Landtank MAF surprising scangauge results

[aclos3]

Scott,

You see, this is where I will conclude that the mod is OK. I am a scientist by profession and you sound scientifically minded as well, so you must know that even when the theory points strongly towards a single conclusion, if that conclusion is not presenting itself in the real world something is being missed by the theory. That's not to say that the theory is flat out wrong, just that it misses the mark in terms of magnitude or that some other variable was not accounted for.

Anyone installing this mod accepts that there is a certain amount of risk. Heck, anything that increases power increases wear and tear on the whole system. However, it seems that a ton of effort is being put into perpetuating the idea that this mod potentially leans out your motor to the point of damage when THAT IS NOT THE CASE.

 

[semlin]

aclos, i appreciate you are trying to reason with him. i did too before giving up. but you are wasting your time. does it not occur to you as at least "odd" that in his latest post, sumotoy just completely ignored the very pointed criticism he has just received and instead carried on with another repetitive post? he is a troll and you are feeding him.

 

[Grench]

Your right. Next time he's posting something harmful to the truck's mechanics or dangerous to the safety of the passengers I'll just ignore him and pray for those who take his advice.

2X -

We could go through a whole montage of potentially dangerous suggestions. The two that irked me the most were the 'CDL on all the time is safer' thread and the 'in spring airbags >= load distribution coupling when hauling large heavy trailers.' (Neither is true - please confine those topics to the threads they were in. Cited here as example only and not to re-open those discussions.)

Despite his cries of, "it's just physics," he denies fault when physics are used to prove him wrong. As Romer said, he has made a few good modification threads - then goes and blows all of the generated goodwill by claims of infallibility. Just plain sad.


/************* Back on topic.... **************/

I am a happy customer of both Slee and Landtank.

I have the LTMAF & new sensor sitting in a box on my shelf - it has been a 'round to it' project.

Slee's results are interesting and very informative. I can see where the long term effects of a lean condition should probably be avoided.

I still like the idea of the cleaner flow with the new design MAF.

What I'm wondering is if the LTMAF may be 'too much of a good thing'. I.e. maybe the newer sensor pushes too far to the lean side. In which case it may (in theory anyway) be possible to push the readings it provides back in the other direction 25%? 50%?. I haven't dug into the electrical spec on it - not my thing really. It may be as simple as a couple of resistors - may not.

So, back around to reality - theory only gets us all so far. What -real- tests would help to work this out? Tools needed?

IMHO - Likely need to be tested on a dyno with computer hook up and sniffer on the exhaust. Pre-post install testing is nice, but it would be nice to have a few 'in between' points where the new sensor's signal is pushed somewhat to look more like the old sensor's. May be as simple as running a potentiometer on one of the leads to attenuate the signal - I don't know.

Yes, this would be expensive - but is there another way to sort it out?

 

[SUMOTOY]

Scott,
You see, this is where I will conclude that the mod is OK. I am a scientist by profession and you sound scientifically minded as well, so you must know that even when the theory points strongly towards a single conclusion, if that conclusion is not presenting itself in the real world something is being missed by the theory. That's not to say that the theory is flat out wrong, just that it misses the mark in terms of magnitude or that some other variable was not accounted for.
Anyone installing this mod accepts that there is a certain amount of risk. Heck, anything that increases power increases wear and tear on the whole system. However, it seems that a ton of effort is being put into perpetuating the idea that this mod potentially leans out your motor to the point of damage when THAT IS NOT THE CASE.

Problem, we don't have the proper data to support *any* conclusion. Only documenting bulletproof motors seems an interesting argument? Any 0-5v MAF will work in the application, and by the 400 posts here, it only has to pass the LTFT to conclude it's doing everything right? In 15 pages, we can observe a lot of smart engineer-type 'scientific' folks present opinion and theory, and most of it doesn't correlate with how the mod, or the ecu it sends output to, works. The device operation, and ecu function are known givens.

Logic train:
If Load Values are calculated from MAF values, then the MAF value must accurately reflect the amount of incoming air. If it doesn't, OBDII LTFT values aren't a good measure of the device. And OBDII LTFT values are not a measure of Load Values, Timing Values, or Trans ECU shift points that have all changed by definition.

I see this as Christo does, a larger MAF causing Lean Fuel vs MAF Calculated Load, Increase Timing vs MAF Calculated load, and decreased shifts vs Calculated Load, explains all the 'observed' subjective data presented in this thread. The objective data is missing, because no one has measured the correlative data required. That backs up to intent vs application. Even with the ability to write software, I tend to be very careful messing with Load Values.

When you conclude it's ok, does that include the boosted motors at 10,000 feet? Does that include the guys that can only get 86 octane fuel? Does a decrease in Calculated Load Values vs Timing, cause one to 'recommend' premium fuel? Or is it all just fine all the time and Mr. T's Engineers just ultra conservative in the risk department?

Many vendors, like Lingenfelter for instance, will give you the MAF slopes for their modified sensors - others give the MAF transfer function to remap the MAF software tables
http://www.lingenfelter.com/sites/l...files/LPE100mmMAFsensorflowcurvesv1.0_000.xls

Just sayin' - carry on!



Scott J
94 FZJ80 Supercharged

 

[cruiserdan]

Lather, rinse, repeat,



Lather, rinse, repeat,



Lather, rinse, repeat,



Lather, rinse, repeat,

 

[Grench]

Lather, rinse, repeat,



Lather, rinse, repeat,



Lather, rinse, repeat,



Lather, rinse, repeat,

Gawd it must be late - I'm dying here.

 

[NaterGator]

SUMO, SUMO, SUMO.

With what idiotic authority do you claim load values are calculated solely by the MAF readings? The TPS, oxygen trim levels, knock sensors, etc ALL feed to a control chain and transfer function that determines the load of the engine. I can promise you Toyota has developed an incredibly detailed solution to the equation that accurately predicts the engine load and adapts to changes in stimuli, and that it is ABSOLUTELY not based entirely nor necessarily on the MAF values.

Unless you can post up the state space representation of Toyota's load calculation and LTFT control system you can just as well stuff it. You're contributing nothing of substance to this thread and obfuscating the issue for people. In short, if you want to have a flame war with Rick send him a PM. If he gives a damn he'll respond, and if he doesn't you'll get ignored. At least you won't be a detriment to some otherwise interesting 80 series tech threads. Learn to pack up your bags and admit you're not being helpful or informative...

 

[landtank]

Well, Sumotoy did send me a PM and I ignored it. IMO he has an agenda. Not sure what it is but to post as he has, it's purpose is something other than sharing information.

But it does seem that he thinks that he's in a position that we all need to satisfy him that this mod is OK. Kind of a power trip if you ask me.

Now I'm not a psychologist and this is only based on my personal experience of being 6'6" tall most of my life, but he seems to have the classic Napoleon Complex.

I'd love to know how short this guy is?

 

[cmiller114]

FYI

I've been running Landtank's MAF for over a year with no signs of a lean condition. It seems with all the computer jargon some guys have forgotten good ole low tech plug reading.

 

[SUMOTOY]

With what idiotic authority do you claim load values are calculated solely by the MAF readings? The TPS, oxygen trim levels, knock sensors, etc ALL feed to a control chain and transfer function that determines the load of the engine. I can promise you Toyota has developed an incredibly detailed solution to the equation that accurately predicts the engine load and adapts to changes in stimuli, and that it is ABSOLUTELY not based entirely nor necessarily on the MAF values.

Unless you can post up the state space representation of Toyota's load calculation and LTFT control system you can just as well stuff it....

I'm no authority, it's in the Toyota Training Manuals. Please re-read the entire thread, but here's the relevant information you appear to be looking for, from Toyota Manuals (referenced on page 5 in this thread)

"The Mass Air Flow Sensor converts the amount of air drawn into the engine into a voltage signal. The ECM needs to know intake air volume to calculate engine load. This is necessary to determine how much fuel to inject, when to ignite the cylinder, and when to shift the transmission.... In fact, basic injection and spark calculation are a function of just 2 sensors, engine speed and engine load sensors... Engine Load - Mass Air Flow meter

Causes of Incorrect Spark Timing
On systems that use the ECM to compute ignition spark advance there are only two conditions which are likely to cause spark timing to be incorrect; initial timing or false input signal to the ECM. If engine load is miscalculated because of incorrect input signals (toyota boldface), spark advance angle will not be appropriate for engine operating conditions. This will result in driveability and emission problems.

Even with initial timing correct, it is still possible that the system is miscalculating ignition timing as a result of incorrect sensor inputs. For example, if an airflow meter indicates light engine load, when in fact, the engine is experiencing high engine load, the ECM may incorrectly respond by over advancing ignition timing to the point of causing detonation... If inaccurate sensor inputs are suspected on earlier EFI and TCCS vehicles, it is recommended that you perform standard voltage checks of all major sensor inputs to the ECM."

http://www.autoshop101.com/forms/h48.pdf
http://www.autoshop101.com/forms/h59.pdf
For concerns regarding a modded MAF on the 80 in this context, see Christo's posts 144 and 150 on page 5 in this thread.

The documentation from Toyota couldn't be more clear on how Load Values are Calculated from the MAF. Fuel, ignition and transmission shifts all use that MAF calculated load. The airflow meter example they use is quite appropriate to this mod, IMO.



Scott J
94 FZJ80 Supercharged

 

[nat88toy]

Now I'm not a psychologist and this is only based on my personal experience of being 6'6" tall most of my life, but he seems to have the classic Napoleon Complex.

I'd love to know how short this guy is?

could he be any shorter then i am....maybe

 

[nat88toy]

I'm no authority, it's in the Toyota Training Manuals. Please re-read the entire thread, but here's the relevant information you appear to be looking for, from Toyota Manuals (referenced on page 5 in this thread)

why read the entire thread he only had to read 2 of your posts and he read everything considering its all the same crap

 

[NaterGator]

why read the entire thread he only had to read 2 of your posts and he read everything considering its all the same crap

Yes, it is readily apparent to me that Scott is not well versed in control engineering and the dynamics a feedback system like the LandCruiser's closed loop EFI system. Nor is he particularly well versed in the design of analog and digital sensor systems. All of this despite his claims of extensive work experience and CIS training (mind you all the CISEs I know don't learn s*** about control theory).

He regurgitates the same trivial knowledge of such systems ad-nauseum and then extrapolates ridiculous and flawed assumptions, repeating them over and over as truth.

Ex: His constant banging on about ECU lookup tables and how the ECU simply performs straight forward tabular lookups to determine its output vectors. This concept is absolutely ridiculous; while ECUs most definitely do contain tabular data from engineering studies, they are used to feed the system's instantaneous input vectors in its state space representation. The ECU stores some values of the state in what is called the "feed through" matrix/vectors, which in turn are tabulated and contribute to the next iteration of the system. Obviously this is all a discrete time process using sampled sensor values, etc. The engineers are very careful to setup the state space system so that it responds to a wide range of possible inputs. Once the system is represented in state space form it is trivial to deal with a vast array of inputs while maintaining BIBO stability.

Your guess is as good as mine re: why Scott doesn't understand feed-forward systems or chooses to ignore them in this case. The point is he's either a) full of crap and doesn't know what he is talking about or b) choosing to ignore what he knows so he can keep blathering on about nonsensical drivel to get attention on an internet forum. I suppose there are other explanations, but none are particularly pretty.

 

[SUMOTOY]

Yes, it is readily apparent to me that Scott is not well versed in control engineering and the dynamics a feedback system like the LandCruiser's closed loop EFI system. Nor is he particularly well versed in the design of analog and digital sensor systems..... to feed the system's instantaneous input vectors in its state space representation....

Hunh?!
That was very impressive wording that doesn't address how the ECU works in the 80, or really any other EFI system. MAF input is strictly a function of voltage drop from airflow through a given size tube. The software dictates what that voltage function is, how accurate it is, it's upper and lower limits, and it's 'resolution'. Why? Because you haven't done anything to recalibrate the A/D converter (FYI - technically referred to MAF transfer function).

What Toyota Engineers do (along with most any other oem or aftermarket tuner), is they take various 'Load Values' and create a Map for fuel and ignition tables (Speed Density uses Manifold Absolute Pressure Sensor for Load, Mass Air uses MAF sensor for Load). If you look at the Toyota documentation regarding ignition tables I referenced, if the MAF reading is off, so is the Load Value. The Load Value is off, so is the fuel and ignition table for any given RPM. Load Value for ignition is strictly a MAF vs RPM table. 80 ignition tables from MAF sensor voltage are *only* corrected by engine temp and knock sensor intervention, that's it. How do I know that? Because Mr. T documents that.

There is no knickity-snickity black box vectoring stuff going into all this. Every stock oem software program I've seen uses this formula for 'ideal' airflow (WOT)
Engine CID x .5 x Volumetric Efficiency x rpm/1728 = CFM
Convert to Mass of Air = CFM x Density Ratio (air temp and pressure correction) = gm/s (WOT ideal)

Load is then corrected via the MAF value. For instance, if at 4000rpm the 80 is 40gm/s using the above tabled Ideal Value (Theoretical Airmass), and you have a MAF airflow value measured value of 20gm/s (Actual Airmass), a Load Value of 50% is used in the fuel and spark Maps. Simply put in a Mass Airflow Ecu:
Load Value = Current Airmass/Theoretical Airmass

The Toyota Engineers spend a lot of time on the dyno and in testing to determine what the highest possible incoming air is for the 80 engine and determine a very accurate VE and Theoretical Airmass using the equation above. They then assign this the software max MAF value of 5v. Then the voltage slope of the MAF has the widest 'resolution' to the actual air demand of the engine.

When Toyota designed the stock MAF, the obstructions inserted into the tube are used to manipulate airflow to give more accurate airflow at low values. If you look in the bore of the MAF B (Stock) vs the MAF A (modded), the airflow is of higher velocity at lower values, to give a more accurate voltage value at low actual airflow. That 'resolution' is *expected* in the software transfer function of MAF voltages. Read: The inserted airflow obstructions in a given tube change the MAF voltage slope, which is then programmed in the software.

This means any deviation from the *expected* airflow voltage value as measured by the MAF, is just summarily: incorrect. Many opin here it's better, more accurate, has better S/N, etc... All incorrect assumptions, because a MAF meter is not calibrated in the tube, it's calibrated in the MAF Transfer function A/D tables of the software (see Christo's comments page 3). And it defines Load.

When doing base Maps for standalone, I use the above formulas to get a car started and running. Then use dyno and driving time to tweek the MAF transfer function values to give the most accurate MAF values possible. FYI in oem systems, this MAF value is in the 1% accuracy range, and most tuners like to stay as close to this as possible. Why? Because the more accurate your actual measure of incoming air, the less work you have to do elsewhere.

YOMV



Scott J
94 FZJ80 Supercharged

 

[landtank]

So I'm clear here Scott, this was all done on an 80 series Land Cruiser? I'm asking because last I knew Slee had great luck tuning a 100 series LC with the AVO turbo but not so much luck on an 80 with an AVO.

This tells me that the ECUs ARE different and what works for one doesn't for the other.

And if this is your approach to Audi's, then whats the point of posting this anyway.

 

[Romer]

Yes, it is readily apparent to me that Scott is not well versed in control engineering and the dynamics a feedback system like the LandCruiser's closed loop EFI system. Nor is he particularly well versed in the design of analog and digital sensor systems. All of this despite his claims of extensive work experience and CIS training (mind you all the CISEs I know don't learn s*** about control theory).

He regurgitates the same trivial knowledge of such systems ad-nauseum and then extrapolates ridiculous and flawed assumptions, repeating them over and over as truth.

Ex: His constant banging on about ECU lookup tables and how the ECU simply performs straight forward tabular lookups to determine its output vectors. This concept is absolutely ridiculous; while ECUs most definitely do contain tabular data from engineering studies, they are used to feed the system's instantaneous input vectors in its state space representation. The ECU stores some values of the state in what is called the "feed through" matrix/vectors, which in turn are tabulated and contribute to the next iteration of the system. Obviously this is all a discrete time process using sampled sensor values, etc. The engineers are very careful to setup the state space system so that it responds to a wide range of possible inputs. Once the system is represented in state space form it is trivial to deal with a vast array of inputs while maintaining BIBO stability.

Your guess is as good as mine re: why Scott doesn't understand feed-forward systems or chooses to ignore them in this case. The point is he's either a) full of crap and doesn't know what he is talking about or b) choosing to ignore what he knows so he can keep blathering on about nonsensical drivel to get attention on an internet forum. I suppose there are other explanations, but none are particularly pretty.

Well Said

I use to try and reason with him. I was always wrong and he was right, there was no budging in his position, no open communication

His response to your post pretty much proved your point

The fact he does the popcorn at the end of his last points shows he is just trying to stir things up and see how it goes.

I have been the Mod here now for five years and in all that time, he has constantly been the biggest irritant to me and other members. I cant think of anyone who has been a bigger problem.

Note: If I was abusing my position as a moderator, I would just delete him. I am a non paid volunteer with a real job and a member here as well and entitled to my opinion. This is more for Sumo as he likes to PM me to tell me what the real definition of my job is because he knows better than woody or me.

 

[concretejungle]

Sumo, do you mind me asking what tuner shop you own/run? Just wanted to look at some of your work.

 

[george_tlc]

Not even February and I just keep thinking of this little guy each time this thread pops to the top again:

cheers,
george.

 

[SUMOTOY]

So I'm clear here Scott, this was all done on an 80 series Land Cruiser? I'm asking because last I knew Slee had great luck tuning a 100 series LC with the AVO turbo but not so much luck on an 80 with an AVO.
This tells me that the ECUs ARE different and what works for one doesn't for the other.
And if this is your approach to Audi's, then whats the point of posting this anyway.

This Mass Air approach (vs Speed Density or hybrid) applies to all standalone systems, and all OEM's. Including the 2010 my. More information regarding what is "great luck in tuning" might help me understand a point you are trying to make.

What this tells me... You have uncalibrated hardware causing a software problem. You either need to recalibrate the hardware to have the same voltage slope (for given software tables), or you need to rescale the software to reflect the new voltage slope of the hardware.

What the 100 does with an AVO system seems irrelevant? Feel free to tell me more about it.


Scott J
94 FZJ80 Supercharged

 

[landtank]

Well Said

I use to try and reason with him. I was always wrong and he was right, there was no budging in his position, no open communication

His response to your post pretty much proved your point

The fact he does the popcorn at the end of his last points shows he is just trying to stir things up and see how it goes.

I have been the Mod here now for five years and in all that time, he has constantly been the biggest irritant to me and other members. I cant think of anyone who has been a bigger problem.

Note: If I was abusing my position as a moderator, I would just delete him. I am a non paid volunteer with a real job and a member here as well and entitled to my opinion. This is more for Sumo as he likes to PM me to tell me what the real definition of my job is because he knows better than woody or me.

A big problem is to keep him in check so somebody doesn't get hurt or have a problem with what he suggests.

Last year a member was asking for advice on using a CT26 turbo on his !FZ-FE.

I had concerns over fitting it to the manifold he got and then actually getting it in the vehicle, others had thoughts that it would be a poor choice in size.

Scott on the other hand applauded the idea and countered all our concerns at length to as why it was the ideal solution and bagged on us all about our lack of knowledge.

Now the poor guy has a turbo that he can only use if he heavily modifies it so it's going to be sold to get a GT35r.

Sumo won that battle at the cost to one of our members.