E90Post
 


 
BMW 3-Series (E90 E92) Forum > E90 / E92 / E93 3-series Powertrain and Drivetrain Discussions > N54 Turbo Engine / Drivetrain / Exhaust Modifications - 335i > The Basics of Tuning and Timing



Reply
 
Thread Tools Search this Thread
      02-24-2011, 04:06 PM   #67
biz77
Lieutenant
156
Rep
598
Posts

Drives: 2021 X3 M40i
Join Date: Aug 2010
Location: Spokane, WA

iTrader: (4)

Quote:
Originally Posted by Mike@N54Tuning.com View Post
...the DME logic starts overreacting with 3 degree drops instead of small 1/4 degree drops and you know it's time to dial back the boost. This is the basis for piggyback autotuning.

Mike
What is it the DME is "overreacting" to?
Appreciate 0
      02-24-2011, 04:07 PM   #68
Jeff@TopGearSolutions
Jeff@TopGearSolutions's Avatar
United_States
3441
Rep
79,212
Posts

Drives: C6 Z06, 09 335i, 10 335xi
Join Date: Dec 2008
Location: www.TopGearSolutions.com

iTrader: (37)

Quote:
Originally Posted by biz77 View Post
Understood. MAP is the main factor in determining load. If we look at manifold pressure on an N54 at a given RPM and IAT at sea level and then at 5,000 ft. ASL, with the same IAT and RPM, we should see almost identical manifold pressure and therefore load, no?
I believe so.
Quote:
Originally Posted by roninsoldier83 View Post
While the DME is setup to attempt to make the same calculated torque regardless of conditions, I can attest that it is unable to make the same torque/peak power numbers at high elevations as it does at sea level.

I say that because I've seen several 335's at a local drag strip (Bandimere Speedway), which has an elevation of about ~5800 ft. The average trap speed for a stock 335 up here seems to be around the 95mph mark, which is a far cry from the 100-105mpg trap speeds most people seem to be pulling at sea level. 91 octane piss water and uber-thin air FTL
I dont doubt that at all. I was just reiterating the fact of what BMW wanted the car to do in theory. In practice it comes up short.
Appreciate 0
      02-24-2011, 04:23 PM   #69
roninsoldier83
Second Lieutenant
roninsoldier83's Avatar
56
Rep
219
Posts

Drives: E82 128i 6MT / AP2 Honda S2000
Join Date: Oct 2010
Location: Denver, CO

iTrader: (1)

Quote:
Originally Posted by Mike@N54Tuning.com View Post
As I said the N54 platform is pretty unique which is why people who come from other backgrounds always have your strong opinions until they do the tuning themselves. Fueling is closed loop during WOT using wideband sensors. Boost is vacuum controlled with a default open position. Timing is closed loop. It uses what we refer to as "maximum set points" and continually adds timing, listens for early knock, and removes timing, until it hits its maximum set point. Which are set so high that it would take race gas to hit them on a stock car. It uses short term trims (knock) and long term trims (octane) analogous to long and short term fuel trims. This is in addition to all the other factors that influence the timing mapping. If you add race gas you will automatically increase advance in a factory stock or tuned car as they are all running on this knock threshold system. As you get close to not being able to support your boost with your octane the DME logic starts overreacting with 3 degree drops instead of small 1/4 degree drops and you know it's time to dial back the boost. This is the basis for piggyback autotuning.

Mike

Ahhh, "maximum set points".... what you just described, is a timing table with "maximum set points" (typically referred to as base timing values) that is based on a load site.

If you need race gas to hit them from the factory (assuming you're correct), that means we're not talking about light load/cruise/idle timing numbers (typically close to the ~40 degree mark on most cars), but actual WOT timing numbers, correct? If there is a timing "limit" that is attainable (with race gas or not), that means there's a table where it is capped somewhere within the DME/ECU, and typically there is going to be a load value associated with this limit... Just my $.02


Honestly, I mean this with no malice, I can't wait for ATR to be released so I can actually take a look into these DME's for myself. How I miss the days of free downloadable tuning software *sigh* haha!!
__________________
2008 Honda S2000 (weekend/former autoX) - 1996 Acura Integra GS-R (occasional autoX) - 2008 BMW 128i 6MT (track/time attack) - 2008 Nissan Xterra (winter beater)
Appreciate 0
      02-24-2011, 04:26 PM   #70
roninsoldier83
Second Lieutenant
roninsoldier83's Avatar
56
Rep
219
Posts

Drives: E82 128i 6MT / AP2 Honda S2000
Join Date: Oct 2010
Location: Denver, CO

iTrader: (1)

Quote:
Originally Posted by JPSlick View Post
I believe so.


I dont doubt that at all. I was just reiterating the fact of what BMW wanted the car to do in theory. In practice it comes up short.
Yeah, but it's to be expected haha! BMW's seem more adaptive to this elevation that several other turbo cars I've had experience with up here... but when you live this high up, well, we come to expect it up here
__________________
2008 Honda S2000 (weekend/former autoX) - 1996 Acura Integra GS-R (occasional autoX) - 2008 BMW 128i 6MT (track/time attack) - 2008 Nissan Xterra (winter beater)
Appreciate 0
      02-24-2011, 04:35 PM   #71
OpenFlash
United_States
1731
Rep
17,960
Posts

Drives: A Lot
Join Date: Sep 2006
Location: SF Bay, CA

iTrader: (0)

Garage List
The idea of relying on "closes loop timing control" to do the job of the tuner/calibrator is poor form. And if either Mike@n54 or Terry had more tuning experience, I would expect them to agree. The idea of tuning any closed loop system is to dial in the raw tables/compensations so as to minimize reactive closed loop corrections. This is done when controlling boost, fuel and timing. With timing control being the least tolerant positive error. Defending the undefensable suggests either ignorance or the attempt to maliciously deceive. These are facts and no one can change that.

Last edited by OpenFlash; 02-24-2011 at 04:53 PM..
Appreciate 0
      02-24-2011, 04:46 PM   #72
Joshboody
Lieutenant Colonel
65
Rep
1,708
Posts

Drives: pickemuptruck
Join Date: Jul 2009
Location: Phoenix, AZ

iTrader: (7)

GOOD READ... made my wait in the airport enjoyable.
Appreciate 0
      02-24-2011, 04:56 PM   #73
BzzzBom
Major
93
Rep
1,216
Posts

Drives: 05 R53 MINI / 09 335i Coupe
Join Date: Jul 2006
Location: Los Angeles

iTrader: (0)

Based on the above discussion, which flash and piggyback systems are the most sophisticated and safe for our engine? GIAC > Cobb > Procede >>> JB?
Appreciate 0
      02-24-2011, 05:10 PM   #74
Joshboody
Lieutenant Colonel
65
Rep
1,708
Posts

Drives: pickemuptruck
Join Date: Jul 2009
Location: Phoenix, AZ

iTrader: (7)

I don't have much tuning experience so, so to keep this a general tuning thread...

For timing maps, in general, can you scale the ingnition point based on just a couple part and WOT observations. Is a load / rpm table linear? Ie. 6000rpm compared to 5000 20% increase at same load? Same thing with load? How do you factor vtec, vvt, vanos , ect... I would think you'd have to take samples before and after the event.
I'm sure the timing table posted would explain much of this, but I don't have it handy now
Appreciate 0
      02-24-2011, 05:15 PM   #75
Clap135
Brigadier General
Clap135's Avatar
102
Rep
3,460
Posts

Drives: 2009 N54
Join Date: Jul 2009
Location: Sticky's Mom House

iTrader: (1)

Quote:
Originally Posted by Mike@N54Tuning.com View Post
As I said the N54 platform is pretty unique which is why people who come from other backgrounds always have your strong opinions until they do the tuning themselves. Fueling is closed loop during WOT using wideband sensors. Boost is vacuum controlled with a default open position. Timing is closed loop. It uses what we refer to as "maximum set points" and continually adds timing, listens for early knock, and removes timing, until it hits its maximum set point. Which are set so high that it would take race gas to hit them on a stock car. It uses short term trims (knock) and long term trims (octane) analogous to long and short term fuel trims. This is in addition to all the other factors that influence the timing mapping. If you add race gas you will automatically increase advance in a factory stock or tuned car as they are all running on this knock threshold system. As you get close to not being able to support your boost with your octane the DME logic starts overreacting with 3 degree drops instead of small 1/4 degree drops and you know it's time to dial back the boost. This is the basis for piggyback autotuning.

Mike
That is simply untrue, the only reason the ecu advances timing with race gas on a stock car is for 1 reason
The stock is timed to aggressive for the conditions its driven in.

You just wrote a huge paragraph and explained nothing. By simplying lowering the TARGET timing, you achieve knock free operation, just like any other modern platform.

I posted BONE STOCK timing vs load charts that cobb provided multiple times. ARe you simply ignoring them?
Appreciate 0
      02-24-2011, 05:18 PM   #76
Sniz
Lieutenant General
Sniz's Avatar
647
Rep
10,587
Posts

Drives: e92 335 - gone // e36 M3 turbo
Join Date: Mar 2007
Location: Ellicott City, MD

iTrader: (1)

Quote:
Originally Posted by Joshboody View Post
GOOD READ... made my wait in the airport enjoyable.
to the CLOUD!!!
__________________
Appreciate 0
      02-24-2011, 05:46 PM   #77
roninsoldier83
Second Lieutenant
roninsoldier83's Avatar
56
Rep
219
Posts

Drives: E82 128i 6MT / AP2 Honda S2000
Join Date: Oct 2010
Location: Denver, CO

iTrader: (1)

Quote:
Originally Posted by Clap135 View Post
That is simply untrue, the only reason the ecu advances timing with race gas on a stock car is for 1 reason
The stock is timed to aggressive for the conditions its driven in.

You just wrote a huge paragraph and explained nothing. By simplying lowering the TARGET timing, you achieve knock free operation, just like any other modern platform.

I posted BONE STOCK timing vs load charts that cobb provided multiple times. ARe you simply ignoring them?
Without coming off an as a prick, I honestly don't think Mike understands what he's talking about. These cars have load vs RPM timing tables just like every other modern vehicle I've ever tuned.... most of which have been posted more than once.

For reference Mike, here's a stock timing map/table for a N54:


On top of that, here's the base timing values for Cobb's stage 1 timing map:




^^^Funny, they look exactly like pretty much every other timing table I've ever played with. They have limits (Mike's "maximum set points") based on load vs RPM.... just like a Subaru... and every other modern car I've ever played with/tuned.


+1 to yourself and Shiv. If base/maximum timing values cannot be obtained under WOT, the timing tables are too aggressive for conditions, and a tuner worth his salt would reduce timing to ensure knock free operation under any load within normal conditions, only relying on the knock sensor as a failsafe.

Most modern cars have knock control systems that decrease/retard timing based on knock, BMW's knock sensors simply appear to merely be more sensitive and the ECU/DME is able to react more quickly.

Cliff notes: N54 timing is controlled in the same manner as any other car I've ever seen- via a RPM vs load table.
__________________
2008 Honda S2000 (weekend/former autoX) - 1996 Acura Integra GS-R (occasional autoX) - 2008 BMW 128i 6MT (track/time attack) - 2008 Nissan Xterra (winter beater)
Appreciate 0
      02-24-2011, 05:51 PM   #78
Clap135
Brigadier General
Clap135's Avatar
102
Rep
3,460
Posts

Drives: 2009 N54
Join Date: Jul 2009
Location: Sticky's Mom House

iTrader: (1)

Quote:
Originally Posted by roninsoldier83 View Post
Without coming off an as a prick, I honestly don't think Mike understands what he's talking about. These cars have load vs RPM timing tables just like every other modern vehicle I've ever tuned.... most of which have been posted more than once.

For reference Mike, here's a stock timing map/table for a N54:


On top of that, here's the base timing values for Cobb's stage 1 timing map:




^^^Funny, they look exactly like pretty much every other timing table I've ever played with. They have limits (Mike's "maximum set points") based on load vs RPM.... just like a Subaru... and every other modern car I've ever played with/tuned.


+1 to yourself and Shiv. If base/maximum timing values cannot be obtained under WOT, the timing tables are too aggressive for conditions, and a tuner worth his salt would reduce timing to ensure knock free operation under any load within normal conditions, only relying on the knock sensor as a failsafe.

Most modern cars have knock control systems that decrease/retard timing based on knock, BMW's knock sensors simply appear to merely be more sensitive and the ECU/DME is able to react more quickly.

Cliff notes: N54 timing is controlled in the same manner as any other car I've ever seen- via a RPM vs load table.
Im so glad you have patience left for this. I would probably curse him out and get banned in my next reply. Hopefully more poeple who understand how tuning works (not saleman) begin to realize the BS the community is being fed here. This is for the good of the n54 tuning world, lying simply makes it good for BMS and screws everyone over.
Appreciate 0
      02-24-2011, 05:57 PM   #79
tscdennab
Banned
149
Rep
1,809
Posts

Drives: BMW 335i
Join Date: Oct 2010
Location: Europe

iTrader: (0)

Garage List
2008 BMW  [6.24]
Quote:
Originally Posted by Clap135 View Post
This is for the good of the n54 tuning world, lying simply makes it good for BMS and screws everyone over.
Then how do you explain the logs that I have posted earlier ?
Appreciate 0
      02-24-2011, 06:00 PM   #80
Clap135
Brigadier General
Clap135's Avatar
102
Rep
3,460
Posts

Drives: 2009 N54
Join Date: Jul 2009
Location: Sticky's Mom House

iTrader: (1)

Quote:
Originally Posted by tscdennab View Post
Then how do you explain the logs that I have posted earlier ?
I explain that as the dme knocking its way multiple times over and over again untill it foudn a timing curve that works, in NO time, it will then being to raise the curve and knock will reoccure everytime it tires. Very simple. Why will try to raise ignition? Because the max timing is set for a stock car.


I got a question for you. Your in the thread that i started, why didnt the stock car in the first post smooth out the timing curve? The op did MULTIPLE logs over a couple months time. Shouldnt the timing curve be buttery smooth.
Appreciate 0
      02-24-2011, 06:00 PM   #81
Mike@N54Tuning.com
Joint Chiefs of Staff
Canada
4899
Rep
115,955
Posts

Drives: 2007 335i, 2015 M3
Join Date: Dec 2008
Location: N54tuning.com

iTrader: (89)

Quote:
Originally Posted by biz77 View Post
What is it the DME is "overreacting" to?
The DME reaction to knock sensor feedback is progressive. Normally it operates within a range than reduces timing in small amounts of say 1/4 degree. If knock sensor noise is above that range it overreacts with 3 degrees. This is when you know it's having a harder time honing in on the right curve which means combustion is becoming less stable. If it continues you get another 3 degrees. Finally if there is still no reduction in knock sensor noise after the two 3 degree drops it assumes you are experiencing preignition or something bad is happening and goes to limp mode / triggers ignition glow codes. The whole time it is learning long and short term trims by RPM, load, gear, etc.

Mike
Appreciate 0
      02-24-2011, 06:08 PM   #82
Mike@N54Tuning.com
Joint Chiefs of Staff
Canada
4899
Rep
115,955
Posts

Drives: 2007 335i, 2015 M3
Join Date: Dec 2008
Location: N54tuning.com

iTrader: (89)

Quote:
Originally Posted by roninsoldier83 View Post
Ahhh, "maximum set points".... what you just described, is a timing table with "maximum set points" (typically referred to as base timing values) that is based on a load site.

If you need race gas to hit them from the factory (assuming you're correct), that means we're not talking about light load/cruise/idle timing numbers (typically close to the ~40 degree mark on most cars), but actual WOT timing numbers, correct? If there is a timing "limit" that is attainable (with race gas or not), that means there's a table where it is capped somewhere within the DME/ECU, and typically there is going to be a load value associated with this limit... Just my $.02


Honestly, I mean this with no malice, I can't wait for ATR to be released so I can actually take a look into these DME's for myself. How I miss the days of free downloadable tuning software *sigh* haha!!
Like I said this platform is pretty unique. Get ready for 14.5:1 air/fuel ratios @ WOT and 5-10 degrees advance @ WOT. The maximum advance set points are based on load, rpm, but also gear, iat, EGT, etc. Then you have the long and short term learned factors that play in. Generally with a JB3 on pump for example the long term octane adaptive is maxed out negative which takes away 2-3 degrees of advance. Which is why you have thousands running tens of millions of miles without knock related damage, ping, or even bad looking plugs. And that is using the old logic that could not even listen for knock / drop outs. The new JB4 logic is going to prove much better. When the ATR comes out it will be interesting to see the actual knock related tables and how it all really works. There may be factors that play in that we're not aware of. But make no mistake, it does work perfectly well as many JB4 logs posted have already shown.

Mike
Appreciate 0
      02-24-2011, 06:13 PM   #83
Jeff@TopGearSolutions
Jeff@TopGearSolutions's Avatar
United_States
3441
Rep
79,212
Posts

Drives: C6 Z06, 09 335i, 10 335xi
Join Date: Dec 2008
Location: www.TopGearSolutions.com

iTrader: (37)

Quote:
Originally Posted by tscdennab View Post
Then how do you explain the logs that I have posted earlier ?
Anyone can grab 1 stellar log. Lets see a bunch more over time and see the consistency and differences with high gears/heat soak/ETC.
Appreciate 0
      02-24-2011, 06:42 PM   #84
Mike@N54Tuning.com
Joint Chiefs of Staff
Canada
4899
Rep
115,955
Posts

Drives: 2007 335i, 2015 M3
Join Date: Dec 2008
Location: N54tuning.com

iTrader: (89)

Quote:
Originally Posted by shiv@vishnu View Post
The idea of relying on "closes loop timing control" to do the job of the tuner/calibrator is poor form. And if either Mike@n54 or Terry had more tuning experience, I would expect them to agree. The idea of tuning any closed loop system is to dial in the raw tables/compensations so as to minimize reactive closed loop corrections. This is done when controlling boost, fuel and timing. With timing control being the least tolerant positive error. Defending the undefensable suggests either ignorance or the attempt to maliciously deceive. These are facts and no one can change that.
If you really believed that there would be no procede autotuning maps. Your autotuning works on knock feedback plain and simple.

Mike
Appreciate 0
      02-24-2011, 06:44 PM   #85
Clap135
Brigadier General
Clap135's Avatar
102
Rep
3,460
Posts

Drives: 2009 N54
Join Date: Jul 2009
Location: Sticky's Mom House

iTrader: (1)

Quote:
Originally Posted by Mike@N54Tuning.com View Post
If you really believed that there would be no procede autotuning maps. Your autotuning works on knock feedback plain and simple.

Mike
Correction his timing offsets work on knock feedback....which is a hell of a lot more valuable then simply riding the knock sensor.
Appreciate 0
      02-24-2011, 07:04 PM   #86
Sniz
Lieutenant General
Sniz's Avatar
647
Rep
10,587
Posts

Drives: e92 335 - gone // e36 M3 turbo
Join Date: Mar 2007
Location: Ellicott City, MD

iTrader: (1)

Quote:
Originally Posted by Clap135 View Post
Correction his timing offsets work on knock feedback....which is a hell of a lot more valuable then simply riding the knock sensor.
and round and round we go again.

I believe those who have a bit of knowledge about tuning and how cars should run will steer clear of JB products but plenty will still use them and not blow their motors.

The world will end when Terry admits his tuning platform was not the best idea. Its not going happen just like not all JB users will blow their motors......99% probably wont.
__________________
Appreciate 0
      02-24-2011, 07:10 PM   #87
ExpensiveTaste
Poor Kid with Dreams
ExpensiveTaste's Avatar
United_States
28
Rep
702
Posts

Drives: BSM 07 E90 335xi
Join Date: Aug 2010
Location: Danbury, CT

iTrader: (4)

Special Request

I understand what the OP is saying in every way and I agree 100%. However, I think there is more to it. Can someone conduct a test when it gets warmer out? Run Map 0 or whatever map reverts to the factory BMW tuning and run 87 octane from an empty tank. It's obvious that there will be knock and the timing will drop down by 3 degrees every time. HOWEVER...I would like to see how long it takes for the ECU to adapt and go back to the normal timing changes of less than 1 degree plus or minus. Also, once it gets there, will it stay there without raising timing back up and then getting back to 3 degree timing drops again. We need to figure out the logic the ECU uses to control adaptive timing. For example, does the ECU only try to adapt back up after x successful small timing advances? Is this WOT adaptation a separate logic, or does cruising conditions affect it as well. I would like to learn more about the ECU logic as I'm sure most of us do...
__________________

'07 BSM 335xi JB4, DCI, CX Racing I/C, ER Chargepipe, Meth, KW V3 - Pending Sale
'95 Mustang GTS 347ci Ported TFS TW Heads,Custom Cam,Victor 5.0 IM,Tubular Front Suspension,Full Exhaust,Big and Littles
Appreciate 0
      02-24-2011, 07:10 PM   #88
roninsoldier83
Second Lieutenant
roninsoldier83's Avatar
56
Rep
219
Posts

Drives: E82 128i 6MT / AP2 Honda S2000
Join Date: Oct 2010
Location: Denver, CO

iTrader: (1)

Quote:
Originally Posted by Joshboody View Post
I don't have much tuning experience so, so to keep this a general tuning thread...

For timing maps, in general, can you scale the ingnition point based on just a couple part and WOT observations. Is a load / rpm table linear? Ie. 6000rpm compared to 5000 20% increase at same load? Same thing with load? How do you factor vtec, vvt, vanos , ect... I would think you'd have to take samples before and after the event.
I'm sure the timing table posted would explain much of this, but I don't have it handy now
I think I understand your question... and if I do, yes, tuning is fairly "linear" per se. For instance, if two of your set points for example are 3000rpm and 3500rpm, and in the same load column, at 3000rpm under XXX load, you have a value of 14.5 degrees of timing, and then under the same load value at 3500rpm you have a timing value of 15.5 degrees, the ECU will "ramp" to that value per se. In other words, even though you do not have a set point at 3250rpm, at that RPM under the same load, the ECU will calculate your timing value to about 15.0 degrees. Does that make sense/answer your question?

It will work the same way when sliding between load values at a fixed RPM. A 3D image is a bit easier to understand, as it really does create a "map".

I've never played with Vanos, but I have played with AVCS and MIVEC (cam timing), and they have their own separate tables/separate from ignition timing tables. Honestly, I never invested too much time into cam/valve timing parameters, as I found that you can usually find a bit of lowend torque playing with them/decrease spool time just a bit, but for the most part I found most of them were pretty well tuned from the factory.

Haven't played with too many Honda's (very minor experience with some of the Hondata software), but VTEC typically has its own set point (separate adjustable parameter) that most people adjust at a set RPM. VTEC is just a simple extra lobe/rocker arm that adjusts lift/duration at a set point:


If/when the extra lift and duration increases load on the engine, the ignition timing map/table will be consulted in order to attempt to target the correct amount of timing advance.

I hope that helps.
__________________
2008 Honda S2000 (weekend/former autoX) - 1996 Acura Integra GS-R (occasional autoX) - 2008 BMW 128i 6MT (track/time attack) - 2008 Nissan Xterra (winter beater)
Appreciate 0
Reply

Bookmarks

Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off



All times are GMT -5. The time now is 06:50 AM.




e90post
Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2024, vBulletin Solutions Inc.
1Addicts.com, BIMMERPOST.com, E90Post.com, F30Post.com, M3Post.com, ZPost.com, 5Post.com, 6Post.com, 7Post.com, XBimmers.com logo and trademark are properties of BIMMERPOST