Tuning - Nissan ECU patent info

[KATwo40]

As per Bernard's request:

From the ecu2 Nissan tuning forum, regarding an issue with a car that ran poorly with a modified TP scale (reduced to match the reduction in K value)

Posted by bill:"Back from the dead! Car mentioned in first post works just fine with uncorrected tp scale. Yesterday I was tuning a GTI-R RNN14 with 550cc injectors and did the same thing. It worked properly only with uncorrected tp scale. Can someone explain please why this happens?"

Posted by CBDrift:"Yeah - i noticed when i tuned an n/a sr20ve that the load scales didnt need changing and made it a nicer car to drive if they were unchanged.... i was also asked to change my payware editor to use a tickbox to choose to correct the load scale when the injectors are rescaled as the tuner who asked was noticing that when tracing and the injectors are rescaled the trace still reaches the same tp value - his theory is that injectors just need K, Null and Voltage correction changes and afm's just need the VQ and K value changed and the load scale should be tuned to suit the engine or boost level desired... Cant say i disagree after seeing the results "

Here's the link, if you wish to read.http://ecu2.forumwise.com/ecu2....html

I think what's being overlooked here is that Edub and myself are not saying, "The ECU doesn't use MAF voltage." We're saying the ECU seems to access the same columns in the table, regardless of MAF voltage.

[Chezedik]

So then what I have gathered from this is that given Throttle Angle and RPM it locates it's position on the table and defines a variable. Then it multiplies that variable by K constant and the relative VQ. This then defines the correct rate of fueling. It is then corrected for lambda.

Does this sound right?

If that is the case, then this doesn't seem to be earth shattering news, more or less, what I expected.

[Edub1]

There is rock solid proof of what I'm saying. I discovered this problem not only by observing map behavior but by the logical method of reduction to absurdity. Here is why the former theory of MAF Vs RPM fails.

1) A/F ratio once set holds a given level regardless of airflow. This contradicts the former theory, but would hold true if TP= (MAF(K)/RPM)(cell value)

2) Hacking the MAF signal affeacts a global change but does not disturb how the ECU scales the tables. If the main fueling info was based only on a MAF based lookup table, piggiebacks would skew the table too much and would not work. They do work because TP is set by a fixed calculation involving MAF, K, RPM and various corrections. They change the base TP independant of any tables by changing the MAF signal. Here is a reductio ad absurdum. A MAF signal cut in half simply could not work according to prior theory but it works perfectly according to the new one.

3) RPM affects flow much more than VE. So, our ECU would move to the right based much more on RPM than on VE. Because flow is dependant on RPM, plotting flow Vs RPM is not feasible. You would basically wind up with RPM Vs RPM.

Also, the patent info clearly states that base pulse width to create a stoich condition is defined as TP=(flow)(K)/RPM. The other factors are simply to correct discrepancies in TA Vs flow and are largely irrelivant to what we are discussing.

I went over the patent with a fine tooth comb. It clearly states that a base, stoich TP is derived by a straight linear formula and that the base TP is then altered by correction factors. In this case our correction factors are found in an independant map and determined by TA Vs RPM. And this is not the flow based map described in the patent. That is something quite different.

[NateDogg]

Edub1 and Bernard: What would you suggest that I alter in order to achieve a more accurate AFR under 0 vacuum conditions?

[Edub1]

Edub1 and Bernard: What would you suggest that I alter in order to achieve a more accurate AFR under 0 vacuum conditions?

Vacuum has nothing to do with it. You only need to alter K to make your base TP stoich. Once this is done your car should behave just like stock if all else is equal. You can then alter your fuel map as needed, if needed.

[Edub1]

So then what I have gathered from this is that given Throttle Angle and RPM it locates it's position on the table and defines a variable. Then it multiplies that variable by K constant and the relative VQ. This then defines the correct rate of fueling. It is then corrected for lambda.

Does this sound right?

If that is the case, then this doesn't seem to be earth shattering news, more or less, what I expected.

I'm not sure what you're trying to say but if it sounds like this then yes.

TP is a base pulse to make a stoich A/F ratio. K is a global adjustment. Map cell values are a specific adjustment for a given throttle angle & RPM. Both K & cell value are multiplied with the base TP to give correct fuel at any rate of flow. Because TP holds stoich for any rate of flow.

[Bernard]

Edub1 and Bernard: What would you suggest that I alter in order to achieve a more accurate AFR under 0 vacuum conditions?

NateDogg, I read your question wrong the first time. You need to work on the k value and latency. You will never be able to get the commanded afr to read out exactly the same on the wideband, but it should be within .5 or so. You can also increase ttpmax to 254.

KATwo40 - RE: linked thread.......c'mon dude. A couple of posts later Wallerio describes exactly what I find and have datalogged many many times. The user Vetal described what he did and I can see why he had problems. Relying on Live edit for corrections and assuming they will be bang on is a sure fire way to at best have a crappy tune and worst case, blow up your engine. In fact the romulator itself has issues that unless you're aware of them, you will think you've adjusted values when in fact they haven't changed at all. He hasn't posted his afr's timing or anything. It could be any number of things.

FSM is the factory service manual. It explains very briefly how the ecu calc's load and how the ignition is mapped. It states that tp is used for one scale and rpm for ignition. I am surprised no has looked at that before.

Now to the question at hand. The ecu establishes the theoretical pulse with the maf voltage, rpm and kvalue. Everything before that in the program flow is to decide what state the engine is in and to decide what enrichments to add. After it has calc'd the tp, it uses rpm and tp to access main fuel map value. Then it uses the last equation to correct for accel, high load, o2, tps position , battery voltage, injector latency etc etc. This is why there is a TTP max table. Of course this only applies to those ecu's that are maf based, I have no idea how the alpha n ecu's work, I've never tuned one of those, nor do I ever expect to.
Modified by Bernard at 5:38 PM 1/23/2007

[KATwo40]

Could you post up a couple logs of maptrace? Maybe comparing map cell locations @ 7psi and maybe 12psi?

[KATwo40]

ROFLMAO! I just read the FSM...showing the timing table referencing TP against RPM, plotting the correction value.

I officially stand corrected on my assumptions.

I guess it's time to just start doin' some datalogging.

[Edub1]

We have been discussing the roll of the fuel table up till now. Timing could possibly be different.

Anyway, my electronic version doesn't show that. It says that timing is determined by "various sensors" and pre-programed values. Pg 12 E-Theory/Operation.

Can someone post up that info? Does it specifically say TP or does it say load?

For those of you who are going to do some maptracing, try it with the TPS unpluged.

[KATwo40]

It specifically says TP against RPM in the EL section (electronic version).

It says this for fueling, then goes on to show the same for timing, with the aid of a table example.

I'll post it up later, gotta go to work right now.

[Edub1]

It specifically says TP against RPM in the EL section (electronic version).

It says this for fueling, then goes on to show the same for timing, with the aid of a table example.

I'll post it up later, gotta go to work right now.

I've found something of this type in the 91-94 FSM.

Funny thing, it says "The ignition timing is controlled by the ECM in order to maintain the best air to fuel ratio." Huh?????

Anyway, it does show such a table and describes the ECU as "detecting information SUCH AS" TP & RPM. It does not say specifically TP and RPM it says info "SUCH AS" TP & RPM. It also gives a list of other things that contribute and goes on to say that, "the ignition timing is revised by the ECM according to the other data stored in the ECM."

Plus, numerous other tables show throttle position as a factor. Also, why does this give a 1-5V signal if that signal is not used by the ECU?

It is probable that the ECU does determine a base ignition slope that forms a straight line and then uses additional maps to further modify this signal.

One easy way to find out would be to maptrace with the TPS unplugged and use some batteries or something to provide a fake signal to the ECU and see what happens. I'm betting with a fake 3 V signal the ECU starts to look right, unless it checks it and rejects it somehow. But playing with the throttle and using the fake signal should skew the results somehow.

[Chezedik]

What section is the Map info on in the FSM? I would guess EC.

[Bernard]

EF - EC 17 and EF - EC 21

I wouldn't recommend searching for a better understanding of nissan ecu's until the basic concept explained on those pages are understood. lol

[Edub1]

EF - EC 17 and EF - EC 21

I wouldn't recommend searching for a better understanding of nissan ecu's until the basic concept explained on those pages are understood. lol

Those are the pages I found. One would be jumping to conclusions to think that that info proves anything about the particular tables we are using. I suggest reading the entire section a little more carefully before jumping to conclusions. Other pages in that section clearly show throttle angle as a factor. Furthermore it is a huge mistake to assume that there is only one map for each process - in reality there most likely is not. Especially since the pages you cite say on no uncertain terms "revised according to other data."

Also, if we go back to the former understanding, the axis labled TP was assumed to be MAF signal rather than pulse width.

It's better to weigh the facts and formulate an opinion than to formulate an opinion and then look for facts to support it. I have to go but later I will show you how I can prove irrefutably that the maps we are dealing with are not TP Vs RPM.
Modified by Edub1 at 5:10 PM 1/23/2007

[Edub1]

Here is a stock fuel map. The original assumption was that the x axis represents air flow through the MAF and the Y axis represents RPM which is certainly incorrect. If we assume that the X axis is the actual pulse width which is somewhat more feasible, then we have the following situation.

Given, TP=(K)(maf)/RPM

Suppose we install 540cc injectors and an SAFCII which cuts the MAF signal and therefore the pulse width in half. Our "tp scale" remains unchanged.

If this is the case, the following must be true. Looking at our fuel map graph we can see how fueling would respond to changes in TP. With TP cut in half and our scale unchanged as in our example, the ECU will only scale to the halfway point of the x axis under WOT.

Clearly, the locations that would be accessed in this example would not allow the vehicle to run within normal operating parameters. The same would hold true if K is altered and the X axis left unchanged.

The exact same thing would happen with the timing map if the TP cut in half and the X axis left unchanged. Only the left half of the map would be used and the car would not run well if at all.

The fact is that many people use SAFCs which cut MAF signal and therefore TP in half, but do not experience these types of problems.

The only explaination is that though the ECU might plot TP against RPM at some point, our maps are defenitly not simple TP Vs RPM tables that determine fuel and ignition without other input.

Look at the first half of the map Vs the second.

[Bernard]

Here is a stock fuel map. The original assumption was that the x axis represents air flow through the MAF and the Y axis represents RPM which is certainly incorrect. If we assume that the X axis is the actual pulse width which is somewhat more feasible, then we have the following situation.

Given, TP=(K)(maf)/RPM

Suppose we install 540cc injectors and an SAFCII which cuts the MAF signal and therefore the pulse width in half. Our "tp scale" remains unchanged.

If this is the case, the following must be true. Looking at our fuel map graph we can see how fueling would respond to changes in TP. With TP cut in half and our scale unchanged as in our example, the ECU will only scale to the halfway point of the x axis under WOT.

Clearly, the locations that would be accessed in this example would not allow the vehicle to run within normal operating parameters. The same would hold true if K is altered and the X axis left unchanged.

The exact same thing would happen with the timing map if the TP cut in half and the X axis left unchanged. Only the left half of the map would be used and the car would not run well if at all.

The fact is that many people use SAFCs which cut MAF signal and therefore TP in half, but do not experience these types of problems.

The only explaination is that though the ECU might plot TP against RPM at some point, our maps are defenitly not simple TP Vs RPM tables that determine fuel and ignition without other input.

Look at the first half of the map Vs the second.

Lol, that is a damn good explanation of how it works. Twice the injector takes half the pw for the same fuel. You did leave out the part where you can change the correction of the maf signal. It's also a good example of why you can run into timing issues when the power starts going up with the safc. Well done.

[Bernard]

Edub1 and Bernard: What would you suggest that I alter in order to achieve a more accurate AFR under 0 vacuum conditions?

NateDogg, I read your question wrong the first time. You need to work on the k value and latency. You will never be able to get the commanded afr to read out exactly the same on the wideband, but it should be within .5 or so. You can also increase ttpmax to 254 in the appropriate column.

[KATwo40]

Good explanation Edub!

Here's what I'm gonna go with:

Open ROM Editor.Change stuff until the car runs well and safely.Save.Drive car.

These threads are great, but remind me of the "X AFR is better than Y AFR because science says so", when actually, you just give the engine what it wants, regardless of what science says.

So, unless I missed some sarcasm, I think you've turned Bernard over to the dark side!

[Bernard]

Good explanation Edub!

Here's what I'm gonna go with:

Open ROM Editor.Change stuff until the car runs well and safely.Save.Drive car.

These threads are great, but remind me of the "X AFR is better than Y AFR because science says so", when actually, you just give the engine what it wants, regardless of what science says.

So, unless I missed some sarcasm, I think you've turned Bernard over to the dark side!

Yep you missed the sarcasm. If you recall in Edub's previous post he points out how the safc wouldn't work. Then he maps out exactly how it would work if the afc did change the tp scale by lower maf voltage. Which is exactly what it does!!! The funny part is he's now saying that the ecu might plot the map based on tp and rpm but that there is more to it. Which is exactly what the patent also says!! It adjusts via previously determined corrections. He just cannot grasp that the theoretical pulse is used to select the load column. He doesn't won't recognize a clear and concise statement in the fsm. I don't think it matters much if you have a low power car, but if you want to run 11's with a romtune then you'll get yourself in big trouble pdq if you entertain edubs theories. Anyway I enjoy the back and forth it's winter up here and it's all in good fun.

[Edub1]

Yep you missed the sarcasm. If you recall in Edub's previous post he points out how the safc wouldn't work. Then he maps out exactly how it would work if the afc did change the tp scale by lower maf voltage. Which is exactly what it does!!! The funny part is he's now saying that the ecu might plot the map based on tp and rpm but that there is more to it. Which is exactly what the patent also says!! It adjusts via previously determined corrections. He just cannot grasp that the theoretical pulse is used to select the load column. He doesn't won't recognize a clear and concise statement in the fsm. I don't think it matters much if you have a low power car, but if you want to run 11's with a romtune then you'll get yourself in big trouble pdq if you entertain edubs theories. Anyway I enjoy the back and forth it's winter up here and it's all in good fun.

No, you have failed to comprehend what I explained just as you have failed to comprehend what the FSM said. You might want to try reading a little slower and paying more attention. Also, try not being sarcastic when you are demonstrably confused.

The first error you have made is that the FSM states in plain English that several factors determine timing and fueling. In fact, it specifically mentions throttle position as one of these factors.

As for the SAFC example you have it completely backwards. The SAFC alters the MAF signal and hence pulse width. What it does NOT do is alter the "load" scale in the map. The former theory claims that the X axis values must be scaled by the same amount as K because like the altered MAF signal, pulse width(TP) will vary directly with K. This was assumed to be neccessary because a given pulse width allegedly denotes the location along the x axis.

The SAFC examples proves that this is wrong. It is wrong because the device cuts the MAF signal which directly cuts pulse width. If pulse width denotes the location on the x axis and TP is halfed only the first half of the maps would be used.

Because one can run an SAFC and clearly utilize the entire map and not just the first half, the notion that pulse width determines the place along the x axis is blown out of the water. I have also proven this on my car by altering K and keeping the stock TP values. Not only did the ECU scale the x axis at the same rate, but it did this at all boost levels.

If you take the time to consider what I'm saying, you'll see that the old theory isn't in doubt, it's flat out disproven.

Now, there might be some base values are determined by TP Vs. RPM. There might even be an unknown map of this nature. But, there are a number of other factors involved and our maps are probably just one of these.

Even as a simple matter of mechanics, load which is properly defined as the motor's attempt to accelerate, must be one of the main determinants of timing due to the change in VE.

Yes, there is some science involved here but it's backed up by reality. I have a car with 460cc injectors and a Z32 MAF. My TP values are stock and my map is accessed just like stock at 0 boost or at 17lbs. And my car hauls a$$!

The main error everone is making with that map trace is to see the effect of the throttle position sensor voltage and mistake it with the MAF voltage.

[Bernard]

Under the basic fuel injection heading the last sentence says that " in other words, the program value is preset by engine operating conditions determined by input signals (for engine speed and airflow) from both the crank position sensor and the mass airflow sensor. This is described as TP in the patent. The next heading is Various Fuel Injection Increase/Decrease - it states that the amount of fuel is compensated for improved engine performance under various conditions: It then lists several conditions where it adds fuel: (these are the corrections stated in the patent - they are allowed to increase fuel up to the ttpmax)

Warmup startingduring accelhot engine

The next list is where it decreases fuel:

deceleration (imagine that)

Explain to me how or where you get whatever it is you're disagreeing on the fuel injection? I cannot for the life of me not see how one would say that basic fuel injection is not determined by the maf/rpm and the "preset" (k value) as stated in the tp equation in the patent.

The system description of the ignition states that it uses the detected fuel injection pulse and crankshaft position sensor ....

Below that is a table labels Injection Pulse Width and Tp (msec) on one axis and rpm on the other.

The injection pulse width is used so that the same calc'd load for fuel can be used for timing. Explain to me why nissan would put these statements into the FSM if they did not hold true?

What are you using for maptracing and datalogging?

What is your k value and latency? Are your knock fuel/timing maps both the same? Is all your emissions stuff installed and working? Do you have any codes? How much timing are you calling for and what is the ecu putting out?

I await comments.... from you or anyone else for that matter.

PS. Have you read the cherrypicker ecu tuning guide? If you have you must think it's a load of crap.

[Edub1]

I have two electronic FSMs, one for the E which doesn't show any of the things you are mentioning and one for the DE which shows some of what you are describing. The one for the DE does say that there is a map that determines what is implied to be a base timing. I am not disputing this. The problem is in jumping to the conclusion that the map we have access to is the map mentioned. This is because the ECU says and I quote,

"IN ADDITION TO THIS"

StartingWarm upIdleSwirl valve operationHot operationAcceleration

"the ignition timing is revised by the ECM according to other data stored in the ECM.

The major problem with relying on maptrace is that there are at least 7 variables mentioned that DO change the total timing. Furthermore, the map you trace might be an acceleration map that is used in conjunction with the base timing map to determine overall timing.

Now let me explain why putting a car on a dyno and bring it to a certain running load while watching maptrace does not prove anything. When you bring the car to a given RPM and load, you are doing two things, one is increasing airflow and the other is increasing throttle position. As stated in the FSM, both variables send info to the ECU to determine timing. One can not claim to have done a proper experiment unless one is able to isolate a single variable. Because you have not done so, you can not claim to have conducted a valid trial.

How do you know that our map is not an acceleration advancement map and that you are not seeing the result of the TPS output? Remember, due to the nature of the motor a rise in TP and a rise in TPS output are going to happen somewhat simultaneously.

In order to answer this question for sure, one would need to somehow devorce the two sensors and test each individualy.

As for the base TP, it is determined by TP=(K)(MAF)/RPM as I have stated 4 or 5 time. Then the ECU accesses our fuel map to retrieve a 3rd value that is most likely based on throttle position and the need for acceleration. This was my original argument.

But this was not the former beliefe. The former theory IYRC held that the TP is actually determined by values on our fuel map which is MAF Vs RPM and that the X axis that is labeled TP is in fact MAF reading. The SAFC example clearly proves this is incorrect. And to some extent it proves that timing is likewise.

Given all these descrepancies, I have always argued, it is unwise to make a similar assumption with regard to timng. Might the ECU determine timing from a base TP, sure but we can not assume that the map we are looking at is that map or that the end result is know because we have watched the ECU scale our map with input from to different sensors.

Personally, my BTM tune kickes butt and I don't even care. But untill someone finds a way to isolate each variable nothing is proven

Try setting your TP scale to stock values and then maptrace, I'll bet there is no difference in cell accessing.

[Edub1]

I re-read some of the Cherrypicker article which does confirm most of what I'm saying about the fuel map. Except, I have noticed that changing the TP values fails to change the way the columns are accessed and the SAFC example speaks for itself. Using a SAFC to correct for injectors would, as a rule cause exceedingly lean conditions accross the board and exceedingly advanced timing. Is this the popular finding?

As for the timing map, the artical does not have a complete explaination. It is possible that the map is accessed in this manner, I'm just not going to trust it untill it is proven. Plus, my stock TP values seem to cause my car to access the maps unchanged - in fact my car runs 1000% beter than when I tried to scale them.

What makes that guy an infallible authority anyway?

[Bernard]

I have two electronic FSMs, one for the E which doesn't show any of the things you are mentioning and one for the DE which shows some of what you are describing. The one for the DE does say that there is a map that determines what is implied to be a base timing. I am not disputing this. The problem is in jumping to the conclusion that the map we have access to is the map mentioned. This is because the ECU says and I quote,

"IN ADDITION TO THIS"

StartingWarm upIdleSwirl valve operationHot operationAcceleration

I can view those maps in realtime and see what they are doing with nistune. I don't know what kind of software you are using, but you can edit those maps anytime you want and view the changes you make in the ram output.

"the ignition timing is revised by the ECM according to other data stored in the ECM.

The major problem with relying on maptrace is that there are at least 7 variables mentioned that DO change the total timing. Furthermore, the map you trace might be an acceleration map that is used in conjunction with the base timing map to determine overall timing. That is a far cry from your first post claiming to have found that the maps used throttle angle and not mass airflow by having read a patent

Now let me explain why putting a car on a dyno and bring it to a certain running load while watching maptrace does not prove anything. When you bring the car to a given RPM and load, you are doing two things, one is increasing airflow and the other is increasing throttle position. As stated in the FSM, both variables send info to the ECU to determine timing. One can not claim to have done a proper experiment unless one is able to isolate a single variable. Because you have not done so, you can not claim to have conducted a valid trial.

I don't have to disconnect the tps. My car doesn't use it except for idle anyway. I am able to effect the afr by merely changing the maptraced load cell. Same thing for timing. Every single time. I can also change the accessed tp scale by editing #'s in the vq table. I can change it by altering the kvalue, ttpmax, all that stuff. So have many others. I have done this with maxima's, skylines (both gtr's and gts's), ka's, L28's, 300zx's, I tune a few ecu's a month. They all behave the same way

How do you know that our map is not an acceleration advancement map and that you are not seeing the result of the TPS output? Remember, due to the nature of the motor a rise in TP and a rise in TPS output are going to happen somewhat simultaneously.

That's why the ecu adds tps enrichment. The patent states at certain throttle angles the enrichment is greater than others. The accel enrich table supports this as well. The values are greatest at the lower end of the scale and then fall off.

In order to answer this question for sure, one would need to somehow devorce the two sensors and test each individualy.

This what you should have done BEFORE making any claims.

As for the base TP, it is determined by TP=(K)(MAF)/RPM as I have stated 4 or 5 time. Then the ECU accesses our fuel map to retrieve a 3rd value that is most likely based on throttle position and the need for acceleration. The ecu accesses the fuel map for a correction value. That value can be an increase or decrease or no change and is used with the other corrective values to provide a final injection pw. This was my original argument.

But this was not the former beliefe. The former theory IYRC held that the TP is actually determined by values on our fuel map which is MAF Vs RPM and that the X axis that is labeled TP is in fact MAF reading. The SAFC example clearly proves this is incorrect. And to some extent it proves that timing is likewise.

TP has never been thought to be determined by values on the fuel map. I have no idea where you got that from. The axis labeled TP has been a calc'd pw derived from the k value, maf and rpm. You showed exactly how the safc works and why it works somewhat. There is a reason why the safc can work with large injectors, but it has nothing to do not using the maf.

Given all these descrepancies, I have always argued, it is unwise to make a similar assumption with regard to timng. Might the ECU determine timing from a base TP, sure but we can not assume that the map we are looking at is that map or that the end result is know because we have watched the ECU scale our map with input from to different sensors. There's a huge difference from stating that you've found evidence and can prove it works differently beyond a doubt and then changing it to a I'm not sure because I think it is wrong or it .. it might work this way. All this from someone who is new to the ecu tuning world, and has given up tuning his ecu because he couldn't get it to do what he thought it should be doing or whatever the reason is. Newbies read this stuff and may not realize that you are not sure what you are doing and take your posts as the truth, thereby costing them a lot of trouble and potentially cash.

Personally, my BTM tune kickes butt and I don't even care. But untill someone finds a way to isolate each variable nothing is proven

You haven't even tuned your own ecu? And you want to tell others how it works?

Try setting your TP scale to stock values and then maptrace, I'll bet there is no difference in cell accessing.

I have rescaled many ecu's for more resolution by expanding the maps. It's never mattered where the tp value is, when the ecu calc's that value, it accesses that scale. It's works everytime, I wouldn't be stating my case if this wasn't what I experienced.

[KATwo40]

Bernard, if you wouldn't mind, could you comment on my post in EFI101, regarding incorrect latency?

Thanks.

[Edub1]

You've done all that on an E motor?

What software are you using that allows you to expand the maps and adjust latency? Can I have the address file?

You're asking me where I got the notion that people believe that the maps are MAF Vs. RPM? If you go to hybridka in the E section that is what is stated repeatedly.

I must admit to being somewhat humbled by your experience so I guess I'll just ask.

If a guy has an NA motor and he doubles his injector size and uses an SAFC to correct for these, therefore cutting TP in half, how is his ECU able to access the entire map? Or does the car run just fine only using the 1st half of the maps?

Also, using romedit I corrected K for 460cc injectors and a Z32 MAF. My car ran like crap with the TP scales adjusted to match. When I put them back to stock my problem was solved. While tuning, I noticed by watching my wideband that a given spot on my fuel map was accessed at the same time regardless of airflow. That is to say, with my WG tied open I had a lean spot at a given RPM and throttle angle. Under boost, my ECU did not shoot passed the lean spot as expected but instead passed through the lean spot exactely as it had previously. I fixed the lean spot and observed again that my ECU tracked the fuel map at a steady rate regardless of boost.

Just for kicks, I tried increasing the TP values of my last few fuel columns by dramatic amounts - nothing changed in the way my car accessed the fuel map.

Why is this the case? Why are other people saying not to alter the TP values?____________________________________________________

Upon further consideration I noticed that the equation for TP, TP=(K)(maf)/RPM simplifies thusly,

Where N=RPM, K is constant and VE is volumetric efficiancy.

flow = 2.4l (VE) * N

TP=(K)(flow)/N

TP= [2.4l (VE)N]/N

TP=[2.4l(VE)]

TP=VE

So, it is does make sense that the TP scale or x axis relates to the motor's deviation in TP from the base which is related to the change in VE which is mostly related to throttle angle. Of course a turbo will add another dynamic.

So, it does make sense that the maps use TP as their x axis and use the cell values as appropriate modifiers to a base TP or base timing. It also seems that the range of the map is small with regard to the overall calculations. Much like a fine trim rather than a master setting.

However, there still seems to be an issue with proper scaling of the E motor.


Modified by Edub1 at 2:35 PM 1/25/2007