What are you waiting for? The A2L has been public and freely downloadable for years. If solving your knock problem is as simple and defining 6 tables and increasing some values then you can get started and post up the results of your custom knock tuning.
N54 N54 False Knock solved in tuning, even for built closed deck motors
- Thread starter nyt
- Start date
You can use forum PMs or just post here, this isn't E90post, you won't get banned for discussing anything technical.
Ive had the chance to use a tuner nerd knock monitor and the false knock doesnt seem to be related to a bad knock calibration from the factory. Retarding timing or leaning out fuel seems to lower the timing reduction. There's a mess of tables related to di but i feel confident the root is injection happening to close to ignition. I am on standalone now but injection timimg seems to be the answer for stock internal guys. That xdf nyt shared has a good bit of fuel timing tables and if i remember right the map ids are in the comments to search the fd docs and find some guidance.
Interesting, you've seen leaning out fueling helped reduce timing pulls? I always thought it was the opposite - richen the mix to make it more knock resistant? More fuel to evaporate = cooler charge temps and richer afr slowing burn rate, all pushing out the knock threshold?
Thats what leads me to it being injection timing. Iam still working on fine tuning injection timing on the s7 but on di we cant spray fuel right up to injection there needs to be a debarable difference (ive seen 46 to 32 degrees from reading) i haven't gone to far into this yet because iam still working on getting the s55hpfp stuff together i just set a 4.8ms limit on di and run pi the rest of the way up so far. I did notice with 117 in the tank i would start to see trouble with injection getting past about 34 degrees to ign but vvt also seems to play a role. Food for thought tho ive seen so many misfire "solutions" but the window between injection and ign seem to play a big role to
Wow, great observation. I wonder if there is a time window when the DME is listening for knock, and if injection gets too close to the ignition point then the sound/vibration from the injector closing could be misinterpreted as knock/pre-ignition?
So when you've seen improvement from leaning up, is the theory then that by injecting less fuel, the injection event finishes further ahead of the ignition point (before the point where that sound could register as knock)?
For the injection timing maps, do you know is it based on what crank angle to start injection, or is it what crank angle to end injection on?
This looks a lot like the same knock control measurement values that the DME can be triggered to broadcast every 10ms. For my own edification, is that correct?
I gather that BMW did much the same the same thing as you suggested on the test bench when conducting the original design of experiments (DoE) to define the maps, curves and labels. I find the whole process of OEM engine calibration fascinating.
Exactly it takes x amount of injector opening time to put enough fuel in mg/stroke. The stock dme fueling is complex and i honestly never completely fugured it out before i converted to the s7.
The timings for injection are start of injection.
There are split angles for single and double injection. All homo tables the strat and homo strat arent used but the new mhd rom does seem to repurpose some of them, posible further complication.
There are probably most certainly more tables to add to the xdf and iam unaware if vvt adjustments are compensated for?
Definitely in need of more logging channels unfortunately i cant test anything now running a standalone but i had this suspicion for a while and with the ease of moving injection timing around on syvecs it seems to add up.
Figured id put it out there tho for you guys still pushing the stock dme
Appreciate the knowlege sharing!!Exactly it takes x amount of injector opening time to put enough fuel in mg/stroke. The stock dme fueling is complex and i honestly never completely fugured it out before i converted to the s7.
The timings for injection are start of injection.
There are split angles for single and double injection. All homo tables the strat and homo strat arent used but the new mhd rom does seem to repurpose some of them, posible further complication.
There are probably most certainly more tables to add to the xdf and iam unaware if vvt adjustments are compensated for?
Definitely in need of more logging channels unfortunately i cant test anything now running a standalone but i had this suspicion for a while and with the ease of moving injection timing around on syvecs it seems to add up.
Figured id put it out there tho for you guys still pushing the stock dme
Something else. Its bin an uphill battle getting the di pump under control. The way syvecs does it is with pwm closed loop with only p and i so no derivitive i had to make some custom maps to keep it mostly stable.
In the process i seen some things...scary things...
The sensor or presure relief valve maxes out at 4351psi. When iam pegged there long enough lp goes up to as high as 115psi ive seen even with the pr hat with the 72.5psi regulator.
Around 95psi the fuel lines start banging off the car like plumbing in an old house
Injectors seem fine at that pressure tho i just got the s55 hpfp and mocked it up to a spare block i have, from what i think looking through syncronous pump control ill have to get a custom gear made and chain posibly tensioner? To get the pump to turn 1:2 for 720 sync to get syvevs to control it. The project is in its infancy just a rainy day bad weather pass time for now but i cant wait to see what we can really get out of these injectors
In the process i seen some things...scary things...
The sensor or presure relief valve maxes out at 4351psi. When iam pegged there long enough lp goes up to as high as 115psi ive seen even with the pr hat with the 72.5psi regulator.
Around 95psi the fuel lines start banging off the car like plumbing in an old house
Injectors seem fine at that pressure tho
i just got the s55 hpfp and mocked it up to a spare block i have, from what i think looking through syncronous pump control ill have to get a custom gear made and chain posibly tensioner? To get the pump to turn 1:2 for 720 sync to get syvevs to control it. The project is in its infancy just a rainy day bad weather pass time for now but i cant wait to see what we can really get out of these injectorsSimilar threads
