[brycey2010]

Originally Posted by Doug007

I agree that IF the car starts knocking, the DME will start pulling timing to save the engine and will really affect performance negatively. Keep in mind he is in FL so probably runs 93 everyday. I don't know what other mods he has (meth?) that would raise the knock threshold even higher.


The factory DME targeting a torque level certainly levels out performance with varying conditions, but performance will still change in cold dry vs. hot humid conditions. When tuned to higher boost levels (any tune) you will feel more variation because the tunes are targeting a boost and this target does not move to maintain a constant power because you are much closer to the knock threshold as you said.




Ugh.

Again, if this thread was about losing performance when hot on a JB3/map 5 with some datalogs showing knock I might agree with you. But it's not. I understand the point on knock hysteresis (i.e., one knock event will cause the car to retard timing across the board and timing will only gradually be advanced), but there's nothing to say his car was knocking here...

He just said the car feels different some days. I have felt this in everycar i've owned. I've felt it stock and felt it on PROcede.

true for me as well, i guess you notice it allot more with higher performance vehicles/ I do run 93 always, only engine mod i have is jb3 2.0 map 5.

[BrianMN]

Originally Posted by Doug007

I agree that IF the car starts knocking, the DME will start pulling timing to save the engine and will really affect performance negatively. Keep in mind he is in FL so probably runs 93 everyday. I don't know what other mods he has (meth?) that would raise the knock threshold even higher.



The factory DME targeting a torque level certainly levels out performance with varying conditions, but performance will still change in cold dry vs. hot humid conditions. When tuned to higher boost levels (any tune) you will feel more variation because the tunes are targeting a boost and this target does not move to maintain a constant power because you are much closer to the knock threshold as you said.




Ugh.

Again, if this thread was about losing performance when hot on a JB3/map 5 with some datalogs showing knock I might agree with you. But it's not. I understand the point on knock hysteresis (i.e., one knock event will cause the car to retard timing across the board and timing will only gradually be advanced), but there's nothing to say his car was knocking here...

He just said the car feels different some days. I have felt this in everycar i've owned. I've felt it stock and felt it tuned.

Doug- I don't think you fully understand how the knock threshold system works. That's fine..not many people do, I probably don't either, haha

How i understand it is theDME/knock sensors are always 'sensing' the combustion chamber and gradually applying the timing curve based on that. The sensors don't just feel an audible Knock and send a signal then; they are constantly sending signals to the DME. They can 'sense' when it the engine is 5 degrees away from knock, when it is 4, 3, 2 degrees from knock, etc. Therefore, the car doesn't need to be knocking to be pulling timing. Result: Pulled timing (hurt performance) without a knock occuring.

Hopefully Shiv/Calvin/Knowledgeable person can correct or corroborate this.

[Slidewayz]

Isn't it true that all other things being equal (like AFR), you'll make more power with a little more timing and a little less boost than vice versa?

[D335I]

Me too, still fun but not consistent.

The first week of adaption was really impressive, then little by little seemed to soften, still fun but not the same.

Pulled tune for 1/2 a day for service put her back in and it was awesome, then flattened out over a two week period, weather about the same if not colder.

Still a great tune and getting better.

[Doug007]

Originally Posted by BrianMN

Doug- I don't think you fully understand how the knock threshold system works. That's fine..not many people do.

The DME/knock sensors are always 'sensing' the combustion chamber and gradually applying the timing curve based on that. The sensors don't just feel an audible Knock and send a signal then; they are constantly sending signals to the DME. They can 'sense' when it the engine is 5 degrees away from knock, when it is 4, 3, 2 degrees from knock, etc. Therefore, the car doesn't need to be knocking to be pulling timing. Result: Pulled timing (hurt performance) without a knock occuring.

Hopefully Shiv/Calvin/Knowledgeable person can correct or corroborate this.

Hmm, they never taught me this in my internal combustions engine class in college.

I was always taught that knock is a product of two flame fronts colliding. If there is only one wave front, there is no knock. Now you can knock on a single cycle or knock over and over again.

The engine can certainly sense conditions conducive to knock (high IATs for example) but it can't know the threshold until a knock event occurs since there is no such thing as an onboard octane analyzer. Please correct me if I am wrong and the DME has another strategy to predict knock.

My point in this thread was that the OP was not complaining of knock but day to day variations in performance. This is totally normal and really has nothing to do with whether he has a JB3, PROcede, GIAC, DINAN, or stock tune!

[Doug007]

Here's Shiv's explanation of N54 knock sensing from another thread:

Originally Posted by shiv@vishnu

2) Knock Status--- Self explanatory. In this particular case, the PROcede is actually counting the occurrence of knock every time the engine is under boost. This knock tendency can eventually be used to auto adjust the PROcede's ignition correction amount. Basically taking role away from the factory DME's reactive logic (which actually lets knock occur to a degree before intervening). But for now, it's simply there for the user to monitor if he/she chooses to fine tune the conservative baseline maps for their particular conditions/car/mods. More on that later.

Here's a log of these two new channels in action:

Sure looks like to me that the engine knocks (at least a bit) before it retards timing...

[BrianMN]

Originally Posted by Doug007

Here's Shiv's explanation of N54 knock sensing from another thread:




Sure looks like to me that the engine knocks (at least a bit) before it retards timing...

So You're agreeing with me??

[BrianMN]

Originally Posted by Doug007

Hmm, they never taught me this in my internal combustions engine class in college.

I was always taught that knock is a product of two flame fronts colliding. If there is only one wave front, there is no knock. Now you can knock on a single cycle or knock over and over again.

The engine can certainly sense conditions conducive to knock (high IATs for example) but it can't know the threshold until a knock event occurs since there is no such thing as an onboard octane analyzer. Please correct me if I am wrong and the DME has another strategy to predict knock.

My point in this thread was that the OP was not complaining of knock but day to day variations in performance. This is totally normal and really has nothing to do with whether he has a JB3, PROcede, GIAC, DINAN, or stock tune!

I agree with the point of the thread, it just turned in this direction, and also regarding choice of tunes.

To be on the same page, When I think of a knock, how it was described /used above, or knock status datalogged with procede or bt tool, Im referring to an audible bang. I believe there are 'levels' of severity with knocks, and thats what info the knock sensors provide.What I was trying to say is that the factory dme doesn't just advance timing until it experiences a big knock/knock status, rather it is very intelligent in knowing/predicting what the threshold will be based on input from knock sensors and probably other sensors like you mentioned. I'm by no means an expert in N54 tuning Or knock sensor engineering, Just relaying how I understand it...so that has to be taken with a grain of salt anyway, HaHa

Hopefully Shiv or someone can chime in

[Doug007]

Originally Posted by BrianMN

...DME/knock sensors are always 'sensing' the combustion chamber and gradually applying the timing curve based on that. The sensors don't just feel an audible Knock and send a signal then; they are constantly sending signals to the DME. They can 'sense' when it the engine is 5 degrees away from knock, when it is 4, 3, 2 degrees from knock, etc. Therefore, the car doesn't need to be knocking to be pulling timing. Result: Pulled timing (hurt performance) without a knock occuring.

Brian, the part I had an issue with was that the DME could "sense when the engine is 5 degrees away from knock..."

BMW certainly uses IATs and other sensors to influence timing, but the DME simply can't know how far it is from the knock threshold until there is a knock. There's simply no way for the DME to know the quality of the fuel...if you put in 104 octane race gas or on the other side of the coin screwed up and put in 87. The DME is VERY sensitive to even a single knock event as shown in the log.

As in the log shiv posted, a single knock is enough to pull timing and is nothing you or I would be able to hear. One knock is enough to retard timing by about 10 degrees! (look at the knock event and how timing responds).

[brycey2010]

something could be done to make the cars performance more consistent, it seems to me that the issue is heat related, we all know cars run better, more efficiently in cooler less humid conditions, with or without a tune, so, formula 1 and nascar etc, have the same issues, what is the most affordable way for us to have consistent performance, it must be in cooling, i brought this topic up, because im not talking a little difference in performance im talking allot, i dont think there is a tune that can sort this out, if it could, then the car would have to be run in a de-tuned state to compensate for varying conditions, in other words tune the car for best performance in the worst conditions, and restrict performance in the best conditions to give consistency. There were days in the 80's where my golf gti felt like a rocket, as indeed my bike did, and there were days when both felt slow, as i said this would be unacceptable in professional racing, so there must be a fix.

[Doug007]

Originally Posted by brycey2010

something could be done to make the cars performance more consistent, it seems to me that the issue is heat related, we all know cars run better, more efficiently in cooler less humid conditions, with or without a tune, so, formula 1 and nascar etc, have the same issues, what is the most affordable way for us to have consistent performance, it must be in cooling, i brought this topic up, because im not talking a little difference in performance im talking allot, i dont think there is a tune that can sort this out, if it could, then the car would have to be run in a de-tuned state to compensate for varying conditions, in other words tune the car for best performance in the worst conditions, and restrict performance in the best conditions to give consistency. There were days in the 80's where my golf gti felt like a rocket, as indeed my bike did, and there were days when both felt slow, as i said this would be unacceptable in professional racing, so there must be a fix.

The car does this stock as described above. The car raises boost in hot conditions to maintain power and lowers boost in colder conditions to keep torque near the same value.

The tunes try to extract the most power the conditions will permit, just like any race team would. Race cars are not immune from this, but it affects every car in the race equally...

[ak335]

yeah deff feel a difference some days then the others, exp when its nice and coldd

[hotrod182]

Originally Posted by Doug007

Brian, the part I had an issue with was that the DME could "sense when the engine is 5 degrees away from knock..."

BMW certainly uses IATs and other sensors to influence timing, but the DME simply can't know how far it is from the knock threshold until there is a knock. There's simply no way for the DME to know the quality of the fuel...if you put in 104 octane race gas or on the other side of the coin screwed up and put in 87. The DME is VERY sensitive to even a single knock event as shown in the log.

As in the log shiv posted, a single knock is enough to pull timing and is nothing you or I would be able to hear. One knock is enough to retard timing by about 10 degrees! (look at the knock event and how timing responds).

Perhaps he needs to know how a knock sensor works, and what it actually can sense. They are usually Piezo sensors, simply bolted to the engine block. The only thing they can sense is physical knock, as they are tuned to basically "listen" to that kind of frequency/harmonics present in a typical knock event. Sometimes you can even tap on the side of the block with a hammer and trigger the knock sensor signal.