How the MAF really works
#123
TECH Addict
Join Date: Mar 2004
Location: Fat Chance Hotel
Posts: 2,336
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by P Mack
So it takes more heat to raise the temperature of humid air, which should make it cool better right?
The question would remain...how much?
#125
late night thinking
Great thread guys! I'm just getting started in tuning and trying to wrap my head around the whole issue here. Let me throw out some recaps and a couple of theories and let me know what you think.
Given these assumptions:
-flow through the MAF sensor is non-compressible (otherwise stated as compressibility affects are negligible, flow is subsonic)
-reversion affects are negligible (maybe a better way to put it would be changes in reversion affects due to system modifications are negligible)
-geometry of the MAF sensor has not been altered
-changes upstream of the MAF are not causing non-uniform velocity profiles
-mass flow increases (or decreases) do not go "off scale" of the MAF
WHY DOES THE MAF NEED RECALIBRATED AFTER SYSTEM CHANGES THAT AFFECT AIRFLOW?
Thermometer analogy - just because I move to a hotter climate (my engine flows more), doesn't mean my thermometer is invalid (the MAF can't read the flow accurately).
So in a pure sense, it seems intuitive; my MAF calibration should still be valid even though I've changed the system on both sides of it (again, with the given assumptions).
But in practice, we've seen the need to adjust the MAF calibration. Why?
Theory:
We are incorrectly assuming a "DC" type flow. That is, flow at a given TP is steady. In actuality, we are dealing with pulses of air with frequency and amplitude. 1000-6000 RPM = 66-400Hz - intake pulses. MAF is "inherently calibrated" (meaning, I know they are actually calibrated with real flow numbers vs. electrical freq input) to the flow pulses of a given engine - for given a frequency, engine "A" has X amplitude of air pulse. When we modify our engines, we are changing that amplitude of air pulse per given frequency. Which is messing with this inherent calibration. hmmm, having a hard time wording this... maybe some more thinking is needed.
I don't know - it's late and my head is spinning . Maybe I'll wake up tomorrow and wish I hadn't posted this. Well at least it will be good for a laugh!
I think I'm gonna put a carb on my car.
Given these assumptions:
-flow through the MAF sensor is non-compressible (otherwise stated as compressibility affects are negligible, flow is subsonic)
-reversion affects are negligible (maybe a better way to put it would be changes in reversion affects due to system modifications are negligible)
-geometry of the MAF sensor has not been altered
-changes upstream of the MAF are not causing non-uniform velocity profiles
-mass flow increases (or decreases) do not go "off scale" of the MAF
WHY DOES THE MAF NEED RECALIBRATED AFTER SYSTEM CHANGES THAT AFFECT AIRFLOW?
Thermometer analogy - just because I move to a hotter climate (my engine flows more), doesn't mean my thermometer is invalid (the MAF can't read the flow accurately).
So in a pure sense, it seems intuitive; my MAF calibration should still be valid even though I've changed the system on both sides of it (again, with the given assumptions).
But in practice, we've seen the need to adjust the MAF calibration. Why?
Theory:
We are incorrectly assuming a "DC" type flow. That is, flow at a given TP is steady. In actuality, we are dealing with pulses of air with frequency and amplitude. 1000-6000 RPM = 66-400Hz - intake pulses. MAF is "inherently calibrated" (meaning, I know they are actually calibrated with real flow numbers vs. electrical freq input) to the flow pulses of a given engine - for given a frequency, engine "A" has X amplitude of air pulse. When we modify our engines, we are changing that amplitude of air pulse per given frequency. Which is messing with this inherent calibration. hmmm, having a hard time wording this... maybe some more thinking is needed.
I don't know - it's late and my head is spinning . Maybe I'll wake up tomorrow and wish I hadn't posted this. Well at least it will be good for a laugh!
I think I'm gonna put a carb on my car.
Last edited by rck329; 09-21-2005 at 09:41 AM.
#126
8 Second Club
iTrader: (16)
that makes a lot of sens to me...but then again...I understand when you use terms liek Amplitude...because I'm an audio guy.
I think you said one phrase that makes perfect sense..
Its been calibrated to the Stock LS1 as it sits unmodified.
and any modification changes airflow.
an increase in airflow changes the way that it crosses the MAF
alos when you change the size of teh hole(i.e. Porting/descreening) you also change the airflow.
IF you change the Pressure at which the air is moving through it then you are changing the speed at which air is moving through it also as teh size of teh hole remains constant...
so now we have an incorrect reading due to the pressure change.
which is why it needs to be recalibrated for any mod
I think you said one phrase that makes perfect sense..
I know they are actually calibrated with real flow numbers vs. electrical freq input) to the flow pulses of a given engine - for given a frequency
and any modification changes airflow.
an increase in airflow changes the way that it crosses the MAF
alos when you change the size of teh hole(i.e. Porting/descreening) you also change the airflow.
IF you change the Pressure at which the air is moving through it then you are changing the speed at which air is moving through it also as teh size of teh hole remains constant...
so now we have an incorrect reading due to the pressure change.
which is why it needs to be recalibrated for any mod
#127
FormerVendor
Join Date: Jan 2003
Location: Melbourne, Australia
Posts: 525
Likes: 0
Received 0 Likes
on
0 Posts
i think you guys are recovering issues answered in earlier posts to this thread.
MAF is only accurate for unidirectional, laminar, non-turbulent flow. The MAF has been calibrated as best it could be given the conditions available at calibration time. Any modification that affect the conditions under which the MAF was calculated will change its readings. Especially mods that remove things that are designed to promote non-turbulent, laminar flow (such as the screen) etc.
Think of the physical process as a guy standing in the middle of a busy hallway trying to estimate the number of people that are walking past him by the number that bump into him. If anything other than a marching band in perfect formation goes by he is bound to get it wrong, now imagine it is a large kindergarten and someone just hit the fire alarm...
MAF is only accurate for unidirectional, laminar, non-turbulent flow. The MAF has been calibrated as best it could be given the conditions available at calibration time. Any modification that affect the conditions under which the MAF was calculated will change its readings. Especially mods that remove things that are designed to promote non-turbulent, laminar flow (such as the screen) etc.
Think of the physical process as a guy standing in the middle of a busy hallway trying to estimate the number of people that are walking past him by the number that bump into him. If anything other than a marching band in perfect formation goes by he is bound to get it wrong, now imagine it is a large kindergarten and someone just hit the fire alarm...
#128
TECH Fanatic
iTrader: (1)
i wouldn't say that reversion is ignorable. there are few big setups (forged stroker/big intake/big head/cam suitable for even a bigger stroker) that i worked on where the car would work fine in SD, but i could (and did!) spend hours on trying to get it to run in MAF an the car just wouldn't have it. instant stalling while pulling out of the driveway, BAD bucking at cruise speeds, etc. pull the maf and car drives normal.
i traveled from chicago to california in SD, and i have a lot of data on how trims react to altitude/gradients/temperatures etc. I've been trying to get in there and chew through the data, but i've been busy setting up a new life in CA. short preview though: VE is voilitile, fickle, and can change quite a bit fairly quickly.
more on that later, motivate me
i traveled from chicago to california in SD, and i have a lot of data on how trims react to altitude/gradients/temperatures etc. I've been trying to get in there and chew through the data, but i've been busy setting up a new life in CA. short preview though: VE is voilitile, fickle, and can change quite a bit fairly quickly.
more on that later, motivate me
#129
TECH Addict
iTrader: (6)
Originally Posted by soundengineer
IF you change the Pressure at which the air is moving through it then you are changing the speed at which air is moving through it also as teh size of teh hole remains constant...
so now we have an incorrect reading due to the pressure change.
which is why it needs to be recalibrated for any mod
so now we have an incorrect reading due to the pressure change.
which is why it needs to be recalibrated for any mod
#130
Can a dirty maf sensor throw lean bank 1 & 2 codes?? Ses is driving me craaazzzzy. I have been getting a little ping here and there at times also. Does fuel octane play a role?
thanks-
greg
thanks-
greg
#131
9 Second Club
iTrader: (104)
The timing the PCM pulls is part of the reason for you seeing the car go lean. I seen this happen also and I think its because of an incomplete burn when the exhaust valve opens. The WB is reading the O2 content in the raw fuel which is misleading (looks like things are leaning out). Its really not lean at all just an incomplete burn being read as such but it isnt.
Hope this makes sense because at first it didnt to me either and i was expecting the same results as you to my surprise it went the other way
I also tested this theory with no fueling tables changed i pulled -2* of timing and everyhting went 4% lean. Flashed back +2 and everthing feel into place
Hope this makes sense because at first it didnt to me either and i was expecting the same results as you to my surprise it went the other way
I also tested this theory with no fueling tables changed i pulled -2* of timing and everyhting went 4% lean. Flashed back +2 and everthing feel into place
#133
On The Tree
Join Date: Jul 2002
Location: Someplace high in the Himalayas
Posts: 113
Likes: 0
Received 0 Likes
on
0 Posts
Right now we fail the maf and tune the VE table. Then turn the maf back on and assume the error is from the maf. Would it be possible to do it the other way? In HPT, the LS2 GTO is set to use the maf above 3600 rpm I believe. That would mean its blending maf and ve below that. What if you set it to like 700 rpm and tuned tuned the maf curve by trying to maintain steady states of flow(which its good at). Then set it to SD and assume the error is in the ve table.
In trying to tune my VE table for over a year on a pretty stock car, I've about pulled all my hair out. I think part of the problem is there are just too many variables. IAT, MAP, MAF, VE, IPW etc. One thing is for sure, HPT needs to give us a way to filter out transients.
In trying to tune my VE table for over a year on a pretty stock car, I've about pulled all my hair out. I think part of the problem is there are just too many variables. IAT, MAP, MAF, VE, IPW etc. One thing is for sure, HPT needs to give us a way to filter out transients.
#134
TECH Senior Member
iTrader: (7)
set the value to change blending to a low number (500 rpms) and you can tune solely on the maf, without the VE being a factor
I did this years ago to prove it could be done, I set my whole VE table to 1.0 (in the GM engineering units of grams/cyl) without any loss of power, or drivability issues.
If you set the VE to all 0's you will have an error
the measure of success for factory calibrations may be looser than your trying to acomplish
Ryan
I did this years ago to prove it could be done, I set my whole VE table to 1.0 (in the GM engineering units of grams/cyl) without any loss of power, or drivability issues.
If you set the VE to all 0's you will have an error
the measure of success for factory calibrations may be looser than your trying to acomplish
Ryan
#135
Launching!
Would it be possible to do it the other way? In HPT, the LS2 GTO is set to use the maf above 3600 rpm I believe. That would mean its blending maf and ve below that. What if you set it to like 700 rpm and tuned tuned the maf curve by trying to maintain steady states of flow(which its good at). Then set it to SD and assume the error is in the ve table.
#136
On The Tree
iTrader: (1)
Join Date: Jul 2008
Location: Vancouver BC, Canada
Posts: 116
Likes: 0
Received 0 Likes
on
0 Posts
My apologies for bringing this thread back from the grave.
After reading this thread:
- My original belief that as long as you did not change the MAF itself or the air path going into the MAF then the MAF calibration does not need to be updated was/is incorrect.
- The best layman understanding I could derive from this thread as to why post MAF mods still require recalibration of the MAF is that the MAF is calibrated with the engine running at a certain "frequency". By modding the engine the "frequency" is altered. If this layman's explanation is incorrect please do clarify. :-)
- After tuning the VE table and enabling the MAF again, I was surprised to see the fuel trims at 25%. The stock MAF calibration has not been altered since a number of changes were made to the car. 364 ci to 416 ci, long tube headers, different intake, different heads, 230/242 cam, etc. If a MAF is calibrated (expected to be used) on an engine of a certain "frequency" (i.e. stock) then it seems reasonable that a car that has undergone significant modifications would require major updates to the MAF calibration.
I am looking forward to working on the MAF calibration for my car to determine if something is wrong with the car or if the stock MAF calibration is now (after the modifications the car has undergone) more than 25% inaccurate.
Jackson
After reading this thread:
- My original belief that as long as you did not change the MAF itself or the air path going into the MAF then the MAF calibration does not need to be updated was/is incorrect.
- The best layman understanding I could derive from this thread as to why post MAF mods still require recalibration of the MAF is that the MAF is calibrated with the engine running at a certain "frequency". By modding the engine the "frequency" is altered. If this layman's explanation is incorrect please do clarify. :-)
- After tuning the VE table and enabling the MAF again, I was surprised to see the fuel trims at 25%. The stock MAF calibration has not been altered since a number of changes were made to the car. 364 ci to 416 ci, long tube headers, different intake, different heads, 230/242 cam, etc. If a MAF is calibrated (expected to be used) on an engine of a certain "frequency" (i.e. stock) then it seems reasonable that a car that has undergone significant modifications would require major updates to the MAF calibration.
I am looking forward to working on the MAF calibration for my car to determine if something is wrong with the car or if the stock MAF calibration is now (after the modifications the car has undergone) more than 25% inaccurate.
Jackson
#137
TECH Senior Member
The air path between the lid/filter and the heads/valves/cambers behaves more like a waveguide: forward/reflected pressure waves interfere with each other... i.e. think of VSWR... it is more than just "frequency", it is the node/antinode pattern (resulting from the interference of the forward/reflected waves) coinciding with the MAF sensor location, effecting the MAF sensor's response... and also the "period" of the node/antinode pattern changes as engine speed changes.
If you change any of cam/heads/headers/cid/intake/filter/duct then you may be invalidating the MAF calibration.
$0.02.
If you change any of cam/heads/headers/cid/intake/filter/duct then you may be invalidating the MAF calibration.
$0.02.
#138
Banned
iTrader: (1)
Does everyone understand how important a MAF is?
What does the MAF control?
Think about this. Could the engine run off the MAF instead of a crank signal? like if a crank sensor failed. It would still have a cam signal so it could fire injectors. The PCM can still calculate load correct?
How much does the Trans use the MAF for?
What does the MAF control?
Think about this. Could the engine run off the MAF instead of a crank signal? like if a crank sensor failed. It would still have a cam signal so it could fire injectors. The PCM can still calculate load correct?
How much does the Trans use the MAF for?
#140
TECH Senior Member
Very good points being made...
It's not important.
It provides one means for the PCM to compute cylinder airmass.
The motor can still run without the MAF, the PCM would use the VE table to compute cylinder airmass.
It doesn't use the MAF at all; from cylinder airmass the PCM computes engine torque which it uses to compute/lookup the trans "throttle signal" pressure (which hydraulically/mechanically modulates line pressure).
+1.
It provides one means for the PCM to compute cylinder airmass.
It doesn't use the MAF at all; from cylinder airmass the PCM computes engine torque which it uses to compute/lookup the trans "throttle signal" pressure (which hydraulically/mechanically modulates line pressure).
+1.