Landtank MAF surprising scangauge results

[semlin]

VAF does not operate on the same principle at all, there is no correlation to MAF sensors. MAF = Mass Airflow Meter. AFM = Air Flow Meter (+ temp input + ecu calculations) = MAF calculated in the tables. The ecu doesn't *learn* anything with regard to MAF. There are just tables, meticulously designed and engineered to Bournulli's laws of flow thru a given size orifice. Just about every OEM manufacturer uses MAF to reflect quite accurate and actual engine airflow to a stock engine. The more accurately you do this, the better the EFI system will work.

By definition the wider you differentiate a 0-5v signal the more accurate the scaling will be of that input device. Add a margin of error for altitude and engine production tolerances, out goes the ecu tables.

Again, the graphs above confirm that, if you use them only as a MAF bore diameter change. In this case, the larger bore diameter is not well suited for the stock ecu programming, because it will never reach the ecu measure limit. The smaller bore diameter MAF has better resolution and uses the best scaling for the programmed limits of the engine air demand.

Welcome to EFI basics!

Scott J
94 FZJ80 Supercharged

thanks for the lecture, but nothing you said contradicts anything i said or addresses my actual point. what i said is that the 1993 landcruiser ecu is able to compensate for large voltage variations between VAF units at the same airflow.

your only clear point in this thread is that landtank's maf will likely send out a different voltage signal at a given airflow from stock. everyone knows that. the rest of what you are saying is speculation. you speculate that the voltage signal will be less accurate than stock by focussing on the housing, citing an extreme example of oversizing a MAF, and ignoring completely the different sensors. you speculate without any evidence that the ecu has some kind of arbitrary maximum permitted voltage input beyond which it will not attempt to map fuel, and you speculate that the differences in the voltage signals between stock and landtank will be so great that the ecu won't be able to keep up.

in other words, you have added nothing to the discussion other than a lot of verbage.

good luck.

and, for that matter, you are also wrong. the ecu clearly can learn about the signal from the VAF in the truck, because there is a large range of permissible variation in voltage signal between VAFs. it quite obviously must learn what voltage signal reflects what airflow condition in the VAF.

 

[SUMOTOY]

thanks for the lecture, but nothing you said contradicts anything i said or addresses my actual point. what i said is that the 1993 landcruiser ecu is able to compensate for large voltage variations between VAF units at the same airflow.

Exactly why it was replaced with the more accurate and simpler MAFS. The 93 80 doesn't compensate for anything, the programming is different because the AFM input is converted and tabled completely differently, not even remotely related. See Search on AFM > MAF conversions. Good luck with that, bttt for many years. And I followed *all* threads here attempting to do so... AFM operation is a complicated topic, that I'm happy to discuss elsewhere. IMO from your more recent statements, I believe you have a lot of mis-understanding of how it works within the ecu.

your only clear point in this thread is that landtank's maf will likely send out a different voltage signal at a given airflow from stock. everyone knows that. the rest of what you are saying is speculation. you speculate that the voltage signal will be less accurate than stock by focussing on the housing, citing an extreme example of oversizing a MAF, and ignoring completely the different sensors. you speculate without any evidence that the ecu has some kind of arbitrary maximum permitted voltage input beyond which it will not attempt to map fuel, and you speculate that the differences in the voltage signals between stock and landtank will be so great that the ecu won't be able to keep up.

A lot of assumptions you make. First, I don't pretend to know anything about the MAF on an 80. I do know how 0-5v MAF's operate. I cite only examples of 'effective' bore increases, and could care less about what sensor is used. The ecu doesn't have some arbitrary max voltage, but it can't exceed the max voltage (4.85v) of a sensor that can only read to 4.85v. If the B replacent MAF follows the A voltage output curve, it's a winner! I claim from just looking at a pic of the mod, you have a Bernoulli Law equation that makes that identical ramp an incidental coincidence at best. Looking at the photos of them, I have a more basic theory that measured voltage will be different between them. And anyone can prove me wrong, and we can move on to Fuel Injector Differential Pressure affects, Load Based Timing Table differences, and Force Induction differences to the MAF application. The table I put up is generic and can be used in all these concepts.

in other words, you have added nothing to the discussion other than a lot of verbage.
good luck. and, for that matter, you are also wrong. the ecu clearly can learn about the signal from the VAF in the truck, because there is a large range of permissible variation in voltage signal between VAFs. it quite obviously must learn what voltage signal reflects what airflow condition in the VAF.

You can make whatever claim you wish. I have no great desire to speak to AFM signal processing in the 80 ecu, because I personally think they are pieces of junk. ECU's aren't really that smart with regards to AFM or MAF's inputs. They don't 'learn' them, they only table them. The only variable in ECU software on the 80 that remotely resembles 'learning' is the baseline injector duration. It's the only value learned and stored in long term ECU memory in the 80, OBDI or OBDII.

I'm happy to discuss AFM anytime, as I have 10 years of of experience with them in my turbo 4R before my 80, and over 20 years in CIS and CISEIII applications. I might also share I have very detailed Bosch L-Jetronic EFI training manuals in my possession, feel free to start a new thread on AFM's, I will do my best to help present better understanding of how they work. They are good devices to understand how much better MAFS are. Here, they have nothing to do with MAF applications in any way.

Hint: start a new thread on AFM if you like.

Scott J
94 FZJ80 Supercharged

 

[Marc311]

1) Let's keep it civil. We're talking about MAF Sensors here - no lives hang in the balance.
2) I don't claim to have read all 14 pages or know much about MAF sensors and engine mgt, but I thought LT's premise was that the newer sensor may give more accurate readings. In other words, the graphs above show perfect parabolic curves. But the reality may be that the ACTUAL plotting of various sensors could show more blocking in one and a smoother curve in another. Or one sensor may react quicker than another. Couldn't it? (I'm asking here, not telling). The result being quicker, more accurate adjustments by the ECU. Or am I way off and I just set the discussion back 12 pages?

Either there is a huge placebo effect, or LT's device does actually do something. It might also be doing bad things which seems to be up for debate. But to say it does nothing when even the scanguage shows differences doesn't seem right. (I am not saying scanguage info is totally accurate, but it should accurately show things have changed.)

 

[semlin]

wow. thanks for setting me straight sumotoy. i guess that even though my actual point about ecu flexibility to vaf voltage signal is correct, and confirmed by the factory service manual for my truck, it turns out i don't know enough about vafs to have been allowed to make it, so i had better just keep quiet.

 

[SUMOTOY]

wow. thanks for setting me straight sumotoy. i guess that even though my actual point about ecu flexibility to vaf voltage signal is correct, and confirmed by the factory service manual for my truck, it turns out i don't know enough about vafs to have been allowed to make it, so i had better just keep quiet.

We can speak to the 93/94 Factory Manual address of AFM diagnostic values found on pages EG-240 and what they mean to the Bosch L-Jetronic software tabling, voltage compensation, and fuel injector duty cycle in an AFM thread if you like. AFM discussions on this thread are not relevant at all to MAF operation. Start a new thread.

Thanks

Scott J
94 FZJ80 Supercharged

 

[SUMOTOY]

1) Let's keep it civil. We're talking about MAF Sensors here - no lives hang in the balance.
2) I don't claim to have read all 14 pages or know much about MAF sensors and engine mgt, but I thought LT's premise was that the newer sensor may give more accurate readings. In other words, the graphs above show perfect parabolic curves. But the reality may be that the ACTUAL plotting of various sensors could show more blocking in one and a smoother curve in another. Or one sensor may react quicker than another. Couldn't it? (I'm asking here, not telling). The result being quicker, more accurate adjustments by the ECU. Or am I way off and I just set the discussion back 12 pages?

Anything is possible. Christo walks through a lot of the documentation that would support my charts above and all observed data to date. I personally can't believe that the comparative voltage values haven't been directly and exactly measured and plotted (solve for x/y axiis in A and B above). Forget a scamguage, a 10dollar voltmeter from sears.

Either there is a huge placebo effect, or LT's device does actually do something. It might also be doing bad things which seems to be up for debate. But to say it does nothing when even the scanguage shows differences doesn't seem right. (I am not saying scanguage info is totally accurate, but it should accurately show things have changed.)

Change the effective bore of a MAF, expect changes in voltage output from the MAF. That has been seen in so many other ways, values, scanguage logging, opinions and claims, but AFAIK output voltage curves were not directly compared or plotted. Until that is done, better, worse or dead nuts on is anyone's guess and Bernoulli's law rules the MAF device as I presented it above.

I'm not here to say a given change in MAF works or doesn't in any 80 application. There is no data to support that statement one way or the other. There is only physics and the basic operating principles of 0-5v MAF sensors I presented in the charts above. For 'expert' reference to the 80 specifically, see posts 95 and 107 in this thread.

Scott J
94 FZJ80 Supercharged

 

[SUMOTOY]

MAF Porting comparisons

Here is a comparison of 2 identical application MAFs with the same bore, the one on the left is a stock LS1 MAF (B to the charts above). The aftermarket "high flow" one on the right has the wing removed (A) for more 'unobstructed flow'. In dyno flow signal testing of this mod isolating only the wing (testing 2006 LS1 stock MAF - identical wing design but no screen), the MAF on the right had a drastically different airflow curve than the stocker. With no other changes than 'effective bore' the MAF on the right measured a decrease in flow throughout the signal output curve.

Without software rescaling of the MAF this mod also created a low speed resolution that was 'worse' than the stocker, and created an overall lean condition in closed loop operation. Propensity for knock also increased over the stock MAF.

Summary: With two changes to a modified 80 MAF (sensor and effective bore), careful signal measures are required to assure no lean condition exists with the new signal.

Scott J
94 FZJ80 Supercharged

 

[landtank]

Here is a comparison of 2 identical application MAFs with the same bore, the one on the left is a stock LS1 MAF (B to the charts above). The aftermarket "high flow" one on the right has the wing removed (A) for more 'unobstructed flow'. In dyno flow signal testing of this mod isolating only the wing (testing 2006 LS1 stock MAF - identical wing design but no screen), the MAF on the right had a drastically different airflow curve than the stocker. With no other changes than 'effective bore' the MAF on the right measured a decrease in flow throughout the signal output curve.

Without software rescaling of the MAF this mod also created a low speed resolution that was 'worse' than the stocker, and created an overall lean condition in closed loop operation. Propensity for knock also increased over the stock MAF.

Summary: With two changes to a modified 80 MAF (sensor and effective bore), careful signal measures are required to assure no lean condition exists with the new signal.

Scott J
94 FZJ80 Supercharged

This doesn't even begin to apply here. In your example the internal shape of the chamber was changed. So the air flow characteristics changed also. The chamber is now not matched to the sensor so yes it works like s***.

That would be like me putting the newer sensor into the chamber of the original one (egg shaped). The air flow differences would screw with the output.

The chamber I designed is exactly what the sensor was designed for. Size is only for scaling. That is how Toyota uses the same sensor on different size displacement engines.


If you are going to post up examples and data lets keep it on topic and that of a Toyota at least.

 

[SUMOTOY]

This doesn't even begin to apply here. In your example the internal shape of the chamber was changed. So the air flow characteristics changed also. The chamber is now not matched to the sensor so yes it works like s***.

Er, the internal shape of the chamber stays the same, or the internal 'shape' of the chamber stayed the same in the one that is being sold on this list. The sensor is located in the same place as the one with the wing. The bore dimension is the same (constant) - the actual sensor element is identical. This is well documented, not even sure what you are saying. If you didn't like the pic 'A', I can replace it with another 'hi flow' LS1 MAF that does the exact same thing and looks better(?). Look at the concept this way, only remove the center wing section of the LS1 stocker (B), the trace output follows (A). Happy to find another picture of 'an' A, if you wish. It's irrelevant to the the example, but if it helps you understand the concept better....

That would be like me putting the newer sensor into the chamber of the original one (egg shaped). The air flow differences would screw with the output.

To what? If you don't know what the signal-out voltage is, you screw the output making *any* change by definition. There is no such thing as a 'matched' sensor to housing. There are possibly better matched airflow properties of a given engine demand, with a given bore to the sensor voltage output profile. Exactly my point. As we can see from the toyota interchange part numbers listed above, the sensor is pretty irrelevant to the bore. The variables are software programming and engine demand. The flaw in your claim is that you matched a sensor to a given bore. If that sensor also works in a smaller bore application, that seems to be a bad match criteria. Specifically, it's irrelevent to how a sensor is calibrated.

Rick, you need to understand that the 'wing' in the Stock B above or B in the stock Toyota was not designed for production or S&G. The engineers that tune these engines know Bernoulli's law, because they program ecu softare tables to it. The 'wing' designs, and all obstructions are inserted to change the airflow/voltage profile of a given sensor. Specifically to the two Stockers (LS1 and Toyota), the airfoil designs *increase* low flow resolution (low voltage for a given flow in the charts above shifts to the right - without necessarily shifting the high voltage flow). The ecu software has a given voltage curve programmed based on the Stock Toyota (B) sensor profile. The chamber is now not matched to the software, so it works like s***. (I don't claim that yet btw, but you have made the statement wrong)

The chamber I designed is exactly what the sensor was designed for. Size is only for scaling. That is how Toyota uses the same sensor on different size displacement engines.

Already said that. In the LS1, the sensor is exactly the same, and the bore size is identical. If you look at Bernoulli's law, didn't you effectively change the size of the bore removing the obstructions to the sensor (all else being equal). Bernoulli's law also dictates that the flow of a liquid through a pipe will change properties based on the internal obstructions (airfoils, deflectors, inserts).

If you are going to post up examples and data lets keep it on topic and that of a Toyota at least.

0-5v MAF concepts are quite general to the above examples on purpose, because there is no magic to them, or how they work in *any* MAF EFI system. The bottom line? The examples are exactly the same. The difference between comparing the two toyota parts (stock vs modified) is no one has properly tested the output voltage. In the LS1 example, the test results are well documented and follow the exact same principles as applied to Bernoulli's laws of flow.

Why no voltage measures? Measuring the ecu interpreted values of a mod without software knowledge, seems to create many unanswered questions and more basic-observed results. To Bernoulli's laws of flow to the examples above, we can go to *any* post in this thread and explain the results as the shift from MAF B (stock) to MAF A (modified). To pulling and plugging the FPR, to Christo's increase in NOx, to increased 02 change interval, to pretty much any result, including the original poster's HP claims. Christo leads you all through this very well. I claim if you don't understand the concepts he presents, then just do a voltage plot of the stock vs the modified MAF. Don't even label anything but the voltage axis for now.

I make all these examples generic, and posted the pics of the LS1 MAF's to get a visual. Take the modified pic off, and just look at the stocker. Remove the wing (all else being equal), the voltage changes as described to the generic B/A charts I posted above.

80's applications are no different.

Scott J
94 FZJ80 Supercharged

 

[Romer]

Scott, I don't see why you would drag this up after 6 months other than to instigate something. Christo had data so he presented it. Your just taking Christo's data and extrapolating theories which don't seem to be agreed to or really make a difference.

Now if you had looked at the data, developed a solution and tested it, then that would be a great reason to revive this thread

 

[SUMOTOY]

Sumotoy EFI Project

Scott, I don't see why you would drag this up after 6 months other than to instigate something. Christo had data so he presented it. Your just taking Christo's data and extrapolating theories which don't seem to be agreed to or really make a difference.

Now if you had looked at the data, developed a solution and tested it, then that would be a great reason to revive this thread

MAF EFI is MAF EFI. I'm happy to exactly quote the technical reference regarding large MAF's - it's not theory really, it's just physics. There is no disagreement... Yet. I *use* this thread as a building block, because we can reference any claim, opinion and result and explain it. And allow me to procede/build on data in my own EFI project. Easy to reference a concept as a "see Christo post/link' in post xxx." He provides voluminous and excellent reference material in this thread.

If you read my posts, I don't claim to know anything about the LT mod, I am only presenting basic MAF operation in MAF based EFI. I am looking to document these basics, and hopefully find some output voltages. Why? Because as I go forward with a piggyback ecu fuel and timing project, 150 folks on a forum with a modded MAF hold quite useful data. Without the basic voltage slopes of the stock vs modified MAF, a lot more work comes into play. I'd rather use the known values of various trucks and table them. I could even 'incorporate' these slopes into software programming! Creating a solution with 150 given 'mods'. Specifically, "here's the software tabling of the stock MAF. Here's the software tabling of the Modified MAF". A MAF based programmable EFI, doesn't really care what MAF is installed. I don't care either.... Yet. I have no data to care either way.

Specific to those running turbo/SC, especially at altitude, understanding these MAF change affects in sensor voltage are critical to proper engine operation. The charts and examples I present, can go back 16 pages and explain every result, without reaching any subjective conclusion with regard to the mod itself...

Specific to your last paragraph, I searched to find 'any' relevant data and could find none. I would welcome any re-direct to that data thread.

Thanks

Scott J
94 FZJ80 Supercharged

I'm not frustrated, nor on a witch hunt. I don't believe Christo demonstrated that either.

 

[firetruck41]

If you don't know anything about LT MAF, and you want to document your own EFI project, why would you post in this long dead thread and why wouldn't you start your own thread about your EFI project? I'll tell you why, no one was biting when you were baiting LandTank in baktasht's EFI thread. If your not just the argumentative blowhard you appear to be, then prove it and start your own thread about your own EFI/MAF operation/etc.

 

[Romer]

. . . . it's just physics. There is no disagreement...

You continually say this and many continually disagree with what you post. There is plenty of disagreement that I read from other posters here.

I don't claim to know anything about the LT mod,

Then I don't really understand why you revived this thread instead of starting your own

I'm not frustrated, nor on a witch hunt. I don't believe Christo demonstrated that either.

Christo was providing real data as information, everyone saw the value and appreciated it. You are not providing anything that I can see other than your personal theories which you claim as fact because they are based on Physics

 

[landtank]

I hope people realize that I don't just off handily dismiss all negative thoughts on what I have designed. In the past I've used them to improve my caster plates for example.

Right now I'm hoping to have pre and post dyno runs with my MAF installed in a few weeks.

And am working towards a tuning sleeve so those who want to run with the FPR hooked up can.

I still believe it's safe to run this as there has been nothing other than speculation to prove otherwise.

I'm just trying to satisfy anyone who has bought one to the best of my ability.

 

[SUMOTOY]

You continually say this and many continually disagree with what you post. There is plenty of disagreement that I read from other posters here.

About what *exactly*? Anyone is welcome to correct what I wrote... MAF work as presented, if I have something wrong, please feel free to correct the *information* presented. I made this very general and basic, because it really is? There is a lot of information that the ecu uses with the MAF voltage, including load, Volumetric Efficiency, timing, and fuel.

Then I don't really understand why you revived this thread instead of starting your own

Because this thread 'should' be able to give us quick references to the resource documentation Christo put up. And quite possibly, answer all the unanswered questions and theories of the last 17 pages. I feel that basic understanding of MAF EFI operation, can answer any and all unresolved data to date. Without voltage slopes of MAF, the Scanguage data is premature, actually irrelevant.

Christo was providing real data as information, everyone saw the value and appreciated it. You are not providing anything that I can see other than your personal theories which you claim as fact because they are based on Physics

I believe I understand EFI theory, application and practice, because I've written code for EFI systems, have measured MAF voltage slopes, changed software tables, including MAF and MAP inputs using large MAF sensors and larger MAP sensors. MAP standalone (speed density) and MAF, and MAF/MAP hybrid programming in stock and standalone EFI... ALL MAF voltage follows a slope, with x-axis voltage, and y axis Mass of air, as presented above. And to Bernoulli's law it's not linear. That's not theory, thats MAF operation, including whatever specific toyota part number is inserted into the modded housing.

I also explained how the MAF sensor mod works in any MAF or MAF/MAP hybrid system. I can use any of the 3 piggyback systems I've used before on other engines, and which MAF used is irrelevant. Because I can scale the MAF in the ecu programming, it can be the stock 80, modded 80 sold here, or a 90mm Ford motorsport monster. I only need the voltage slopes. Once they are installed, I can adjust very a wide variety of parameters within the software to maximize it's given slope.

All MAF EFI systems start with Voltage/MASS of Air slopes. Without knowing code, trying to chase upstream leaves a lot of unexplained anomolies found by Christo and others.

I can get extremely specific to the 80 ecu, but there's no apparent MAF voltage slope data to do so. The best part of getting thru the basics here, is any 'other' solutions to 80 EFI, has to follow the same basic operation. I can chase a piggy back fuel/timing computer in a separate thread, and Rick can continue his MAF development thread without me in it... Right now, we have 150 MAF meters that have never been voltage sloped to identify what they are doing. And thousands of stock MAF meters that have never been voltage sloped either. I find that changing to the former without the latter, is not relevant yet.

Cheers

Scott J
94 FZJ80 Supercharged

 

[Romer]

I am not going to play your game and argue with you

My post still stands and all you did was further emphasize to me what I had already said

You should start your own thread to expand on your theories and those that want to engage can

 

[SUMOTOY]

I am not going to play your game and argue with you

We aren't arguing anything! We have no data on which to do that. And I can explain *any and all* measures in the last 17pages here. I believe to know Christo can too. In the last 17 pages, no one else seems to be able to. After reading each and every post, I started where I identified the main focus should be, at the MAF signal. How it works and it's operation has no opinion, in the 80, or any other MAF equipped ecu.

My post still stands and all you did was further emphasize to me what I had already said
You should start your own thread to expand on your theories and those that want to engage can

Problem. EFI and MAF operation are going to come back to this thread again, and again. I can start a new thread on EFI basics and reference this thread over and over. OR, we can have one continuous thread that includes references and answers to the unattended questions in the last 17 pages. I can supply a lot of documentation as well, I own a lot of it. And, I can pick up the phone and get answers to a lot more.

Christo identified some issues which weren't at all new to me... You, as a moderator forbid me to take them on... Before 150 MAF sensors were sold. A year later, a well respected 80 technician and 80 member takes these issues head-on, and got nowhere. The good thing, is he documented well, how the Toyota 80 EFI system works.

I will document my EFI piggy back project in another thread. Just as LT documented his MAF development in another thread. Both/either can and should reference this thread for baseline. I understand you can make the call. I present, the MAF install will all come back to this thread anyway.... It's the root cause in every anomolly found in this thread.

Scott J
94 FZJ80 Supercharged

 

[concretejungle]

i have no idea what most of this technical mumbo-jumbo is. However, it is a technical thread on a board open for discussion. I say let it continue. If you are not enjoying the discussion or arguments i would just not click on this thread.......

I do hope something constructive comes out of all this. I'm always interested to see facts and fiction on our heavily modified rigs.

 

[SUMOTOY]

I hope people realize that I don't just off handily dismiss all negative thoughts on what I have designed. In the past I've used them to improve my caster plates for example.

I'm not 'negative' about this mod, I'll restate clearly I don't have enough information to be so. MAF's operation is pretty well known, to oem's and tuners. I do share Christo's concern about the FPR disconnect, as does any EFI system tuner, it is a constant in every oe and aftermarket EFI system, AFAIK *without* a single exception. Rising Rate FPR's just raise the 'constant' value. Those at altitude, especially with Force Induction, are at the highest risk with this mod. My suggestion, until someone gets the voltage slope of the MAF, leave the FPR connected.

Right now I'm hoping to have pre and post dyno runs with my MAF installed in a few weeks.

To properly slope MAF voltage on a dyno, a lot of data is required, btdt. My suggestion would be to have a set dyno load, and use 1000rpm increments, then use 500rpm increments as you transition to open loop. Then log some tip in throttle at various throttle angles (I'm pretty sure I read Scamguage will give you TPS angle - very useful here). A force induction truck won't be able to use any of this data, that will require a retest.

And am working towards a tuning sleeve so those who want to run with the FPR hooked up can.

Premature, and likely a waste of time. Changing MAF slope with hardware is a massive undertaking with very little benefit. My suggestion, hook up the FPR, and go elsewhere for 'corretion' for whatever result you get.

I still believe it's safe to run this as there has been nothing other than speculation to prove otherwise.

"I think a EFI system that dumps fuel, will not blow up a motor." Direct quote from an EFI tuning guru I spoke to about this mod. I still believe the unsafe part of this mod is the FPR disconnect. Christo documented it, happy to get a direct quote for you, I had a very lively discussion about this. With regard to 'it's safe' to run any MAF mod, MAF voltage slopes are needed to make that claim.

The biggest issue with not software scaling MAF, is that a LTFT at +18% means that an otherwise known good 02 sensor will not last nearly as long, since you have a much narrower 2% 'wear' limit. That explains many of the 02 sensor half-life observations in this thread.

Looking forward to the dyno results.

Scott J
94 FZJ80 Supercharged

 

[landtank]

The biggest issue with not software scaling MAF, is that a LTFT at +18% means that an otherwise known good 02 sensor will not last nearly as long, since you have a much narrower 2% 'wear' limit. That explains many of the 02 sensor half-life observations in this thread.

luckily for everyone with my MAF installed the ft% are near identical to that of the stock MAF throughout the whole range of operation. No where near 18% at any time.