[steve341]

A friend of mine wrote this article. It's a very informative read. Enjoy.

MAF basics



Some basic info about MAF and MAF transfer curves.


Understanding a mass air meter and why it so critical to have the correct meter curve in a tune.

This is where 99 percent of the tuners get themselves in trouble and it all starts from step one.(The Mass Air Meter and Meter Curve) Lets not even worry about the rest of the tune (that's a completely different story)the first and most important step all starts here.
The MAF meter is the first point where air enters the motor and every calculation and I mean every calculation the computer does will depend on that being an actual and real measurement. (Load, Idle, part throttle, WOT, A/F, timing, torque calculations, etc. all depend on it)


The main purpose of a Mass air meter is to measure airflow.
The Meter operates in a range from 0-5volts
5 volts being the highest it could read.

Every voltage point from 0-5volts for a given mass air meter and sensor represents an actual air flow measurement.

So if you put a MAF meter housing with a sensor on a flow bench you can obtain a mass air transfer function for that given meter sensor combo.(A mass air meter will only work correctly/measure correctly if these values are accurate)
So lets say this is the data you obtained after flowing the meter on a flow bench.(And by the way this is exactly how Ford does their mass air meters and curves)

example
0 volts = 0 airflow
1 volt = 500 CFM
2 volts =800CFm
3 Volts =1200 cfm
4 volts = 1500 cfm
5 volts =1800CFM
every voltage point in-between those valves will also equal a CFM valve somewhere between them .(Just using whole numbers for this example.)

If you needed A MAF meter to measure more CFM, there are two ways to do it correctly.
One, use a different sensor that has a different range in the same housing.
Two, use the same sensor but increase the size of the housing

Another note, since CFM is a direct relation to HP this is how a computer can/will calculate the torque the motor is making at a given RPM.(one way how torque management works/is calculated)

It takes 1.5 CFM to make 1 hp

Knowing this, you can take a mass air voltage/reading from your MAF and calculate how much HP your engine is making.(But again only if you know the meter and meter curve is accurate)
Believe it or not this is how top fuel teams figure out how much HP their engines make(Hard to dyno an engine making 5000hp plus HP, no dyno can really read that high, until maybe the last couple of years) They measure how much CFM the engine is moving and then calculated that into HP


Now going back to one of my first statement's that 99 percent of the tuners get themselves in trouble here and cause all kinds of issues ranging from bad startups, pinging, drivability issues, idle issues etc.
Here is why

Most tuners out there were TAUGHT.... to adjust the A/F ratio by moving the transfer curve in the tune. (Will it change an A/F ratio it sure will...........) They really don't understand how a mass air system works. (If they did they would NEVER do this) It’s really a shame because most tuners feel this is the correct way to tune (they really believe they are tuning correctly).
Why ... again because this is the way they were shown. Since they were shown that, it has to be the correct way. They were never actually shown or understand what that mass air meter is doing or suppose to do.

They were also taught to create a mass air curve from scratch or adjust a curve by data logging short terms or using an A/F gauge to tweak or enter a value in the curve depending on the short term voltage or the a/f ratio measured coming out the tailpipe.
They will move a curve up or down until they get the a/f ratio they are looking to obtain.

So let’s see why this is the wrong way of doing things
Lets look at the example data above created from a MAF metering using a flow bench.

Let’s take the 2 volt measurement which equals 800 CFM (which we know is a real and actual air flow measurement for this example):

We are driving our car and the mass air meter is at 2 volts.
OK, the computer will take that 2 volts reading and start calculating how much to open the injector to hit a target A/F, referencing the air fuel ratio target map. (If the injector slopes and all are correct in the tune and depending on your fuel pumps/system etc. Whatever you have in your target a/f will be what you will obtain out the tailpipe)

It will also look/calculate load, timing maps etc. from this measurement but lets just look at the A/F ratio here.

If for some reason (and can be many) the target a/f is suppose to be 12 to1 and 10 to 1 is coming out the tailpipe.

What the tuner will do instead of adjusting the correct parameters they will take the 800 CFM value at 2 volts and make it a smaller CFM value/ number in the MAF curve to trick the engine in thinking its moving less air.
If the engine thinks it’s moving less air it will calculate the injector not to open as much delivering less fuel, which will cause that 10.1 A/F to hit the 12.1 A/f that is targeted.

So what did they do and how they tune, is by looking or only caring only about the A/F ratio.
Every other calculation done at that point the computer does will be wrong..... Why ....................because it thinking its moving less air then it really is now!!!!

This is a real basic and simple example.
I hope this helps your understanding of why and how important it is to really know what that mass air meter is measuring and what the MAF curve does in the tune.

A few years ago it was more forgiving when the drive by wire didn't exist to tune by moving or creating a MAF curve the wrong way.

Drive by wire and or highly modified cars will only amplify issues when a car is tuned with an inaccurate MAF or MAF curve.

There are many more other unintended negative affects that will occur as a result of a tuner who changes a known MAF curve that has been established on a professional flow bench,( like the one Ford uses) for the purpose of adjusting air/fuel but I’ll leave that for another day.



Just look in the 2011 Ford racing catalog similar statements online catalog

http://www.fordracingparts.com/2011-catalog/

Read page 154 bottom paragraph page 220 explains a little about the MAF and its calibration

[85GT]

:thumbup:

[garyd85gt]

so what are we supposed to change?? Is it not conceivable that by adding ram air, a large air filter, large diameter piping and a supercharger pulling more air through it, that the MAF flow numbers will be off???

[steve341]

If you have inputted the maf transfer function correctly, the meter will measure the proper amount of air until it runs out of range.

Just a question: Would you recalibrate your oil pressure sending unit to create a reading that you were looking for?

[85GT]

so what are we supposed to change?? Is it not conceivable that by adding ram air, a large air filter, large diameter piping and a supercharger pulling more air through it, that the MAF flow numbers will be off???

You work on the fuel.

[cgrey8]

I have a few issues with this article, and I'll point out the things I find issue with, why, and my opinions on the subject.

...example
0 volts = 0 airflow
1 volt = 500 CFM
2 volts =800CFm
3 Volts =1200 cfm
4 volts = 1500 cfm
5 volts =1800CFM
every voltage point in-between those valves will also equal a CFM valve somewhere between them .(Just using whole numbers for this example.)

I understand this is just an example to give the idea, but even examples should use the correct units of measure. CFM is a volumetric measurement. MAFs measure mass, not CFM. Common measures of mass for MAFs are in kg/hr, or sometimes in lbm/min. Either of those can be converted to CFM given an air temp. But because the disconnect between mass-flow and cfm flow is temp, it is NOT correct to insinuate that a MAF measures CFM. It is almost as incorrect to say Widebands measure AFR. They do not. But that's another subject covered in the Tech Docs here:
Widebands DO NOT measure AFR.

...Most tuners out there were TAUGHT.... to adjust the A/F ratio by moving the transfer curve in the tune. (Will it change an A/F ratio it sure will...........) They really don't understand how a mass air system works. (If they did they would NEVER do this) It’s really a shame because most tuners feel this is the correct way to tune (they really believe they are tuning correctly).
Why ... again because this is the way they were shown. Since they were shown that, it has to be the correct way. They were never actually shown or understand what that mass air meter is doing or suppose to do...

I disagree with this claim. As someone else already said, if you don't adjust the MAF, what do you adjust? The main things in a tune to adjust AFR are the MAF curve, Injector Parameters (Slope/BP/Offset), OL fuel tables, and the MAF backflow table. Granted any one of these could be off and many may need to be adjusted, but when you are working with an engine that won't run or runs horribly, you often feel like you are trying to move a king size mattress by yourself. The thing flops all over the place. So you keep it simple just to get a handle on things. Enter Injector and MAF parameters as best you know them and adjust as needed. If that means tinkering with the Injector parameters, you do that. If that means gross changes to the MAF curve, then you do that just to get the engine to run so you can get better data without washing the cylinder walls with fuel. But just because you got it running doesn't mean the values you entered to get it running were correct and should stay that way. Often make-shift values that don't make sense to a DIY tuner need to be revisited. And with more info from datalogs and research, you can often find what it was that was messing things up, and return the make-shift values to more sensible values later. But this is often an iterative process. And to the inexperienced, I admit it is EASY to forget what got changed and what didn't early on in an effort to just get the engine to run. This is exactly why so many DIY tuners tweak-n-tune, learn, tweak some more, then revert back to a stock tune (wiping away history) and re-entering the good info they've learned just as a sanity check. Doing that, you quickly find what parameters were pertinent while reverting parameters that may have gotten changed in the throws of desperation-tuning.

Experienced or not, a tuner does need to keep their tweaks in check. Minor adjustments to the MAF curve are not disastrous and in many cases are appropriate. After all, the way a MAF flows in the engine bay with an engine aspirating air through it and the way it flows in the controlled conditions of a flow bench with steady flow could be different hence the common need to make MAF adjustments. But probably the more common reason for MAF curve massaging is due to differences in intake piping that flows the air through the MAF differently than the MAF was flowed with on the bench. It's easy to forget that the sensor is only at 1 position in the MAF flow region and flow through the MAF is not always constant at all locations. If you doubt that, clock an edge-sensing MAF 90° in various directions and see if your tune behaves exactly the same. Slot MAFs or MAFs that locate the pickup in the center of flow will be less affected by clocking hence why they are better. Another example is take an LMAF and install it in a highly turbulent blow-thru application and then try to tune that.

Regardless of what the reason is, to assume that a MAF is dead-on accurate in all possible applications and conditions once it has been flowed is a BAD assumption...IMO.

...They will move a curve up or down until they get the a/f ratio they are looking to obtain.

So let’s see why this is the wrong way of doing things
Lets look at the example data above created from a MAF metering using a flow bench.

Let’s take the 2 volt measurement which equals 800 CFM (which we know is a real and actual air flow measurement for this example):

We are driving our car and the mass air meter is at 2 volts.
OK, the computer will take that 2 volts reading and start calculating how much to open the injector to hit a target A/F, referencing the air fuel ratio target map. (If the injector slopes and all are correct in the tune and depending on your fuel pumps/system etc. Whatever you have in your target a/f will be what you will obtain out the tailpipe)

It will also look/calculate load, timing maps etc. from this measurement but lets just look at the A/F ratio here.

If for some reason (and can be many) the target a/f is suppose to be 12 to1 and 10 to 1 is coming out the tailpipe.

What the tuner will do instead of adjusting the correct parameters they will take the 800 CFM value at 2 volts and make it a smaller CFM value/ number in the MAF curve to trick the engine in thinking its moving less air.
If the engine thinks it’s moving less air it will calculate the injector not to open as much delivering less fuel, which will cause that 10.1 A/F to hit the 12.1 A/f that is targeted.

So what did they do and how they tune, is by looking or only caring only about the A/F ratio.
Every other calculation done at that point the computer does will be wrong..... Why ....................because it thinking its moving less air then it really is now!!!!

I disagree with this, but only because in practice, it doesn't matter. The assumption here is that the stock tune is going to be 100% correct for all possible engine combos out there assuming the MAF and Injector Parameters are correct and that the only thing you have to do, as a tuner, is get the AFR correct. In practice, that's just not the way it is. All the things that would be affected by Load not being exactly perfect due to MAF curve tweaks will likely need to be tweaked anyway. So regardless of whether the Load is 50% but the EEC is calculating the condition to be 60%, the spark table for that engine condition is likely going to be reworked on the dyno/track/road anyway. The Load value associated to that engine condition is irrelevant to the engine as long as the tune is giving the engine the right amount of fuel and spark for that condition. That's all that matters. Now it is much nicer when the computer is closer to actual because that usually results in less tuning. And I think that's the goal of most tuners...get-r-dun, so the comments have merit. But it is not the end of the world to have the computer and reality not 100% in sync in all aspects as long as the important stuff (delivered fuel and spark) is in sync.

So how do people adjust for AFR changes? The two major camps of thinking when it comes to getting mix correct. One is to tweak until the commanded lambda (you can think of lambda as AFR) matches the measured lambda. The other is to do whatever it takes in the tune to get the measured lambda correct for the engine running condition. As long as actual is on target for the engine condition, it doesn't make a hill-a-beans difference if the commanded lambda is too high or too low.

Regardless how correct AFR is gotten, being worried that MAF curve mods are going to throw off Load values is not really a big concern if you'll be visiting the rest of tune anyway. And in the case of boosted applications, every single one of these tables is going to get touched regardless due to table rescaling. So unless the tuner has a specific need to have the MAF perfectly accurate, worrying about whether the MAF is +/-10% of actual is hardly a concern.

Now that being said, there does come a time when adjusting the MAF by excessive amounts is an indication of a problem. For instance people that have to raise or lower the entire curve by X % most likely have some other major parameter wrong in their tune such as bad slope values or the wrong MAF curve for the MAF they have installed.

But this does lead into another topic that is not talked about here and that's why using "calibrated" MAFs is simply not a good idea. But that subject is discussed here:
Is having a MAF calibrated to your size Injectors important?

...Drive by wire and or highly modified cars will only amplify issues when a car is tuned with an inaccurate MAF or MAF curve...

I agree that newer vehicles get more and more complex making simple tuning changes, not-so-simple. Add transmission control to the mix, and now you have to retune shift points and line pressures or you burn up trannys. And as you get more and more advanced with VVT, Direct Injection, and full-time EGR (even at WOT), things will get a lot more difficult for the average DIY tuner to deal with. If people are intimidated by DIY tuning on these old 90s era EECs, check out the tune controls available for a 2011 GT 5.0 computer. You damn near have to make a career out of learning these newer EECs to be able to deal with them correctly. Another caveat that is different today than of the 80s and 90s is the factories are tuning closer and closer to the actual engine limits. It used to be a chip could gain you 20hp on a stock engine without too much trying. Today, dyno tuners struggle to find 5hp that the factory left out of the tune.

Bottom line is my attitude towards tuning is results are all that matter. If the engine is getting the fuel and spark it needs, and getting it consistently at the appropriate times, how the tune attained that is not nearly as important. So if a tuner has to modify a MAF curve to get the desired results, so be it. That's my take on the subject. Others feel free to chime in with their opinions. But since this is more of a hardware discussion at this point, I'm moving the thread.

[garyd85gt]

""Experienced or not, a tuner does need to keep their tweaks in check. Minor adjustments to the MAF curve are not disastrous and in many cases are appropriate. After all, the way a MAF flows in the engine bay with an engine aspirating air through it and the way it flows in the controlled conditions of a flow bench with steady flow could be different hence the common need to make MAF adjustments. But probably the more common reason for MAF curve massaging is due to differences in intake piping that flows the air through the MAF differently than the MAF was flowed with on the bench. It's easy to forget that the sensor is only at 1 position in the MAF flow region and flow through the MAF is not always constant at all locations. If you doubt that, clock an edge-sensing MAF 90° in various directions and see if your tune behaves exactly the same. Slot MAFs or MAFs that locate the pickup in the center of flow will be less affected by clocking hence why they are better. Another example is take an LMAF and install it in a highly turbulent blow-thru application and then try to tune that. ""

Wow, this is what I was trying to say, just wayyy more eloquently done. Thanks cgrey, I am one who had to change MAF at beginning just to get car to run, was way lean with stock meter and injector readings. I tried to change the injector slopes, even using 30# slopes for 42# injectors, but that didn't really change anything. Now my slopes are good, but my MAF curve is 10 to 15% higher then the stock 90LMAF curve and it runs great....

[BIGDATOWN]

You can also send your mass air sensor and housing to c&l for a small fee they can give you all the exact data to enter into the maf scalar. Alot less guess work also.

[cgrey8]

Don't get me wrong, starting from a curve that was found on a flow bench is certainly better than using a "canned" curve. Although even with flowed curves, I don't think it is unreasonable to expect that some parts of the curve may need to be nudged. Although once the tune is "done", the final curve and the original curve should be obviously very closely related. If they don't look similar, and instead one looks like it is "peeling" off the other where they are getting further apart as the flowrate goes up, then that's often a tell-tale sign that the Injector High slope is off.

When you compare my tune's MAF curve to the published C&L Curve for a 73mm MAF Blue Tube 55mm electronics listing in EA, you see they are very similar, but not exactly the same. Check it:

MyCurveVsOriginalCurve.jpg

My curve is Green.

(51.12 KiB) Downloaded 5976 times

Notice my points don't align on the same voltages as the original curve. That's because I redistributed them into the ranges of the MAF curve that I use. The MAF is slightly oversized for my engine, so I never see voltage over 4.1v. So I took some of the points that were up above 4.1v and moved them down into the lower parts. I also put a tighter concentration of the points towards the cruising flowrates where I wanted more accuracy and control. That'll have to change once I get the new 331 in there since it'll be capable of flowrates that'll have the MAF in the upper 4v range for sure. But to the point of this thread, the places where the two curves are furthest apart probably don't throw the Load off more than a few % if that. And for that little bit of error, I'll gladly modify the MAF curve when that makes things simpler to get the engine running better.

[cougarnut281]

The original post totally ignores the inlet tract the maf is plumbed in. Bends, wripples, screens, etc. I have seen a swing of over 1 volt just by rotating the maf before. So since the motor is moving the same amount of air at a constant idle speed what voltage (air flow) is the correct one for the EEC calculation?

The way I do it is find the most stable reading and then readjust my Maf curve to suit. My load values always seem to fall in line when I am done so I can't be doing it all wrong.

[steve341]

Here's another part of the article:

1) ...only the MAF tube with the sensor inserted needs to be flowed. Anything upstream (e.g. filter) only affects *how much* flow the engine sees but does not affect the accuracy of a properly flowed MAF across it's flow range. Unfortunately most tuners use existing curve functions (possibly with other sensors) and then scale/adjust them for different sensors/MAFs based on target A/F’s ...that's where the problems start since everything downstream in the tune is then effectively 'corrupted' ...garbage-in/garbage-out.

2) ...companies like PMAS (Professional Mass Air systems) have the ability to natively flow (no scaling) anything these engines can handle -- they do it for pro racers all the time. Only the meter-in-tube needs to be flowed. Then the [mathematical] curve that relates that flow range to the 0-5v ECU voltage range can be created. That accurate mapping of low-voltage is the key and is what's called the Mass Air transfer curve/function -- it's a tiny piece of the overall calibration but affects so much because nearly *everything* in a mass air managed engine depends on that curve being 100% accurate to determine when and where it dynamically pops into/out-of many, many control-calibration tables (and the paths it takes through them) as you run our engine.

3) Only the MAF gets calibrated -- every tuner subscribes to using the mass-air transfer function because the engine simply won't run (at all) without it and will run poorly if it's just off a bit off from true/accurate flow. The problem is *how* they use it. The *only* correct/proper factory-engineered way is to base the curve's calibration on dead-nuts accurate flow -- and then it *never* gets changed for that MAF sensor and MAF tube combination (i.e. the MAF). The improper way is to get it close enough (sometimes through trial and error and sometimes based on scaling/estimating working off an existing curve/sensor -- yikes!! -- really!) to get the engine to run and then tweak the short-terms to fake it out to hit 'target' A/Fs, with the result of inducing inaccurate 'pathing' through those many, tables that are especially critical at WOT (where being off a couple % on a table-pop can be the diff between nice-engine and scrap-metal), but also affect closed-loop cruise, idle, etc.

This dangerous tweaking practice grew out of a long legacy of 'approximate' tuning with it's roots back in the days of Speed-density engine management (an widely-used but inferior pre-cursor of Ford's Mass-air engine management) back when Ford's calibration methods were still being decifered/hacked by many racers. While those days are mostly gone, the 'addiction' of tweaking has lived on -- I must conclude at this point it's 10% history and 90% interlock of the customer to the tuner. Why? When the MAF is flow-accurate, many mods can be swapped and none of them will affect the MAF so long as it still has meter capacity. When the MAF curve is 'tweaked' to a given set-up virtually any change requires re-tweaking (just to get back to a running, but inaccurate, calibration) ...if you were a tuner competing for customers and repeat business, you might be inclined to 'play' the characteristics of the latter -- ya think?

[steve341]

The original post assumed that one has done his/her homework regarding the engine build, fuel system, air intake plumbing, proper voltage coming from the alternator and/or maf location are all done correctly. All vaccum and exhaust leaks are corrected. Fuel is proper. Grounds are of factory caliber.
Ford added a screen in front of the mass air meter to smooth out the turbulance before the air got measured.

[decipha]

who is the author of this publication?

they figured out the same thing i did years ago, i feel the exact same way, with a known maf curve you can adjust your injector parameters and get your fuel 100% dead on

i've tuned countless vehicles this way, even stock vehicles that just needed fuel dialed in i adjust the injection and get the kam's to stay at 1

its fast accurate and the only adjustments needed to transient fuel and inferred load is at the very most top range

[cgrey8]

It seems this whole topic only proves there's more than 1 way to skin a cat. To each his own how they go about it.

[decipha]

the biggest problem with raping the maf curve is that inferred load gets screwed, and if anything ever happens your dead on the side of the road, where as using the injectors to dial in fuel, your inferred load is almost dead on balls, only adjustment needed is at the top end usually WOT where the engine is more efficient, on a stock engine inferred load is perfect and everything works out perfectly as well

this isn't just my method of tuning, this is the preferred method of tuning nissan's as well, direct relation would be the k_constant which is a global injection multiplier (IE injector slope)

[cgrey8]

People talk about messing with the MAF curve like making any change to the MAF curve is going to throw the Load calculation completely off. Yeah if you multiply/divide the entire curve by some fixed amount (+/-10 to 20%), then sure Load will be off. However if you are making small changes, the change in Load will be a few %, if that. It certainly isn't going to be enough to throw the whole tune into turmoil. Look at comparison of my curve to the original curve above. I don't think there's enough difference in those curves to throw the Load off any significant amount at all. Although the difference in the curve is the difference between good driveability or some leanout.

[liljoe07]

I've found the same thing chris is talking about. I've made small changes to the MAF curve around idle. And it never moves from 19-20%Load.

[steve341]

I posted the MAF basics article because too many people don't understand the mass air meter's function. Adjusting the transfer curve to meet an acceptable AF ratio DOES screw with the load tables. Can you get an acceptable AF at WOT? Sure you can. But I can guarantee you'll wind up with drivability issues. Idle will be irratic. Stalling coming to a stop. Partial throttle issues. It's all related to the maf transfer function.

Start with a flowed mass air meter and input that transfer function table into the program and you are on your way. Why one would consider avoiding this fact is beyond me when this is how it was designed. Whether you like it or not, ALL calculations are taken from this table and simply moving the curve is telling the PCM that the meter is flowing more/less air than it actually is. Would you recalibrate your oil pressure sensor to read X when it's flowing Y? Then why would you move the curve?

From the transfer table and a data log, you can very quickly tell how much air is flowing and how much hp you are making that day. As soon as you fool with that curve, you have NO clue how much hp you are generating not to mention, on a boosted car, how to get back to times you ran on a day with better weather.

Here's another little tid bit I'll give you....Ford PCMs pre '05 only accepted hg/hr values under 1739 hg/hr. So on a boosted car with a maf meter that is flowed double that value, you must cut it in half. To match that, you must also cut in half the displacement AND the injector size.

You may disagree all day but fact is fact. Touch the transfer function table and you will never get the tune 100%. Ford doesn't do it that way and neither should any tuner. A quick read in the Ford Racing catalog will tell you not to move it. And a call to them will net the same result. But this is where many tuning instruction booklets(not Ford) tell you first and foremost to move the transfer curve to adjust AFR. Is it faster this way? Sure it is. Will you be chasing gremlins forever? Absolutely.

Of course, the disclaimer is that the installation of all hardware needs to be as perfect as possible. If it isn't correct, then your tune will always have an issue.

[cgrey8]

...Can you get an acceptable AF at WOT? Sure you can. But I can guarantee you'll wind up with drivability issues. Idle will be irratic. Stalling coming to a stop. Partial throttle issues. It's all related to the maf transfer function...

There are quite a few people here on this forum that would disagree with this. There are MANY people that are quite happy with their driveability and have touched the MAF curve in some form. So to guarantee that problems will arise if the MAF is touched is probably a bit presumptuous.

...Start with a flowed mass air meter and input that transfer function table into the program and you are on your way. Why one would consider avoiding this fact is beyond me when this is how it was designed. Whether you like it or not, ALL calculations are taken from this table and simply moving the curve is telling the PCM that the meter is flowing more/less air than it actually is...

I still argue that the way a MAF flows on a bench and the way it flows in an application, particularly a less-than-optimal installation, will be different and that has to be taken into consideration. As mentioned above, the way a MAF flows when installed straight up can be different than when installed at an angle or on its side (clocking the MAF). And I can say, 1st hand, clocking a MAF does affect its behavior. I've done it, I've seen it, and on my application, it is not a small change. It is significant the difference in MAF reading. But I also have a very non-optimal setup. If I had more space in the engine bay, I'd have a much better intake track than I do for the MAF. If the MAF were absolutely perfect in every possible configuration, there should be zero difference in MAF behavior despite installation. But anybody that has done and read about tuning knows that no sensor is infallible and the installation most definitely influences the sensor's accuracy and behavior.

I don't disagree that any changes to the MAF affect the Load calculations. But put into perspective, small changes to the MAF are going to be even smaller small changes to the load calc. And if you are touching all the other things load influences anyway (spark, EGR, idle controls, etc), then my argument is as long as you are only making minor changes to the MAF curve to get the AFRs dialed in, it won't matter. Gross changes to the MAF curve should be avoided for exactly these reasons. And in fact, I would say if gross changes are found to be required on the MAF to get the desired AFRs then that's likely a sign that something else is wrong...either in the tune or in the installation.

...From the transfer table and a data log, you can very quickly tell how much air is flowing and how much hp you are making that day. As soon as you fool with that curve, you have NO clue how much hp you are generating not to mention, on a boosted car, how to get back to times you ran on a day with better weather...

Lets think about this a little harder. If I changed the MAF curve by some minor amount and the difference between what the MAF would've read and what the MAF reads with edits is 10-30kg/hr, that's maybe 1-2% difference at WOT on most people's setups. To say you have no clue what the HP is after making that change is just misleading. You still have a ballpark idea of what the HP is. But even if you had the exact MAF curve and had 100% confidence in the MAF's ability to deliver an accurate flow value on each and every sample, you still wouldn't know exactly what the HP output is without a dyno run to tell you what the engine's efficiency in Brake Specific Fuel Consumption (BSFC) ratio is that day, in that humidity, with that tank of gas. You could have 5 different vehicles aspirating 1000kg/hr, but each making different HP values to the ground. Knowing that an engine is aspirating 1000kg/hr just gives you a ballpark idea as to what power should be. But it is far from an estimate of exact HP unless you have extensive knowledge of the engine's performance behavior from numerous dyno/track runs. So for the average DIY tuner, if the actual reading happened to be 1020kg/hr because of a MAF tweak, that's still in the same ballpark range as if it had been 1000kg/hr.

...Touch the transfer function table and you will never get the tune 100%...Is it faster this way? Sure it is. Will you be chasing gremlins forever? Absolutely.

Of course, the disclaimer is that the installation of all hardware needs to be as perfect as possible. If it isn't correct, then your tune will always have an issue.

I get nervous when people give blanket advice that includes words like never & always. Thus I feel the obligation to post alternate opinions or way of looking at the problem. There are few absolutes in DIY tuning and I don't believe that not touching the MAF curve is one of those absolutes. The MAF curve is one of those areas of the tune where the people will need to decide for themselves what's right based on their understanding of the problem(s) and the data collected. After all, they are the ones that have to live with the results. I just want the topic presented to others to be kept in perspective so they can better make a decision for themselves what's right for their application. After all, this thread advocates for never touching the MAF curve. Other threads advocate for never touching the Injector Parameters. This is the state most DIY tuners begin at anyway. If this plan is so great, then there should never be fueling issues. But there almost always are on most modified applications where MAF, Injectors, and major engine mods have taken place. So what methods of AFR adjustment would you recommend? So far, all I've heard is rhetoric telling people what not to do.

From an inexperienced DIY tuner reading the forums and trying to get an idea as to what to do and where to start, they will probably read this and think...If I can't touch either MAF or Injector parameters, then the only option left is to simply poke in whatever values into the Open Loop fuel tables to get the target AFRs even if the commanded lambdas are not consistent with the actual measured lambdas despite the intuitive desire to see their tune delivering what it is commanding. So short of now messing with obscure controls like the MAF Backflow table, which is almost the same as messing with the MAF curve, there aren't many more AFR-related adjustment options left. And of course, this only covers for Open Loop conditions. How do you tune for Closed Loop issues? I guess one option is to pretend any fuel-related driveability issues aren't there or have faith that one year the computer will eventually learn them away, while reminding yourself you didn't change the MAF curve or Injector Parameters. While this might work for some people, it doesn't work for everybody. In some cases, adaptive learning actually causes problems...at least early on in the tuning phase.

It's no surprise that the most common action for DIY tuners is to ask questions, get datalogs, analyze their data, ask questions about the data, and eventually determine/guess if the problem is in Injector parameters, MAF curve, or some other area of the tune. Make tune changes based on their analysis of data. Test the changes. Repeat until desirable behavior is attained. That's pretty much the formula of the DIY tuner. A professional tuner can often get the job done in far less time. But part of the desire to be a DIY tuner is the learning and the experience. NOBODY should get into DIY tuning expecting it to be a money saver.

[sailorbob]

The OEM MAF transfers are calibrated for the induction system they are installed in. This can be seen by comparing the transfer functions from two different vehicles that use the same MAF sensor. The differences are not huge but they do show how the installation affects the reading.

[cgrey8]

The OEM MAF transfers are calibrated for the induction system they are installed in. This can be seen by comparing the transfer functions from two different vehicles that use the same MAF sensor. The differences are not huge but they do show how the installation affects the reading.

Proof straight from Ford that the MAF curve isn't absolute.

[Cougar5.0]

So I agree with the basic premise of the article.

I'll just post a few bullet points to make it easier to read:

• Getting the MAF spot on is especially important on newer vehicles - as mentioned above. So Fox body people who are getting anxious, chill out a bit. However;
• Tweaking the high-end of the curve to dial in AFR will not effect idle/cruising/drivability at all. Shifting the entire curve up or down is not generally a good idea unless it's your first adjustment after installing a MAF with a known flow curve into your plumbing. I used to shift the curve up top for WOT AFR issues and blend it to rest of curve. Now I simply use FN303 as Ford did to tweak AFR at WOT. The MAF curve is never touched and I can easily adjust AFR to compensate for H2O/meth injection and a shot of N2O. If you have to tweak your curve to accommodate your bends etc., do it once and then never touch it again. Remember though, you want to be as sure as you can be that your injector slopes, breakpoint and offset are accurate before you touch your MAF curve.
• Getting the lower part (idle/cruise) of the curve accurate is much more important, so I agree with the article here. Many times people will mess with the MAF curve down low to fix Slope, Breakpoint or Offset issues. This is a bad practice for the reasons described in the article. Besides, I prefer my fuel mileage calculations in the trip computer to be accurate and that is only possible if those parameters are accurate. If you can get your setup flowed then do it. If not, have a straight shot of at least 1 diameter of your tubing before and after the MAF.
• Flow curves can be scaled based on MAF diameter, but you have to have a uniform flow in the MAF to ensure accuracy (i.e., no bends right before MAF.) Even then it's complicated. I used a curve fitting tool to create a 10th order polynomial equation to scale my slot MAF. If you want to see how this works, Google "HPX tool".
• This little "tidbit" is dated information - "Ford PCMs pre '05 only accepted hg/hr values under 1739 hg/hr." We now know how to "max out" FN037 to get MAF values as high as 2500 kg/hr @ 6000 RPMs (notice the FN037 limit of about 1740 kg/kr @ 6000 RPM.)
  

So while I generally agree with the O/P and went to great lengths to ensure that my slot MAF was starting at a known accurate place based on the Ford OEM curves, let's not forget that much of the aftermarket plumbing on a lot of the cars we're tuning here effects the MAF transfer function. And many aftermarket MAFs do have questionable/non-repeatable flow curves. And unless you know your slopes etc. very accurately, you've got more potential error (BTW, I used the data for the 42lb. Green Tops directly from Ford's datasheet and it's spot on.) So I wouldn't get too caught up in this if you have already dialed your setup in. If you are having issues and aren't sure about your MAF, getting a known good reference wouldn't hurt though - a good MAF is as critical as the O/P suggests, though even more so in newer ECUs.

[decipha]

i've been saying it for the longest time, with a flowed maf you can simply put those values in the maf transfer and dial in your fuel using your injectors, this is the way i tune, it works perfect everytime and with a known curve you can ALWAYS get the kam's to stay at 1 in closed loop

you then don't need to spend as much time dialing in fn1036a, which is very time consuming if you rape the maf curve

take a look at the lightnings, they come factory equipped with 42lb injectors and a 90mm maf, with the stock/actual injector values and maf transfer loaded they are at .95 lambda at WOT with no kam correction with lambses at .82, with kam correction WOT drops down to .90, still too lean

by simply dropping the high slope from 42 to 33.5 the kam's stay at 1 and WOT is exactly on the money at .82 lambda, inferred load stays perfect and everything is well in the world, they typically pick up 9rwhp just by dialing in fuel, the low slope usually ends up around 36 or so to get the kams at idle and what not to stay at 1

I did 7 of these trucks just last month (5 harleys), they all responded the exact same way

[cgrey8]

If you are tuning on a vehicle that has the stock MAF and stock intake track, then I tend to agree that there is little reason to change the MAF curve. However if you are using the LMAF on a different vehicle that is NOT using the Lightning intake track, then it is plausible that the MAF curve might need tweaked here and there.

Also keep in mind, I'm talking from the standpoint of tuning with a vehicle where I didn't know the Injector parameters and I was working with a C&L 73mm MAF with a filter mounted right on the MAF and butted up against the battery and hood. I could not use the published MAF curve as-is. And it took some time tweaking the Injector Slope/BP values particularly since at the time, EA was NOT calculating the slopes correctly due to conversion errors in the GUFx defs for Injector PW. So it is quite possible I'm coming from such an extreme application that the "norm" can be treated this way. If that's the case, then so be it.

[tremec77]

For those that are saying to change the injector slopes and leave the MAF alone......wouldn't that be just as bad....if not worse than changing the MAF curve? It would seem that since any given MAF could have countless variables such as intake track before the meter, type of filter used, and so on...it would only make sense that your MAF curve would/could need changing a little. At the same time, it seems to me that the injector setting would tend to stay much close to stock perameters. Since injectors are not measuring anything, they are only spraying fuel at the set times, intervals, and lengths. I would think injector settings would tend to stay at what their tech sheets say and the MAF would need to be what is changed. Am I wrong in my thinking?

[cgrey8]

For those that are saying to change the injector slopes and leave the MAF alone......wouldn't that be just as bad....if not worse than changing the MAF curve? It would seem that since any given MAF could have countless variables such as intake track before the meter, type of filter used, and so on...it would only make sense that your MAF curve would/could need changing a little. At the same time, it seems to me that the injector setting would tend to stay much close to stock perameters. Since injectors are not measuring anything, they are only spraying fuel at the set times, intervals, and lengths. I would think injector settings would tend to stay at what their tech sheets say and the MAF would need to be what is changed. Am I wrong in my thinking?

I tend to agree with this. While I still advocate at least checking your injector slope/bp values in EA, it seems based on all the logs I've analyzed from others that there are only a few instances where people's logs show slopes different than what their tune was setup for to start with. Although there are a few out there that have slightly different rail pressures or have the wrong Injector Offset curve in and thus have logs showing different injector parameters are required. Either way, running the check can tell you if something is a miss even if it isn't the slope/bp values that are at fault.

But I do agree with your sentiment that there is far fewer ways to affect injector behavior than MAF behavior. In fact, the only way I know of is to modify the rail pressure.

[decipha]

i've tuned countless engines with ford 90mm mafs, on every single one of them i've NEVER had to adjust the curve, using just the injectors i've always managed to get the kam's to stay at 1

i even found an issue with an 02gt once where it hit turbulence at 3.5-4.2v and the afr went crazy, yet as soon as it past that everything went back to normal, a screen element (chicken wire) fixed that issue right up

if i didn't use the injectors to dial in injection i would have dismissed it and just increased the maf transfer to compensate which would have fixed the issue but jacked the maf transfer all up

People talk about messing with the MAF curve like making any change to the MAF curve is going to throw the Load calculation completely off. Yeah if you multiply/divide the entire curve by some fixed amount (+/-10 to 20%), then sure Load will be off. However if you are making small changes, the change in Load will be a few %, if that. It certainly isn't going to be enough to throw the whole tune into turmoil. Look at comparison of my curve to the original curve above. I don't think there's enough difference in those curves to throw the Load off any significant amount at all. Although the difference in the curve is the difference between good driveability or some leanout.

it will make a significant difference if your maf ever fails, not so much a concern on the guf ecu's but the newer ecu 94+ with inferred load this is an issue, this could be the difference between getting your vehicle towed home and driving it home with the maf unplugged

chris.. looking at your curve posted above, you can very easily drop the high slope and kick up the low slope to get your fuel dialed in with the actual maf curve loaded, then adjust the breakpoint and get the crossover perfect

For those that are saying to change the injector slopes and leave the MAF alone......wouldn't that be just as bad....if not worse than changing the MAF curve? It would seem that since any given MAF could have countless variables such as intake track before the meter, type of filter used, and so on...it would only make sense that your MAF curve would/could need changing a little. At the same time, it seems to me that the injector setting would tend to stay much close to stock perameters. Since injectors are not measuring anything, they are only spraying fuel at the set times, intervals, and lengths. I would think injector settings would tend to stay at what their tech sheets say and the MAF would need to be what is changed. Am I wrong in my thinking?

since we don't have global injection multipliers other than the slopes, thats the only alternative we have, using the afr scalar is only going to turn everything into turmoil because calculated pw at a richer lambse will be off by that factor (i'm talking the actual afr scalar for the ecu's that have them not the afr hack in guf)

but concerning the afr hack in guf, thats a multiplier on slopes as well so the same principle applies

[tremec77]

Since I'm about to have a few months off from work (school teacher :bananadance: ). I may take a few days and rework my current tune and try changing the injector settings and leaving the MAF curve alone and see what happens. I can see valid points for both theories. I will see how it goes and report back.

[decipha]

:) sounds like a plan

keep in mind using this method the wideband isn't used
the wideband will tell you if your running out of fuel though or if you have other issues

you'll need to be in closed loop and use the long term fuel trims (kam's) to dial your fuel in

i usually work my way down from WOT as its easiest for me but i don't recommend that method, i have enough experience that i know where the fuel is going to be to do something like that

kick the breakpoint really high and adjust the low slope to get your kam's at idle and just off idle to stay at 1

be mindful that adjusting the low slope and breakpoint while the engine is running will not take until after the engine is shut down and restarted

don't spend too much time with idle as adjusting the high slope is going to effect it as well so you'll want to get it close enough that you can start beating on her, half throttle is all thats needed to get the high slope dialed in, this is assuming your low slope and breakpoint isn't in the way


as a reference you can load your actual maf transfer in EA and use the ltft and cl filters under the maf to GET AN IDEA of what needs to be adjusted, if you see in the higher mafv the calculated flow is lower then you know you need to remove fuel by increasing the high slope

if you see there's a point where the calculated flow is non linear you know your at your breakpoint, reduce the breakpoint to increase fuel which will reduce the kam's from above 1

looking at the lower voltages of the maf curve you can determine how you need to adjust the low slope

remember low slope=low load like idle and very faint part throttle possibly cruise

high slope=high slope, WOT aggressive acceleration and what not


if you can't get the kam's to stay at 1 with a known curve then you know the injector offset needs adjustment :)

i find the published injector offset to be near perfect most times, its not too often i have to adjust it and when i do its not anything that any normal person would need to adjust anywho, I'm a perfectionist though so i settle for nothing less than perfect

[tremec77]

Since I have longtubes and will be using the HEGOs for this, will I need to change anything before attempting this?