some high knock sum values even at low knock voltages, and sometimes high voltages show low knock sum. Obviously the word sum means it is adding things up, over which can only be time.
Part of todays testing was to plug in another knock sensor and leave it hanging so it does not pick up anything at all from the motor, and monitor intensely for knocking without any ecu influence of retarding timing. Got less than 10 seconds up the road and it throws a CEL for the sensor. Damn ECU seems smart enough to realise that no knock voltage at all is a fault even with the sensor plugged in! I tied it to a bracket off the rocker cover so it picks up some background noise and no more CEL..... and what do you know.... no more stupid knock retarding and a perfect running motor. 100% my problem is phantom and not real knock.
Next test is to plug back in the sensor in the block and bring the octane number down. I usually set a large area of light throttle parts of the ignition maps to a set number so I can see when it is using that map. 100 octane definitely uses 100% of the high octane map, so I will see how low the number needs to get to use 100% of the low octane map. I might also try varying the knock filter map I was talking about earlier and try to get the system to drop octane on free-rev rather than needing load to do it. If this works then at least we might learn what one of the maps truly does.
The sensor is designed to be GROUNDED at ALL TIMES. Whether its plugged into the block or not is up to you.
Jun 5, 2008, 09:02 AM
#25
Lucas' engine doesnt have balance shafts and is fine. I dont know if the aluminum rods have anything to do with it (dont see why they would) but it versus some stock bottom end cars I have seen seem to indicate something about his is helping.
Lucas-0 counts and 24* out the top
Other cars - 15 part throttle or less, and none above 4500
Not sure what noise it is picking up lower if its the balance shafts or not. I am still wondering though about the actual disassembly since the last time we talked about it was deemed "really complex". That is understandable but I am wondering if since the periphery hook is already known if anyone had seen something hopeful or not?
Lucas-0 counts and 24* out the top
Other cars - 15 part throttle or less, and none above 4500
Not sure what noise it is picking up lower if its the balance shafts or not. I am still wondering though about the actual disassembly since the last time we talked about it was deemed "really complex". That is understandable but I am wondering if since the periphery hook is already known if anyone had seen something hopeful or not?
Jun 5, 2008, 10:57 AM
#27
Jun 5, 2008, 04:27 PM
#28
Evolved Member
Join Date: Apr 2008
Location: Sydney, Australia
Posts: 604
Likes: 0
Received 0 Likes
on
0 Posts
All made sense, but still it is good to prove it in the real world. It is why I said "I dont think zero octane is the low octane map. It could be, but at the moment I am not sure about it." But if the code says it is then it is. I also am full aware of there being various corrections applied to the maps. Definitely there is coolant temp with ignition and air temp with ignition, along with many others probably undiscovered yet. The worst part is that I am not an assembly programmer, I am a tuner. The other thing is that I am working with a GT-A Automatic, and virtually none of the development work you guys have done in the states is applied to these vehicles (Def 80700010), so I am flying blind here.
I also didnt just unplug the sensor for tuning. I used a second sensor to plug into that wasnt mounted in the block. Having it earthed makes sense and was something that went through my mind at the time, which is why it was grounded to the rocker cover. So yeah, maybe the ecu is not that smart, just needed the ground. :-) Thanks for the confirmation TTP. Part of going limp in early testing was to find if one of the maps directly controls which parts of the ignition map and by how much is adjusted.
Trying to wrap my head around what goes on in the ECU is the hardest part when it comes to tuning with them. Trying to find this relationship between the high and low octane maps is important for an accurate and safe tune. The problem is that to get the octane number near to zero, there is already retard values stored in RAM that make the low octane map almost useless for tuning. By the time I have the octane below around 70 it no longer follows any of the timing maps.
Since the disassembly guys are still working on how to discover what everything does, it doesnt hurt to do some real world testing and try to work from the other direction with discovering what is going on. I am not trying to run permanently without knock control, I am trying to find a way to adjust it and use it to our advantage. But why do I get the impression from you guys that you dont want to hear about the experimentation?
I also didnt just unplug the sensor for tuning. I used a second sensor to plug into that wasnt mounted in the block. Having it earthed makes sense and was something that went through my mind at the time, which is why it was grounded to the rocker cover. So yeah, maybe the ecu is not that smart, just needed the ground. :-) Thanks for the confirmation TTP. Part of going limp in early testing was to find if one of the maps directly controls which parts of the ignition map and by how much is adjusted.
Trying to wrap my head around what goes on in the ECU is the hardest part when it comes to tuning with them. Trying to find this relationship between the high and low octane maps is important for an accurate and safe tune. The problem is that to get the octane number near to zero, there is already retard values stored in RAM that make the low octane map almost useless for tuning. By the time I have the octane below around 70 it no longer follows any of the timing maps.
Since the disassembly guys are still working on how to discover what everything does, it doesnt hurt to do some real world testing and try to work from the other direction with discovering what is going on. I am not trying to run permanently without knock control, I am trying to find a way to adjust it and use it to our advantage. But why do I get the impression from you guys that you dont want to hear about the experimentation?
Jun 5, 2008, 04:50 PM
#29
Since the disassembly guys are still working on how to discover what everything does, it doesnt hurt to do some real world testing and try to work from the other direction with discovering what is going on. I am not trying to run permanently without knock control, I am trying to find a way to adjust it and use it to our advantage. But why do I get the impression from you guys that you dont want to hear about the experimentation?
Jun 6, 2008, 03:41 AM
#30
What we do know for sure is that the low octane and high octane map are linearly interpolated from the octane number. The octane number will not stay on zero if the knock sum is <=3, so you will never see the timing lining up as closely as it would on the high octane map with an octane number of 100 and knock sum of zero. Practical observations are always useful as sometimes other bits of code do funny things, but I did play with manually setting the octane value to 0 and it behaved as expected from the code. I think it would change the boost as well, but I wasn't using ECU controlled boost at the time to test.
"are the branches to subroutines part of a larger algorithm?"
We can see where the code is, but yes the calculations are made up of lots of branches that check lots of bitmapped flags. Following through the code to write an equation or algorithm to summarise results in me rapidly getting lost.
With sufficient time and motivation I daresay it would be better to actually rewrite the whole knock control routine. If approaching this I would keep the knock control working on a car where it functions well and try to develop my own loggable variable alongside it which would process knock ADC with filters and then use a threshold or difference system.
I've proven the quality of knock control on my own Evo with det cans. I now trust the knock sum on my own and most other standardish cars. For a pump fuel high boost Green setup I chose to use the OEM ECU over a MoTeC M800 because of the quality of the knock control, which thankfully works despite carbon clutch, big wheels, aftermarket manifold, cams. To hold a pump fuel high boost setup together but still get performance I really wanted the knock control. The variations in atmospheric conditions, engine temperatures and fuel would mean either many sessions mapping it in all these conditions, a conservative map or race fuel, or in this case OEM knock control.
On a previous car I had a forged engine which had excessive clearance. The builder had another three goes at it before giving me my money back. One of the nightmares with it was that I couldn't hear the detonation above the piston noise even with det cans. I've not fiddled with enough quality forged engines since and have been unfortunately put off them for life if I can avoid them.
"are the branches to subroutines part of a larger algorithm?"
We can see where the code is, but yes the calculations are made up of lots of branches that check lots of bitmapped flags. Following through the code to write an equation or algorithm to summarise results in me rapidly getting lost.
With sufficient time and motivation I daresay it would be better to actually rewrite the whole knock control routine. If approaching this I would keep the knock control working on a car where it functions well and try to develop my own loggable variable alongside it which would process knock ADC with filters and then use a threshold or difference system.
I've proven the quality of knock control on my own Evo with det cans. I now trust the knock sum on my own and most other standardish cars. For a pump fuel high boost Green setup I chose to use the OEM ECU over a MoTeC M800 because of the quality of the knock control, which thankfully works despite carbon clutch, big wheels, aftermarket manifold, cams. To hold a pump fuel high boost setup together but still get performance I really wanted the knock control. The variations in atmospheric conditions, engine temperatures and fuel would mean either many sessions mapping it in all these conditions, a conservative map or race fuel, or in this case OEM knock control.
On a previous car I had a forged engine which had excessive clearance. The builder had another three goes at it before giving me my money back. One of the nightmares with it was that I couldn't hear the detonation above the piston noise even with det cans. I've not fiddled with enough quality forged engines since and have been unfortunately put off them for life if I can avoid them.