[Sniz]

Originally Posted by KennyPowers

and that ^^^^^ is what everyone has been posting for 35+ pages.

And the JBx tunes as of yet have no way to limit the occurance of these knock sensor related timing drops. BMW, procede, cobb, giac do by either lowering the max set ignition limits, CPS offseting. The jb does none of the above.

So by virtue of your own statements, the jb in it's current form is deficient because it is unable to limit these events as other tunes are able to do.

[Terrance28]

Originally Posted by InCityPhoto

And what is your definition of knock? Full detonation?

The knock sensors react to anything that resembles knock frequencies even in the slightest, so would it not make sense that they are reacting as the knock frequencies starts?




If you are seriously thinking that a aftermarket tuner has done more research then a company like BMW...

Then you must be crazier than a bed bug. Sorry I had to finish it.

[Bagelfast]

Originally Posted by AlterZgo

My 2 cents... As people have already pointed out, there are many here who just refuse to learn and understand basic tuning principles.

I think the problem is, many of the younger posters on this forum grew up reading about overclocking and hacking computers and cell phones and iPods and whatnot. Difference is, you overclock a computer and until it starts freezing and all you have to do is back off the speed a bit and no harm no foul. You jailbreak your iPhone and if it gets a screwed up, you can always flash back to stock ROM and you're OK.

However, if you run a car - ANY car - with excessive amts of boost, timing, excessive lean A/F ratios, you can instantly grenade your motor to the tune of a $10K+ repair. This isn't something you can recover from by just turning the boost down to stock or running a less aggressive map. Once your motor is fried, that's it.

So, with this amt of risk involved, why would anyone still prefer to use a tune that solely relies on the stock engine's computer to control timing and boost at over 2x stock boost levels when they can purchase a tune that provides the proper timing and boost controls for just a small amount more?

Those who believe BMW engineers know everything about N54 motors and provide the perfect strategies for managing detonation in the stock DME, do you really believe BMW engineered these motors to run 16+ psi for sustained periods of time with the stock maps? Hell no! Look at how conservative they are with the 335is. They add extra oil coolers and they only allow overboost under pre-set conditions for 7 seconds and if it's too warm, it won't allow overboost and the timing is definitely reduced compared to a stock N54. All that for just 7 seconds of 14 psi overboost.

So, if you really think BMW engineers are so sharp (and they are), why the fuck would you want to run 16+ PSI sustained with the stock timing maps and rely solely on the stock DME to pull timing?

Just playing devils advocate, can you explain, taken my last post in this thread into consideration aswell, why i ran Map7, an overly agressive map on pump 93 octane for 3 years on a otherwise stock car, and nothing happened to my motor, and its still runing on the same tune till this day, healthy and strong. Just interesting, since we have people giving explanation, but there explanantions seem to give an idea that the motors should explode or break, which hasnt been the case.

[JB4-335]

Originally Posted by Clap135

Post the logs showing ambiant then. What front mount do you have and how long was the actual log?

Here is my JB4 the log done in high 80's with stock front mount.

[ZTUNER]

I tried to keep up with this thread all day at work today and a few things bothered me. The first one is about people talking about what the knock sensor can and cannot do.

The knock sensor itself does not know or care about "knock" per say. Infact it knows absolutely nothing. It is a simple , crude microphone with max sensitivity at a particular frequency band. All it produces is "noise" which is filtered by the DME based on frequency, amplitude and crank angle. A lot of the noise is discarded based on crank angle. If there are multiple knock sensors they can be used to triangulate the noise to figure out which cylinder the knock is coming from. There is a table in all ecus which describes the intensity of this noise (usually voltage) vs rpm. This would be the most basic setup.

For example take this theoretical table rpm vs knock sensor voltage. (not real just an example)
rpm knock voltage
1000 1
2000 2
3000 3
4000 3.5
5000 4
6000 5

If say at 5000 rpm the voltage measured by the knock sensor is greater than 4 volts it refers to another table that looks at "how much more"
If its .1 volt more it pulls 1 deg
If it .5 volt more it pulls 3 deg etc etc
If its off the charts it throws a code and drops you into fail safe

This can be further refined by running this against a "duration" table which controls the "decay" of ignition retard. For instance, if the knock voltage is hitting .5 volts higher than 4 volts at 5000 rpm 5 times a sec, it might pull timing 6 deg vs 3 deg for 2 times a sec and it might keep the timing pulled down longer.

All this can be further affected by other variables like boost level, AIT, Ambient temps or water temps or oil temps, EGT's or whatever predictive element BMW deems appropriate. Apparently Cobb logs show a MAF output which imples that MASS of air is somehow being measured or calculated even though this a MAP based car. I did not see a MAF sensor on here. IMO the decision tree controlling the sensitivity of the knock sensor output could be highly complicated in this car and i am quite sure that unless we know and understand fully how the DME manipulates the knock sensor data we will never know why timing is pulled. Obviously we can see patterns and try to guess but it does not mean our guesses are accurate.

All this being said with higher boost comes , higher AIT's and higher EGT's and maybe higher water and oil temps. Maybe there is a table or two or three that looks at these variables and drops timing if the rate of increase of these variables excedes a certain threshold ? Who knows ? I am certain the piggy tuners and even the flash tuners do not know ALL these variables. Maybe there is a predictive table in the DME that uses a bunch of variables that that are used to figure out mathematically what the cylinder pressures are and it pulls timing based on THAT and not knock itself ? You WILL control Cylinder pressures if you advance or retard timing. Thats why you don't see a huge change in a/f between gears but there is a significant in timing.

ONLY way to truly tell if you have knock is to hook up a high quality knock sensor and hear it thru headphones and tune accordingly. You can also log the knock count PID if that is indeed loggable with this DME to see what the ecu is seeing. You cannot log timing alone and look at drops in timing and assume it is knock. It is a guess at best.

Just wanted to add, another way to figure out if its knock is to dump in high oct fuel to see if the timing drops go away (not meth because it affects ait) . If timing goes up with high oct then its obvious its not a compensation map of some kind but actual knock.

All this being said we should quit attacking vendors and try to understand how this s@#$ works. Its facinating at the very least

Harry

PS : i didn't get into the other things because the first thing was long enough

[Joshboody]

Originally Posted by JB4135

Here is my JB4 the log done in high 80's with stock front mount.

that's a pretty good log, so keep logging every once in awhile and make sure it stays smooth. i'm guessing 93oct and any bolt ons?

without some type of timing control, I would NOT have logs like that in AZ in the summer at 700ft above seal level. But i have timing control and very smooth tuned timing curves.

[JB4-335]

Originally Posted by Joshboody

that's a pretty good log, so keep logging every once in awhile and make sure it stays smooth. i'm guessing 93oct and any bolt ons?

without some type of timing control, I would NOT have logs like that in AZ in the summer at 700ft above seal level. But i have timing control and very smooth tuned timing curves.

I'm in Australia and it's summer here. 300ft above sea level

I'm running 98RON which is the same as your 93MON in the US.

Just DCI at the moment. 3" C@tless Downpipes to install next week.

[Joshboody]

i remember 4 knock sensors on real oem i think. 2 on each side of the block.

[AlterZgo]

Originally Posted by 1LUMI335

Just playing devils advocate, can you explain, taken my last post in this thread into consideration aswell, why i ran Map7, an overly agressive map on pump 93 octane for 3 years on a otherwise stock car, and nothing happened to my motor, and its still runing on the same tune till this day, healthy and strong. Just interesting, since we have people giving explanation, but there explanantions seem to give an idea that the motors should explode or break, which hasnt been the case.

Answered multiple times in this thread with the smoking and eating McDonalds every day analogies. Just because YOUR car hasn't blown up yet doesn't mean it's wise to continue running high boost with the stock DME timing retard strategies.

You can smoke cigars every day and live to the ripe old age of 96 like George Burns... or you can die at 45.

Question is, if you can add a huge safety margin by purchasing a tune that costs $160 more to protect your $10K motor, why wouldn't you? Aside from the $10K+ cost to replace an engine, you'd also have to factor in down time, renting a car for a few weeks if not longer, etc.

But, this is really beside the point. If one reads anything about engine tuning, they would understand why it's critical to avoid detonation as much as possible... even running stock boost levels let alone double. It's like saying I always carry my Sig Sauer with a round chambered and safety off and it's never accidentally discharged for the 5 years I've left it in the glove box of my car. So, it must be 100% safe to do so, right?

Then there's the whole principle thing... As a relatively new member on this forum, I find it fascinating how a lot of new members who know nothing of how these "tuner wars" started jump on Shiv for being a dick when the truth of the matter is, Shiv has been remarkably civil about the whole situation.

But, then again, I also pay for my downloaded music and I think inferior, knock off products are pretty lame. So WTF do I know?

[Joshboody]

Originally Posted by ZTUNER

I tried to keep up with this thread all day at work today and a few things bothered me. The first one is about people talking about what the knock sensor can and cannot do.

The knock sensor itself does not know or care about "knock" per say. Infact it knows absolutely nothing. It is a simple , crude microphone with max sensitivity at a particular frequency band. All it produces is "noise" which is filtered by the DME based on frequency, amplitude and crank angle. A lot of the noise is discarded based on crank angle. If there are multiple knock sensors they can be used to triangulate the noise to figure out which cylinder the knock is coming from. There is a table in all ecus which describes the intensity of this noise (usually voltage) vs rpm. This would be the most basic setup.

For example take this theoretical table rpm vs knock sensor voltage. (not real just an example)
rpm knock voltage
1000 1
2000 2
3000 3
4000 3.5
5000 4
6000 5

If say at 5000 rpm the voltage measured by the knock sensor is greater than 4 volts it refers to another table that looks at "how much more"
If its .1 volt more it pulls 1 deg
If it .5 volt more it pulls 3 deg etc etc
If its off the charts it throws a code and drops you into fail safe

This can be further refined by running this against a "duration" table which controls the "decay" of ignition retard. For instance, if the knock voltage is hitting .5 volts higher than 4 volts at 5000 rpm 5 times a sec, it might pull timing 6 deg vs 3 deg for 2 times a sec and it might keep the timing pulled down longer.

All this can be further affected by other variables like boost level, AIT, Ambient temps or water temps or oil temps, EGT's or whatever predictive element BMW deems appropriate. Apparently Cobb logs show a MAF output which imples that MASS of air is somehow being measured or calculated even though this a MAP based car. I did not see a MAF sensor on here. IMO the decision tree controlling the sensitivity of the knock sensor output could be highly complicated in this car and i am quite sure that unless we know and understand fully how the DME manipulates the knock sensor data we will never know why timing is pulled. Obviously we can see patterns and try to guess but it does not mean our guesses are accurate.

All this being said with higher boost comes , higher AIT's and higher EGT's and maybe higher water and oil temps. Maybe there is a table or two or three that looks at these variables and drops timing if the rate of increase of these variables excedes a certain threshold ? Who knows ? I am certain the piggy tuners and even the flash tuners do not know ALL these variables. Maybe there is a predictive table in the DME that uses a bunch of variables that that are used to figure out mathematically what the cylinder pressures are and it pulls timing based on THAT and not knock itself ? You WILL control Cylinder pressures if you advance or retard timing. Thats why you don't see a huge change in a/f between gears but there is a significant in timing.

ONLY way to truly tell if you have knock is to hook up a high quality knock sensor and hear it thru headphones and tune accordingly. You can also log the knock count PID if that is indeed loggable with this DME to see what the ecu is seeing. You cannot log timing alone and look at drops in timing and assume it is knock. It is a guess at best.

All this being said we should quit attacking vendors and try to understand how this s@#$ works. Its facinating at the very least

Harry

PS : i didn't get into the other things because the first thing was long enough

my understanding, from Cobb mainly, is that BMW does NOT have tables for IAT, baro, etc... just load and rpm... one map and then knock signal. not sure how deep they have gotten though.

[BrianMN]

Originally Posted by AlterZgo

Answered multiple times in this thread with the smoking and eating McDonalds every day analogies. Just because YOUR car hasn't blown up yet doesn't mean it's wise to continue running high boost with the stock DME timing retard strategies.

You can smoke cigars every day and live to the ripe old age of 96 like George Burns... or you can die at 45.

Question is, if you can add a huge safety margin by purchasing a tune that costs $160 more to protect your $10K motor, why wouldn't you? Aside from the $10K+ cost to replace an engine, you'd also have to factor in down time, renting a car for a few weeks if not longer, etc.

But, this is really beside the point. If one reads anything about engine tuning, they would understand why it's critical to avoid detonation as much as possible... even running stock boost levels let alone double. It's like saying I always carry my Sig Sauer with a round chambered and safety off and it's never accidentally discharged for the 5 years I've left it in the glove box of my car. So, it must be 100% safe to do so, right?

Then there's the whole principle thing... As a relatively new member on this forum, I find it fascinating how a lot of new members who know nothing of how these "tuner wars" started jump on Shiv for being a dick when the truth of the matter is, Shiv has been remarkably civil about the whole situation.

But, then again, I also pay for my downloaded music and I think inferior, knock off products are pretty lame. So WTF do I know?

Me like common sense.

[ZTUNER]

Originally Posted by Joshboody

my understanding, from Cobb mainly, is that BMW does NOT have tables for IAT, baro, etc... just load and rpm... one map and then knock signal. not sure how deep they have gotten though.

Did they say that in a thread ? and if so can you link me please ? Also where the heck do they get a maf count ? Is it calculated or is there asensor somewhere ?
Harry

[AlterZgo]

Originally Posted by ZTUNER

All this being said we should quit attacking vendors and try to understand how this s@#$ works. Its facinating at the very least

Exactly! And, if you don't want to learn how this shit works, PLEASE don't install ANY tune in your car. Drive it stock and enjoy. This is not an iPhone or an Xbox you're trying to hack. It's a $50K car! You can't box that shit up and return it to Best Buy!

[ZTUNER]

Originally Posted by AlterZgo

Exactly! And, if you don't want to learn how this shit works, PLEASE don't install ANY tune in your car. Drive it stock and enjoy. This is not an iPhone or an Xbox you're trying to hack. It's a $50K car! You can't box that shit up and return it to Best Buy!

my reason for modding is to "see what happens when". I don't want the solution in a pre boxed package. To me the journey is more important than the destination. I have no interest in the vendor wars and i honestly don't care who lied when and how much and to whom.

Harry

[AlterZgo]

Originally Posted by ZTUNER

my reason for modding is to "see what happens when". I don't want the solution in a pre boxed package. To me the journey is more important than the destination. I have no interest in the vendor wars and i honestly don't care who lied when and how much and to whom.

Harry

FWIW, when I said "you" I meant it generically, not you and as in Harry.

[Joshboody]

Originally Posted by ZTUNER

Did they say that in a thread ? and if so can you link me please ?

Harry

it was in the long Cobb thread and not stated clearly, but alluded to. i understand its how they wrote the intial code for Cobb flash... 1 timing map. Cobb is very confident in the BMW knock retard logic, but of course they rewrote the table.

[Alext]

Originally Posted by AlterZgo

It's a $50K car! You can't box that shit up and return it to Best Buy!

Yes that is right. In my country we get ripped off...my N54'd BMW cost me >US$140k...so I follow this thread with great interest. (& income tax rates are higher here too!!).

I hate to think what a new engine would cost...

[Morpheus]

Originally Posted by Alext

Yes that is right. In my country we get ripped off...my N54'd BMW cost me >US$140k...so I follow this thread with great interest. (& income tax rates are higher here too!!).

I hate to think what a new engine would cost...

Almost no one here has had to buy a new engine. I posted a link to pics of damage on cylinder #5 on an n54 motor that was smoking and getting progressively worse. Not a JB car.

Someone mentioned overclocking. Even done correctly, overclocking will shorten the life of the computer. If you're really worried about a new engine. Leave it stock.

Since you guys are so into metaphors and analogies, how about this? Remember Geraldo and Al Capone's vaults???

[michael1284]

i just don't understand how something can still be beta when it's completely open to the public lol

[Sniz]

Originally Posted by michael1284

i just don't understand how something can still be beta when it's completely open to the public lol

shhhhhhh

[Black Gold]

Originally Posted by michael1284

i just don't understand how something can still be beta when it's completely open to the public lol

What better way to

a) make it sound cool and top secret using the word "beta"

b) alleviate all blame if things go wrog saying, "hey man, it IS just a beta product"

and not just bms does this, others do as well. I share your sentiments on this subject

[SennaZ]

Originally Posted by ZTUNER

I tried to keep up with this thread all day at work today and a few things bothered me. The first one is about people talking about what the knock sensor can and cannot do.

The knock sensor itself does not know or care about "knock" per say. Infact it knows absolutely nothing. It is a simple , crude microphone with max sensitivity at a particular frequency band. All it produces is "noise" which is filtered by the DME based on frequency, amplitude and crank angle. A lot of the noise is discarded based on crank angle. If there are multiple knock sensors they can be used to triangulate the noise to figure out which cylinder the knock is coming from. There is a table in all ecus which describes the intensity of this noise (usually voltage) vs rpm. This would be the most basic setup.

This would be a good explanation if if was 1980 but most manufacturers have taken knock detection circuitry well beyond this. Most use multiple frequencies, multi-stage filtering, FFT and other sophisticated algorithms to analyze knock. What Mike has been calling 'pre-knock' they would probably term 'light knock' which has it's own set of amplitude parameters and is definitely not harmful to the engine. Honestly, don't know what BMW does, but it's likely they have kept up with modern technology.