Printable View
Quote:
Originally Posted by johnsZ06
ROTFLMFAO!
RHS - The MAF is not necessarily a bad thing. Dry shot users need/love it. Other than keeping papers from flying off my desk and deep internet forum discussions of how it works, I have no need for it on my time machine.
So is there a way to run MAF-less with the stock OS and use both spark tables...?
Copy high to low. Otherwise, I beleive the PCM reverts to the low table under a failed MAF condition.Quote:
Originally Posted by joecar
i'm just asking because the way MAF gets talked about by it's opponents makes it look so evil like it's something between Jack the Ripper and public transportation...Quote:
Originally Posted by TAQuickness
Quote:
Originally Posted by redhardsupra
I'd say it's more along the lines of public transportation. It's got it's purpose, but it's not for everyone.
Oh, that's what I do now; I meant without throwing DTC's.Quote:
Originally Posted by TAQuickness
I think some of the contributors to this thread need to back up and get a more realistic perspective on the subject. From 1986 to 1989, GM used MAF-based systems on Camaros/Firebirds and Corvettes. Then they got into a pissing contest with Bosch (the MAF supplier at the time) and from 1990 through 1993, L98 and LT1 engines employed speed density systems. Beginning in 1994, GM reverted to mass air systems. Although each of these systems have advantages and disadvantages, there are an awful lot of MAF-equipped cars that run extremely well, so I don't think it's correct to categorically classify a MAF sensor as "evil". It compensates for a variety of environmental conditions and also makes up for all types of tuning errors. And, all things being equal, anyone would have a difficult time determining whether an engine was mass air or speed density controlled if the evaluation had to be done from the driver's seat with no data logging equipment at hand.
Now, when you start dealing with heavily modified engines, especially those with a lot of camshaft, a MAF sensor is really out of its element. For these applications, a speed density system is clearly more advantageous because the pulsing and reversion that's typical of a long duration camshaft can give a MAF sensor fits. But without programs like FlashScan that tremendously simplify tuning, running speed density wouldn't be a viable option for most people. Building spark and fuel maps with the rudimentary systems that were the only choice just a few years ago was tremendously time consuming.
The simple fact is mass air is superior for some applications and speed density is superior for others. In the end, the BEST system is the one that the individual tuner or car owner is most comfortable with.
Yes, it's all about choices and making the best one for your particular conditions or applications. There's no one right way or wrong way....just whatever way works best for the individual. Having said that, making the right choice means having a full understanding of the pros and cons of each method in order to truly take advantage of the best means to achieve the desired results.
That's what a forum is intended to do and this forum and all who contribute to it make it the great place that it is.
:thankyou2: Now it's time for a beer! :beer:
You ain't kiddin! I got more out this post then I ever expected! It put it all on the table. It seems to me that there is more then one way to skin a cat with respect to tuning and what works for some may not work for others. It is truly amazing how many things play into adjusting fuel and timing.Quote:
Originally Posted by VetPet
:thankyou2:
Oh my God! Relativist tuners.:Eyecrazy:
Bout 101 IIRC, but that doesn't mean the other 100 are correct :muahaha:Quote:
Originally Posted by Redline Motorsports
And when it comes right down to it, it's either 1 or 0 :Eyecrazy:Quote:
Originally Posted by Redline Motorsports
I get the AFR's where I want them in SD mode by tuning the VE. I then tweak the MAF by percentage so that the AFR's match no matter what mode the car is operating in. Most of the time I don't have to change the curve, just add or subtract a percentage of the whole curve.
I also tune Ford EEC's and In the Ford system they rely on the MAF for fuel and load calculations for timing adjustments. For that reason the only correct way to tune a Ford is by MAF adjustments. I use the same prinicples on both Ford and Chevrolet to make sure that the vehicle is thouroughly tuned.
EFILive is great and I use all available tables that I need to make the adjustments that are needed.
Very interesting thread guys. I must say that from a noob's point of view, its very enlightening. Definitley worth a read.
Yeah it is interesting. So if I took thoes little tubes that direct airflow into the airbox I probably just messed with the Maf. I have a truck so nobody probably knows what im talking about, but its a tube gm put in the inside bottom of the airbox.
The latest Hitachi sensor is a marvel compared to the old bare wire clunkers in the foiled, screened housings. The new sensor is shielded from reversion and lots smarter. I talked to the GM PCM guys @PRI and they caution on proper placement in the air stream, but are betting the ranch (and the LS9- check out the MAF table for the 08 LS3) on MAF technolgy to to produce predictable emissions (AFRs).
If you watch your fuel trims trying to second guess a maffless tune, it is indicative of the problem. Have watched my carefull prepared SD tune go to hell in the course of a summer evening at the track. Temps drop 25*, humidity changes, and the tune drifts further from reality. Not to say a "right on" SD tune is not great but, if we all lived in Honolulu it would be a different story, and crowded.
hey Bruce, how can MAF be shielded from reversion? isn't the whole idea behind MAF that air goes through/around the hot element, which requires you to have both ends open? could you elaborate a bit?
A few pics> LINKQuote:
Originally Posted by redhardsupra
I was in the process of tuning my MAF after switching back to a stock op system and the whole feeling of the tune felt terrible. I was throwing various codes, it was getting high KR and in general was running terrible.
There's so much info on here....it's sometimes hard to work out which way to follow.
I asked a buddy of mine who works for GM how they go about setting up the their base tunes and dialling it all in.
This is a reply I just got from him.
Quote:
Checking the MAF frequency and changing the curve to suit should work. I have actually done this once before to improve emissions in an early experimental HSV engine (new air intake completely changed the airflow characteristics). Normally it would be done on a proper test bench with the entire air intake setup - but that isn't going to be easy to do in your case.
A real quick and dirty way to do it is to see how much your fuelling is compensating (for example, -12%), and just pull out 8% or so from the predicted airflow in your MAF curve . If the fuelling heads the right way, then chances are you are on the right track. From there you can fine tune each frequency point as required.
The MAF curve is obtained by testing the entire intake system with the MAF in it, not just the MAF itself. That alone indicates that the intake plays a part in the MAFs response to airflow changes. Otherwise we would just get a curve from the MAF manufacturer and use that.
I'd imagine that the difference is due to changes in airflow/reversion etc around the MAF depending on the geometry of the intake.
As an example: We had a car a while ago that all of a sudden would hardly idle. After some investigation, it was found that a thermocouple had been placed in the intake, about 4 inches in front of the MAF. That was enough to make the airflow so turbulent in the MAF that it was getting airflow swings as high as 50% at steady state.
Basically the MAF behaves like a waveguide with constant+alternating (DC+AC) pulses (waves) going in both directions (interfering with each other to create high/low pressure nodes) and being very dependent on the physical geometry... if you change something in the tract or the motor, then the MAF calibration is wrong...
Thanks Joe..yeah...alot of people say "don't touch the MAF it's a calibrated instrument".
Well for me..this kinda proves it's the (one of) right way to go.
One of the "challenges" of MAF tuning is having the proper "tools" to get it right. As an example, some "high flow" air cleaner systems introduce vortexes in the intake tract that don't exist with a stock air cleaner. So, at any given throttle setting, how do you tell whether a MAF is reading laminar air flow, or a vortex induced by the air cleaner? The same situation may exist when an aggressive camshaft is installed. Reversion and other abnormal (compared to a stock system) intake flow characteristics may leave the mass air meter reading a vortex rather than laminar flow at various points. And these vortexes may change considerably depending on degree of throttle opening and engine load. So the question is, how do you determine what the MAF sensor is actually reading, and whether it's reading consistently at a particular flow rate? I'm not suggesting that it's impossible to make those determinations, only that most people won't spend the time, or don't have the tools to do it. As a result, many MAF tuners have become known as big wheels because they run in circles.
i believe the pcms protocol for fueling is a weighted average of all the fueling tables, if and all the supporting systems are working properly, although some tables play a larger or smaller role in computing the final fuel for a given eng. but. above 4krpm the maf has a major role in fueling and more so than the ve table. also if you install a bigger cam, the shape of the maps will change, but no other table will change its shape more than the ve, especially the peak torque section of the map, the maf should also encrease but the basic curve should remain the same unless the track is modified, for example: removing the maf screen. suppose the maf fails at wot, the pcm must be set up so that there is enough fueling and less timing as a safety measure. if you zero out the ve table the eng. will not run. if you compare 2 stock tunes from two types of cars, each one having different intake track routing and different cams and injectors. you will see that the shape of the ve tables will be different and reflect that specific cam and the maf table will also be different but not in the same way, the change is more subtle. in the end having the correct balance is what its all about or somewhat close to it. just my humble opinion, i have only tuned and owned one muscle car since 04. so i'm probably wrong. still learning.
Quite the contrary, you're undoubtedly right. As has been stated previously, MAF readings are influenced by a variety of factors, and I think the real challenge is in finding a legitimate way to correct MAF readings after camshaft and intake tract changes are made. I think a lot of people cause themselves unnecessary problems because they change MAF tables, when the real problem is the data in the VE or IFR tables. In essence, they're correcting one error that was actually caused by another. At the end of the day, all that really matters is that the commanded and actual AFRs are close enough to be considered the same. So the question becomes, is it better to crutch the VE or IFR tables to compensate for errors in the MAF table or vice versa?Quote:
Originally Posted by jetblast
if you wanna find out what's really going on with MAF, SD and tuning using a wideband, read this:
http://redhardsupra.blogspot.com/200...ss-models.html
TAQ and Marcin and all,
I think what we need to end up with here is an agreed upon MAF "dial in" procedure. There are a few short versions around and some refer to portions of the AutoVE tutorial but for me and others, I think, it is hard to know reference step by step procedure how and where to enter and how to exit the current AutoVE tut. which itself has changed, (thankfully).
B3605 setup in AutoVE should end up @ ~13:1 in what will end up in PE range I assume, and the sum of all the influences in RHS's math models should tune to the 14.7 and ~13 combination that can be tweaked.
After the AutoVE without MAF> MAF dial in> MAF re- enabling> then PE tweaking we seek a happy balanced tune where all the GM guru and tuner work yields smooth safe results. I hope that is what we all want what EFILive is all about.
Some of us are big MAF fans (for various reasons) and mother GM seems committed to them. I for one, feel better if I send a tuned car loose, into the climate extremes we suffer, knowing a MAF (preferably, a big new non restrictive MAF:)) is watching over the tune.
that article seems to imply that dynamics of the operating range of a gasoline engine are infinite and so to arrive at an operating model as good as the gm, ford, chrysler guys when modding, would require using their testing facilities, and so therefore using their model as a reference point is worth some coin, coupled with fuel consumption {mpg} info. a reasonable tune for a modded eng. is then possible. so if you have two identical modded engs. with two different tunes one having a bit more bias towards the maf than the ve, if both have good starting, driveability, wot characteristics, and good satisfactory fuel consumption under various temps and altitudes than i would be satisfied, except for wot which if you go the the track being flexible can gain you a .1 or more on any given day. i'm just a beginer and my perception of how things work maybe wrong.Quote:
Originally Posted by redhardsupra
this writeup says few things:
1. we finally know the exact models for airmass in both SD and MAF
2. know how to know how to convert one to the other
3. the only 'starting point' as far as airmass goes is fuel consumption. unfortunately this is the 'noisiest' airmass model we got, so it's gonna take some interesting math to get the junk out.
and to answer Bruce here...
From where I'm sitting, there's not much to agree on here. if you read my derivations, you know the steps that would go into the MAF calibrations:
1. go pure MAF mode
2. estimate airmass from fuel consumption
3. clean the data
4. convert airmass to airflow
5. fit 3rd order poly to the airflow estimations
6. use the fit to generate the calibration for points needed (since start/end/intervals change on different cars)
or, if you already have a perfect VE table, then just convert one airmass model to the other.
A "simple" matter of starting from the "least" number of "unknowns"...
:crash::chair::bash::bangin: :wallbash:
...are there any other "banging" icons...
Lol! that's me in all of those Joe.
Unknowns????...What are 'KNOWNS"?
:Eyecrazy: :throw: :bawl:
i dont mean to undermine some of the great thinkers here. i have no doubt that my tune is not perfect, but i'm satisfied that i have great fuel economy, i can idle for long periods of time and not foul my plugs, cold and hot starts are normal, winter or summer, and wot, the most fun of all, is giving me great results. now if i could just break 10s!
I think it's as simple as your post #58.Quote:
Originally Posted by hquick
Jetblast,:cheers:
Just one point that I think needs to be plastered somewhere.
You should not and cannot tune VE's with the MAF enabled.
Sorry for yelling!
Yell louder.Quote:
Originally Posted by jfpilla
I think many people do not understand the maf system in these computers. The maf is an accurate way of measuring airflow to an engine but there is nowhere in the software that allows you to tune afrs accurately at every load point.
We only see rpms as the input and hz as the output. The is no tables like we see in a Turbo Nissan where we have a proper fuel map.
That is why the maf table only takes effect after 4000rpm. The throttle butterfly will be all the way open anything above this.
Anything below 4000rpm the maf tables are useless. I can drive on the freeway at 20% throttle and 3000rpms and have low maf readings. Or I could have 80% at 3000rpms and have high maf readings.
I still dont know why GM didnt implement the MAF as the main source of fuel calibration and implenting proper fuel tables.
After all with WOT 4000rpm+ loads the pcm is still using the MAP vs rpm table for fueling calculations.
As discussed already. The only thing we can do is correct calculated frlow rate to actual flow rate.
I know the whole MAF tuning thing is bigger in the states. But speed density is more popular in Aus. I can obtain more accurate WOT fueling with sd as with MAF tuning afrs oscillate at high rpms even after the VE and MAF are both dialled in.
MAFhz is not an output, it's about as much of an independent variable as we got.
as per SD vs MAF tuning, read this: http://redhardsupra.blogspot.com/200...ss-models.html
Frankly, I think all this conversation is great mental exercise, and although it does lead to a better understanding of system functions, most of it is irrelevant. All we really care about is actual air/fuel ratio being as close to commanded as possible. If you're commanding 14.68:1 and a wide band indicates 14.68:1, what else really matters? If tuner A achieves stoichiometric bliss be altering VE and tuner B achieves it by tuning the MAF table, it really doesn't make much difference-- personal preferences aside.
As for "should not, cannot" tune VEs with the MAF enabled I have to disagree. I've found that the VE tables for MAF and SD are significantly different. Try optimizing a VE table in SD mode and switch back to MAF or vice versa and see what happens. If you tune in SD and switch to MAF, you have to make adjustments either to the VE table or the MAF table. As an experiment, I tuned a stock 5.3 in a truck by only altering the VE table until I got the long term fuel trims between 0 and -4%. So if the fuel trims are where I want them, and the wide band is telling me that what I want is what I've got, how does it get any better?
The MAF systems used from 1986 through 1989 were true MAF systems- there was no MAP sensor installed and no VE table. The 1990-93 speed systems were true speed density systems with no MAF included. The LS systems are actually hybrids that use MAF, MAP and a combination of both, depending on the operating environment.
From a theoretical viewpoint, this is certainly not the most elegant solution, but it's relatively straightforward and it works. I'm not saying this is the "be all, end all" of the subject, but I just don't see the payback in spending all kinds of time working with variables that don't seem to do much besides add complexity to a relatively simple situation. If someone has a good explanation with practical application, please post it.
From a practical viwpoint, if MAF is enabled, how do you determine at any instant if airmass is coming from MAF or VE (you want to know which table to apply the correction to)...?
Quote:
Originally Posted by dfe1 http://forum.efilive.com/images/buttons/viewpost.gif
Frankly, I think all this conversation is great mental exercise, and although it does lead to a better understanding of system functions, most of it is irrelevant. All we really care about is actual air/fuel ratio being as close to commanded as possible. If you're commanding 14.68:1 and a wide band indicates 14.68:1, what else really matters?
xxxxxThis must be in SD, so it's fine.
If tuner A achieves stoichiometric bliss be altering VE and tuner B achieves it by tuning the MAF table,
xxxxxThe MAF does not command Stoich.
it really doesn't make much difference-- personal preferences aside.
As for "should not, cannot" tune VEs with the MAF enabled I have to disagree. I've found that the VE tables for MAF and SD are significantly different. Try optimizing a VE table in SD mode and switch back to MAF or vice versa and see what happens. If you tune in SD and switch to MAF, you have to make adjustments either to the VE table or the MAF table. As an experiment, I tuned a stock 5.3 in a truck by only altering the VE table until I got the long term fuel trims between 0 and -4%.
xxxxxNext time you do this disable the MAF and see how badly the car
runs.
-VE's have 3 functions, as far as I know, in an MAF system.
open loop startup.
-Rapid throttle transitions
-MAF failure.
xxxxxxWith your method the MAF better not fail.
So if the fuel trims are where I want them, and the wide band is telling me that what I want is what I've got, how does it get any better?
The MAF systems used from 1986 through 1989 were true MAF systems- there was no MAP sensor installed and no VE table. The 1990-93 speed systems were true speed density systems with no MAF included. The LS systems are actually hybrids that use MAF, MAP and a combination of both, depending on the operating environment.
From a theoretical viewpoint, this is certainly not the most elegant solution, but it's relatively straightforward and it works. I'm not saying this is the "be all, end all" of the subject, but I just don't see the payback in spending all kinds of time working with variables that don't seem to do much besides add complexity to a relatively simple situation. If someone has a good explanation with practical application, please post it.
xxxxxThis is the relativist approach usually argued for by tuners that won't spend the time to do it right. We have tuners that do it right on this forum. I would rather tune it the "best way" possible and avoid the learning out problem .[/quote]
Basically, how do you "correct" some thing based on feedback from output derived from other sources mutually orthogonal to this thing...? :nixweiss:
i've read that volumetric efficiency continues to encrease up to 20k miles and then begins to fall off, the fellows that tune in op loop first,do you take this variable into account?
The proper way to alter the afr for maf calculation is to dial in the VE table first and then scale the maf air flow output to suit.
I would like to see someone do this with an e38 though. With the MAF connected the afr oscillate at high rpm. You can fix this 2 ways. Fudge the maf table even more to correct the difference in fuel mixtures or go to the VE table and correct it there.
I have tuned using solely the VE table with the MAF still connected and also disconnecting the MAF and tuning the VE table first.
I said previously. The delphi computers use solely a rpm vs MAP table to control injector duty cycle. The MAF is only a patch over the top. At full throttle the MAF gives an output reading which then corresponds back to the VE table to obtain the correct figure required. SO why have it there at all.
I understand the VE table is only accurate in the weather conditions it was tuned in and that a correctly scaled MAF will correct any slight changes in air pressure and temperature due to the simple fact that air flow is a much better way of calculating and tuning for optimum conditions than Manifold air pressure.
I believe some earlier models used a MAF vs rpm table for fuel mixtures. No MAP sensor existed. Why didnt GM keep using this type of MAP. The Japanese use it religously on all there performance cars and with the 6.0L having 90mm MAF sensors it would be nice to see a table like this in the later model computers.
Some engineers I studied with a while back always believed the manifold air flow was the best data to collect for accurate fueling.
But all I see is people perservering with scaling MAF tables by a few percent here and there.
Does anyone see my point?