N54 N54 False Knock solved in tuning, even for built closed deck motors

nyt

Sergeant
Sep 15, 2019
331
1
289
0
Ride
2010 335xi
Thanks to @Shitrockn54 for defining the tables, I've managed to tune out the false knock issues, even with with built closed deck motors.

My car was constantly pulling timing and throwing super knock codes on straight e85 before, especially as I turned the boost up. The combination of ije0s, xdrive, and auto seemed to be the main culprits.

These are the tables I've changed, in addition to using the MHD desensitize setting. The 12 pre-amplification tables are using the stock values scaled by 0.45. This is likely too aggressive for blocks without a closed deck. In order to scale them this low, I had to lower the minimum noise signal level to prevent throwing a code. I think 0.4 would be the lowest I'd scale the preamp tables, and less depending on the engine build. The closed deck transmits combustion to the sensors at much higher levels.

knockfix.png


Logs
log.png

log2.png



The XDF for ije0s with these tables defined is attached. Careful if not running flex fuel, I'm just using this to edit the knock tables.
 

Attachments

  • Screenshot 2022-06-19 010417.png
    Screenshot 2022-06-19 010417.png
    8.1 KB · Views: 161
  • 1655368481876_IJE0S (Standard Units) - Flex Fuel customxdf.xdf
    1 MB · Views: 103
Last edited:

nyt

Sergeant
Sep 15, 2019
331
1
289
0
Ride
2010 335xi
Trying this after I swap some plugs out tomorrow... The 12 amp tables are scaled by 0.45. I think this should do the trick, but will see. That may be too aggressive for some, but if you have a closed deck and forged internals, have at it.

(data updated above)
 
Last edited:

fmorelli

Lieutenant General
Staff member
Aug 11, 2017
3,748
3,592
0
57
Virginia
Ride
E89 Z4 35i, F10 535d
@nyt very interesting ... thanks for the details and please keep us updated. Good luck!
 
  • Like
Reactions: nyt

JohnDaviz

Lieutenant
Jan 6, 2019
863
577
0
Ride
335i E92 DCT
I am tuning a guy on IA80S (and me INA0S once the issues are figured) with a closed Deck and forged internals. Would somebody be so kind to define those tables on that Firmware (and INA0S)? I can test those settings on both cars.

Begging at @all4bspinnin and @AzNdevil :-D
 
  • Agree
  • Like
Reactions: Sava666 and TiAgAu

nyt

Sergeant
Sep 15, 2019
331
1
289
0
Ride
2010 335xi
Looking good with these latest changes. After lowing cl_nl_min_diag by half I was able to scale the 12 preamp tables by 0.45. Car has never held timing through shifts before. Even with factory block and twins.

1-4.png


getting boost on target... still looking good.
1-4.png
 
Last edited:

Jake@MHD

Major
Platinum Vendor
Nov 7, 2016
1,593
2,058
0
Philly
Nice to see some results from this. But I really don't recommend just making changes to these willy nilly. I had these all defined for 2yr now, and offered to provide them numerous times to someone willing to go on a dyno w/ det cans / knock ears / plex knock monitor etc to properly adjust them. Never had any takers. Being a JB4 car you can't even log the right custom params like current noise level vs. knock, etc.
 
  • Informative
  • Like
Reactions: Torgus and wheela

135boost

Corporal
Oct 28, 2017
243
1
136
0
Finland
Ride
135is
For sure im more than interested if my vanos issues is solved with new vanos solenoids... volunteer..

I has allready new orginals but i get out of hospital today so first possible time to test solenoids is tomorrow

If issue is solved, then just time and i be there 😀.. place is known

*************Edit************

Vanos issues seems to be solved so maybe next a meeting with p33p33 and arrange that dynoday to do some logs.
I think we must sit down and take few map cleaning before that dyno cause now i have done awful lot of testing to resolve that odd ans now clearly superknock related issue so there can be some supid valuea in my bin what needs to be cleaned.
 
Last edited:
  • Like
Reactions: nyt

nyt

Sergeant
Sep 15, 2019
331
1
289
0
Ride
2010 335xi
Nice to see some results from this. But I really don't recommend just making changes to these willy nilly. I had these all defined for 2yr now, and offered to provide them numerous times to someone willing to go on a dyno w/ det cans / knock ears / plex knock monitor etc to properly adjust them. Never had any takers. Being a JB4 car you can't even log the right custom params like current noise level vs. knock, etc.

And yet here we are, with the issue resolved and the potential to make further progress. Instead of having to use pots or resistors to trim out the knock sensors, we can use these tables.

On that topic, I asked about the tables a number of times, but you did not share them. Not that you're beholden to anyone to do so, but not sharing the tables has held the platform back. Weren't donations even taken at some point for work on this?

Not everyone has access to rollers and expensive knock monitoring tools or are willing to run their engine into knock. While it's the best way to move forward on this, it's not the only way to work through the issue. Once I had the continental docs, table list, and someone kind enough to take the time to define the tables I requested, I had the false knock dialed out within 3 revisions.
 
  • Like
Reactions: iminhell1

carabuser

Lieutenant
Oct 2, 2019
870
1
765
0
UK
Ride
Z4 35i & 335i
I think Jake doesn't want to release tables without assurance that the people using them know what they are doing. If he released the full knock tables and ended up with a swathe of blown motors as a result then it would rest on his name.

I gave the tables required to completely disable knock control to 135iboost to fix his issue with the warning that all knock control would be disabled. I'll do the same for anyone that asks because I'm not employed by a popular tuning platform and have no reputation to uphold.

By reducing the values in those tables in that manner you've not really fixed anything, you've just reduced the knock thresholds by an arbitrary amount so that the knock control is no longer applying corrections. How do you know that there is no damage being done?
 

P33P33

Specialist
Apr 3, 2017
57
21
0
Ride
BMW 335i n54 e93
@carabuser has a valid point. I was in dyno with customer car with stock block and we used plex monitor to see if there was any room to desensitiz/raise knock threshold. It was e50 in the tank and to my suprise plex identified exactly the same knock the stock sensors/calibration did. So this definitely should be tested with det cans or plex etc.
 

fmorelli

Lieutenant General
Staff member
Aug 11, 2017
3,748
3,592
0
57
Virginia
Ride
E89 Z4 35i, F10 535d
I think Jake doesn't want to release tables without assurance that the people using them know what they are doing. If he released the full knock tables and ended up with a swathe of blown motors as a result then it would rest on his name.
I completely disagree with this assertion. It's information. What people do with it is their responsibility, not the provider of the information. Pure and simple.

We don't need more people asserting that being responsible adults is not the responsibility of adults.

And in fact, worse - that you expect the reputation of a responsible person to ride on other irresponsible people. That's absurd.
 

all4bspinnin

Corporal
Jun 12, 2017
178
98
0
Ride
135i TT
I am tuning a guy on IA80S (and me INA0S once the issues are figured) with a closed Deck and forged internals. Would somebody be so kind to define those tables on that Firmware (and INA0S)? I can test those settings on both cars.

Begging at @all4bspinnin and @AzNdevil :-D
I can define them for you if you want. Can you message me the table names and ill make it happen
 
  • Like
Reactions: wheela

nyt

Sergeant
Sep 15, 2019
331
1
289
0
Ride
2010 335xi
@carabuser has a valid point. I was in dyno with customer car with stock block and we used plex monitor to see if there was any room to desensitiz/raise knock threshold. It was e50 in the tank and to my suprise plex identified exactly the same knock the stock sensors/calibration did. So this definitely should be tested with det cans or plex etc.

Sure, but when you're running e85 on a 10.2:1 build with a closed deck and getting super knock at 20psi without massive timing pulls, things need to be adjusted. People were reverting to using pots and resistors, and that's just an ugly solution. I pulled my tables down to the level where I wasn't seeing random false knock show up. These values will be different for everyone depending on build, and should be tuned properly on the dyno. At least those wanting to do so know which tables to work with now and have it available. Using Plex monitor is without a doubt the best way to do monitor while making adjustments. Hopefully you can get a car on the dyno with this and validate things sometime soon.
 

Jake@MHD

Major
Platinum Vendor
Nov 7, 2016
1,593
2,058
0
Philly
I had the false knock dialed out within 3 revisions.

No offense, but how do you know for certain you only tuned out the false knock? Without logging any of the appropriate DME parameters? And / or without a proper external knock monitoring solution.

For reference, I've attached the "small" list of logging params I prepped for proper tuning of this for IJE0S.
 

Attachments

  • IJE0S_UserChannels_Knock.xml.zip
    1.3 KB · Views: 94
Last edited:

nyt

Sergeant
Sep 15, 2019
331
1
289
0
Ride
2010 335xi
No offense, but how do you know for certain you only tuned out the false knock? Without logging any of the appropriate DME parameters? And / or without a proper external knock monitoring solution.

For reference, I've attached the "small" list of logging params I prepped for proper tuning of this for IJE0S.

For certain? I don't have rollers anymore and no plex to test, so I can't say for certain. I can say that running <20psi using e85 at stock compression with a single turbo, a cool intake charge, and good afrs, there should not be knock or timing pulls. All fuel and ignition components are new and tested. I haven't run my setup to a point where it would have real knock to test, but I have gradually lowered it so the false knock is gone. I trust this method more than throwing a pot inline, which was the only other option to address this and have a running car. I've pushed the car to 38psi and there are no audible signs of knock.
 

djslapdaddy

New Member
Oct 3, 2022
5
0
0
I would be more than happy if somebody would define the tables for IA80S (and may be INA0S)
fucking 2023 and im still looking i believe its just 6 tables and you increase values to have less detection but someone with more exp can chime in.