jpurban
Jr. Member
Karma: +5/-0
 Offline
Posts: 40
|
 |
« on: July 20, 2018, 12:21:00 AM »
|
|
|
Was reading the latest Car and Driver today (ZR1 vs GT2RS)... There was an article about BMW 4.4L turbo that really confused me. The same engine is used in both the X5 and the M5, but here is the interesting part... The X5 version has a MAF and the M car version is MAFless from the factory. Article said the choice was made to reduce intake restriction due to MAF, which could be as high as 26%. So, they clearly indicated that MAFless is their preferred option for performance cars.
After reading, I asked my boss about his tune 2003 M3 -- He tells me his car is converted to MAFless and runs great.
Here is my confusion... Why is MAFless generally maligned on this forum when some new factory high performance cars are MAFless? Is it just an ME7 thing? Or, are there physical characteristics that make MAFless a bad idea generally -- suggesting BMW's design choice isn't ideal?
I'm asking because I really don't understand and am curious. I'm hoping someone here has some insight to share.
|
|
|
|
|
Logged
|
|
|
|
bobbyz0r
Jr. Member
Karma: +3/-0
Offline
Posts: 43
|
|
« Reply #1 on: July 20, 2018, 01:23:39 AM »
|
|
|
It's when people try to run MAFless on a ECU software not designed with proper speed density, or haven't made the proper changes to allow it to work properly without a MAF.
|
|
|
|
|
Logged
|
|
|
|
|
aef
|
|
« Reply #2 on: July 20, 2018, 01:44:40 AM »
|
|
|
A 2003 M3 without Turbo and single throttle bodies has nothing todo with a highly boosted me7 car.
Feel free to google TPS or alpha n injection to understand how they work.
|
|
|
|
|
Logged
|
|
|
|
|
gman86
|
|
« Reply #3 on: July 20, 2018, 02:50:53 AM »
|
|
|
True MAFless cars usually use two MAP sensors. Well, technically only one is a true "MAP"; the other is a charge pressure sensor. By having a reference pre and post throttle, the ECU is able to calculate actual load rather than running off OEM calibrated VE maps which become less accurate the more the breathing path is modified.
|
|
|
|
|
Logged
|
|
|
|
!nfern0
Full Member
Karma: +10/-1
Offline
Posts: 91
Audi TT 1.8T BAM
|
|
« Reply #4 on: July 20, 2018, 03:39:38 AM »
|
|
|
Hi there,
modern engine setups run mostly MAFless... They've got TMAP-Sensors, so they use the modelled RL-Values dependent on MAP, which are in fact more precise and not affected that hard by aging as the MAF is... As far as I know MAF is only used for adapting the model when the engine is running stationary. A secondary reason is, that modern cars run in some load areas valve overlap which can actually cause air going directly into the exhaust system and not taking part in the combustion -> logically the MAF-reading is not usable for load...
greets!
|
|
|
|
|
Logged
|
|
|
|
jpurban
Jr. Member
Karma: +5/-0
Offline
Posts: 40
|
|
« Reply #5 on: July 20, 2018, 10:13:17 AM »
|
|
|
Awesome responses. Thanks, guys.
So i'm taking away... MAFless is a more resilent/robust/efficient solution AS LONG AS your control scheme uses actual measurements for BOTH 1) pressure upstream throttle AND 2) pressure downstream throttle to predict relative load.
Further curiosity... ME7 can accept both of these pressure inputs, right? Are there any "proper" MAFless ME7 factory applications? Does BMW use Bosch ECUs?
If so... just thinking out loud... That would imply it is possible to convert a factory ME7 MAF application to a "proper" MAFless, right? (Probably not worth all the trouble, but we're not here because we're the kind of folks that leave well enough alone.)
Thanks again!
|
|
|
|
|
Logged
|
|
|
|
|
woj
|
|
« Reply #6 on: July 20, 2018, 10:50:15 AM »
|
|
|
Bosch ME7.9.10 is MAFless by factory, Fiat/Alfa stuff, but I was told here that transplanting its MAP based code to other MEs is a PITA (and I believe that after seeing this code). Answering your earlier question - very many Fiat and family designs are MAFless since like forever, be it Bosch or MM, NA or FI. Seems to be a (good) tradition. Gasoline of course.
|
|
|
|
|
Logged
|
|
|
|
!nfern0
Full Member
Karma: +10/-1
Offline
Posts: 91
Audi TT 1.8T BAM
|
|
« Reply #7 on: July 20, 2018, 02:57:33 PM »
|
|
|
The biggest problem is developing a proper model which shows the dependency MAP to RL. This is IMHO never ever possible without weeks of work on the engine test bench...
|
|
|
|
|
Logged
|
|
|
|
!nfern0
Full Member
Karma: +10/-1
Offline
Posts: 91
Audi TT 1.8T BAM
|
|
« Reply #8 on: July 21, 2018, 01:41:34 AM »
|
|
|
Awesome responses. Thanks, guys.
So i'm taking away... MAFless is a more resilent/robust/efficient solution AS LONG AS your control scheme uses actual measurements for BOTH 1) pressure upstream throttle AND 2) pressure downstream throttle to predict relative load.
Further curiosity... ME7 can accept both of these pressure inputs, right? Are there any "proper" MAFless ME7 factory applications? Does BMW use Bosch ECUs?
If so... just thinking out loud... That would imply it is possible to convert a factory ME7 MAF application to a "proper" MAFless, right? (Probably not worth all the trouble, but we're not here because we're the kind of folks that leave well enough alone.)
Thanks again!
Maybe some more Info... Don't forget that BMW isn't using the throttle plate in nearly all of their engines, at least in normal operation. --> https://www.youtube.com/watch?v=qZ9k4Ohssu8I'm not really sure whether the M engines use valvetronic... As far as I know BMW uses Bosch and Continental for some model years... |
|
|
|
|
Logged
|
|
|
|
Cloudforce
Full Member
Karma: +5/-13
Offline
Posts: 106
|
|
« Reply #9 on: July 23, 2018, 04:33:59 AM »
|
|
|
A 2003 M3 without Turbo and single throttle bodies has nothing todo with a highly boosted me7 car.
Feel free to google TPS or alpha n injection to understand how they work.
Feel free to google BMW M and try to understand how BMW M engines work. |
|
|
|
|
Logged
|
|
|
|
|
nyet
|
|
« Reply #10 on: July 23, 2018, 07:08:44 AM »
|
|
|
The biggest problem is developing a proper model which shows the dependency MAP to RL. This is IMHO never ever possible without weeks of work on the engine test bench...
That model already sort of exists (in the reverse) in ME7 for the ps_w calculation. But somebody would have to make significant rewrites in ASM to make it happen (although they could, in theory, reuse the existing maps). |
|
|
|
|
Logged
|
|
|
|
|
prj
|
|
« Reply #11 on: July 23, 2018, 12:05:59 PM »
|
|
|
It's not that big of a rewrite when you understand how it works.
The only bigger thing is you need to make your own MAP filter, but other than that nothing HUGE. Just need to be clever.
|
|
|
|
|
Logged
|
|
|
|
|
