Tuning - Nissan ECU patent info

[Edub1]

For those following discussion on ECU tuning. Here is some patent info regarding Nissan ECUs.

What is particularly significant here is that this info clearly states that TP is based on throttle angle and not air flow. Air flow or MAF signal is used to determine the correct A/F ratio but not map position.

This is made quite clear in the patent info if the document is read carefully.

http://eccs.hybridka.com/viewtopic.php?t=35
Modified by Edub1 at 7:11 PM 1/18/2007

[KATwo40]

Ok, lemme make sure I read that correctly...

According to what I read, the ECU looks at throttle angle and RPM for the lookup table.

Then, it corrects for O2, coolant temp.

This would indicate that it uses throttle angle and rpm as the guiding inputs for the correction tables.

However, what about the box below that, where it clearly states TP = KQa/N?

Does that not say that TP is calculated according to K value, MAF input and speed?

[Edub1]

OK, I read through it all the way and it seems that there are some complicated check mechanisms to make sure everything is within normal perameters.

I think what you are looking at is that the ECU first looks at throttle angle Vs rpm to find the spot on the map, then it looks at the other factors and calculates how much fuel is needed.

I think where we are all confused is that we have been thinking of the cell values as the end factor in determining fuel when really they are the instructions.

Put another way, say a given cell set to 45 gives 12:1 A/F ratio. The ECU looks at throttle angle & rpm to land on our cell. Now it says, "ok I see from this 45 that I need to make things 12:1 in this spot." Then it looks at how much air is flowing (MAF) and water temp, and battery voltage, etc and calculates the amount of fuel needed to make 12:1 with the conditions at hand.

You see, the map value is not the ultimate determinate of fuel, it is the initial instruction to tell the ECU what needs to happen at that spot. Then MAF reading comes in and the rest is plug and chug.

[KATwo40]

In reading the summary, which breaks down the process, step by step, I found the following:

" A step 111 subsequent to the step 107 calculates or determines a basic desired fuel injection quantity Tp during a coming fuel injection stroke in accordance with the air flow rate Qa and the engine speed N by referring to the following equation.

Tp = K Qa/N

where K is a preset constant.

A step 112 subsequent to the step 111 calculates or determines a final desired fuel injection quantity Ti during a coming fuel injection stroke in accordance with the basic desired fuel injection quantity Tp and the corrective factors KMR, KAC, KTw, a, and Ts by referring to the following equation.

Ti = a - Tp(1 + KMR + KAC + KTw) + Ts

A step 113 following the step 112 controls the fuel injection signal Si in accordance with the final desired fuel injection quantity Ti. Specifically, the duration or pulse width of a fuel injection pulse Si is adjusted in accordance with the final desired fuel injection quantity Ti so that the actual fuel injection quantity will be equal to the final desired fuel injection quantity. After the step 113, the fuel control segment of the program ends."

What does this mean? It means that the ECU uses MAF, K value AND RPM to determine TP.

This tells me that the tuning axis which is normally labeled TP is incorrect. How can you cross reference TP with RPM when RPM is a function of TP?

So, I backed up a few paragrahs and had another go at it. Here's what I found:

"The step 110 determines the corrective factor KAC in accordance with the difference *diff. The ROM 20C of the control unit 20 holds a table in which a set of non-zero corrective factors KAC are plotted as a function of difference *diff. Specifically, the corrective factor KAC increases with the difference *diff. The corrective factor KAC is derived by referring to this table. After the step 110, the program advances to the step 107."

Since *diff is a function of throttle angle, now we're seeing that this table is a function of RPM vs throttle opening.

But wait! There's something else!

"A step 103 following the step 102 compares the throttle opening degree * to the reference throttle opening degree *N. (note here that N = RPM) When the throttle opening degree * is greater than the reference throttle opening degree *N, that is, when the throttle valve 9 is open at a degree higher than a reference degree, the program advances to a step 104. When the throttle opening degree * is equal to or less than the reference throttle opening degree *N, that, when the throttle valve 9 is open at a degree equal to or less than a reference degree, the program advances to a step 105."

Let's forward to step 105...

"The step 105 derives the current engine load EL from the air flow rate Qa or both the air flow rate Qa and the engine speed N. It should be noted that the current engine load EL may be derived from a basic desired fuel injection quantity Tp or from both the engine speed N and the basic desired fuel injection quantity Tp which will be described hereinafter. The step 105 compares the engine load EL to a reference engine load ELref to detect whether or not the engine load EL is in a high value range. When the engine load EL exceeds the reference engine load ELref, that is, when the engine load EL resides in a the high value range, the program advances to a step 108. When the engine load EL is equal to or less than the reference engine load ELref, that is when the engine load EL resides outside the high value range, the program advances to a step 109."

Here's the really juicy part...

"The step 108 determines the corrective factor KMR in accordance with the engine load EL. The ROM 20C of the control unit 20 holds a table in which a set of non-zero corrective factors KMR are plotted as a function of engine load. Specifically, the corrective factor KMR increases with the engine load. The corrective factor KMR is derived by referring to this table. After the 108, the program advances to the step 107."

So, the best I could get from this whole deal is this:

The ECU looks at throttle angle vs RPM to determine if the engine is accelerating, decelerating or just cruising along steadily.

If the actual throttle angle is in sync with the angle shown in the refernce table, it moves on to determine engine load.

If it sees high engine load operation, it uses Load (MASS air flow rate) vs RPM. If it sees not-so-high load operation, it uses throttle position vs RPM.

Modified by KATwo40 at 8:39 PM 1/18/2007

Modified by KATwo40 at 8:41 PM 1/18/2007
Modified by KATwo40 at 8:42 PM 1/18/2007

[KATwo40]

OK, I read through it all the way and it seems that there are some complicated check mechanisms to make sure everything is within normal perameters.

I think what you are looking at is that the ECU first looks at throttle angle Vs rpm to find the spot on the map, then it looks at the other factors and calculates how much fuel is needed.

I think where we are all confused is that we have been thinking of the cell values as the end factor in determining fuel when really they are the instructions.

Put another way, say a given cell set to 45 gives 12:1 A/F ratio. The ECU looks at throttle angle & rpm to land on our cell. Now it says, "ok I see from this 45 that I need to make things 12:1 in this spot." Then it looks at how much air is flowing (MAF) and water temp, and battery voltage, etc and calculates the amount of fuel needed to make 12:1 with the conditions at hand.

You see, the map value is not the ultimate determinate of fuel, it is the initial instruction to tell the ECU what needs to happen at that spot. Then MAF reading comes in and the rest is plug and chug.

I'll be doing some map tracing with different TP scales to verify this, but I think I get what you're saying. It certainly appears that the ECU is using the table as a target, not a modifyer.

[Edub1]

I wasn't able to fully understand what all the checks were for but it seems like they are there for non-standard conditions like a bad sensor or something of that sort.

And, there is no question that MAF readout largely determines fueling. But it is quite clear that the overall method of accessing the main fuel map is throttle angle Vs RPM. This is stated on the first page.

I bet the laod Vs rpm scale is the VQ scale.

Anyway, I have clearly observed my ECU follow the x axis according to throttle angle. Just what this load table does is beyond me but it's not the main fuel table.

[NateDogg]

When I'm cruising up a hill at 0 vacuum, my target AFR is much richer than my actual AFR. For example, I have a target of 12.8:1 in the fuel map but I am achieving approximately 14:1 AFR.

Does this mean that the VQ table of the Z32 maf is off or is there another way to correct this?

[KATwo40]

Did you use maptrace to find that you're in that particular cell?

Is the O2 function enabled there?

So long as you imported the VQ scale from a Z32 MAF, it really can't be wrong, per se.

This goes back to what I was pointing out earlier. Sometimes the ECU uses engine load and sometimes it uses throttle position degree as the X axis in the correction table.

Additionally, there's a load enrichment process that happens AFTER the engine load lookup table is referenced. That might be another issue for ya'.
Modified by KATwo40 at 7:24 AM 1/19/2007

[Edub1]

When I'm cruising up a hill at 0 vacuum, my target AFR is much richer than my actual AFR. For example, I have a target of 12.8:1 in the fuel map but I am achieving approximately 14:1 AFR.

Does this mean that the VQ table of the Z32 maf is off or is there another way to correct this?

Could be on account of closed loop or, this could be one of those situations that sets off one of those check values. Because throttle position would hold constant but RPM not increase, so it might fail a test and the ECU might look to load to compensate. Interesting.

Would your A/F be on target under normal cruising conditions?

[NateDogg]

Closed loop the target and actual AFR are accurate.

Open loop the two AFR are only accurate on full throttle.

[KATwo40]

Closed loop the target and actual AFR are accurate.

Open loop the two AFR are only accurate on full throttle.

Which is congruent to what we read in the description.

Unless you're WOT or cruising, it looks at some "other" mapping and pulls a number out of it's arse for the injection pulsewidth.

AEM lookin' better every day. LOL

[Edub1]

Closed loop the target and actual AFR are accurate.

Open loop the two AFR are only accurate on full throttle.

Interesting. So, this only happens when going up hill. That kind of makes sense because the ECU does make a comparison involving load, ta and rpm and weighs it against a set value. So here it might see that ta Vs rpm isn't cutting the mustard and it looks to another table.

It's starting to seem like our ECU does a lot more than first suspected. Now I wonder how it's handeling timing.

[KATwo40]

^^ Agreed. (about the timing)

[KATwo40]

Edub, didn't you recently have some driveability issues that you found to be a result of reducing the TP scale values in accordance with the reduction in K value to compensate for larger injectors? I think I recall you saying the car ran great when all you did was change the K value and leave the TP scaling alone.

[eazye2000]

Edub, didn't you recently have some driveability issues that you found to be a result of reducing the TP scale values in accordance with the reduction in K value to compensate for larger injectors? I think I recall you saying the car ran great when all you did was change the K value and leave the TP scaling alone.

I'm just getting into this tuning stuff and I'm sooo lost! K values, VQ maps, just stick a hot p0ker in my eye for fvks sake!! I'm working on it though!

[Bernard]

I have briefly read the linked patent. It looks to me like the ecu is checking to see if it needs to add momentary pulse width based on a preset compare of the present throttle angle to maf voltage. (pumpshot in carbspeak) If maf voltage is this, then the throttle angle (voltage) should be this, if not calculate correction factors based on this. The ecu still uses the maf/rpm/k value as the heart of it's calc's. The fuel maps (how many are there for the KA again?) are a correction in addition to a calc'd injection pulse width. What you might be seeing is that when going up a hill, the enrichment is activated, and without correcting the value for the larger injectors, you'll be going rich. On the other hand, who has datalogged a stock setup to see if the car was not tuned like this in the first place? Some food for thought.

[KATwo40]

I have briefly read the linked patent. It looks to me like the ecu is checking to see if it needs to add momentary pulse width based on a preset compare of the present throttle angle to maf voltage. (pumpshot in carbspeak) If maf voltage is this, then the throttle angle (voltage) should be this, if not calculate correction factors based on this. The ecu still uses the maf/rpm/k value as the heart of it's calc's. The fuel maps (how many are there for the KA again?) are a correction in addition to a calc'd injection pulse width. What you might be seeing is that when going up a hill, the enrichment is activated, and without correcting the value for the larger injectors, you'll be going rich. On the other hand, who has datalogged a stock setup to see if the car was not tuned like this in the first place? Some food for thought.

Glad to have another thinker in here on this.

What I found in the description is that there are circumstances that cause the ECU to set the correction factors from the tables to 0, using a preset enrichment. After these, it goes directly to the next step, which is calculating a TP, but not referencing the table again.

[Bernard]

Glad to have another thinker in here on this.

What I found in the description is that there are circumstances that cause the ECU to set the correction factors from the tables to 0, using a preset enrichment. After these, it goes directly to the next step, which is calculating a TP, but not referencing the table again.

where in the flow chart does the ecu reference the map? i see it as the point where Ti is calculated. And who can say for certain this patent applies to the KA ecu? Is anyone in this thread using nistune for maptracing and viewing what the other tables are doing?

[KATwo40]

where in the flow chart does the ecu reference the map? i see it as the point where Ti is calculated. And who can say for certain this patent applies to the KA ecu? Is anyone in this thread using nistune for maptracing and viewing what the other tables are doing?

"The step 110 determines the corrective factor KAC in accordance with the difference *diff. The ROM 20C of the control unit 20 holds a table in which a set of non-zero corrective factors KAC are plotted as a function of difference *diff. Specifically, the corrective factor KAC increases with the difference *diff. The corrective factor KAC is derived by referring to this table. After the step 110, the program advances to the step 107."

"The step 108 determines the corrective factor KMR in accordance with the engine load EL. The ROM 20C of the control unit 20 holds a table in which a set of non-zero corrective factors KMR are plotted as a function of engine load. Specifically, the corrective factor KMR increases with the engine load. The corrective factor KMR is derived by referring to this table. After the 108, the program advances to the step 107."

This is the Nissan ECCS system from the US Patent office, which applies to Altima, Sentra, Stanza, 300zx, 240sx.

[Bernard]

"The step 110 determines the corrective factor KAC in accordance with the difference *diff. The ROM 20C of the control unit 20 holds a table in which a set of non-zero corrective factors KAC are plotted as a function of difference *diff. Specifically, the corrective factor KAC increases with the difference *diff. The corrective factor KAC is derived by referring to this table. After the step 110, the program advances to the step 107."

"The step 108 determines the corrective factor KMR in accordance with the engine load EL. The ROM 20C of the control unit 20 holds a table in which a set of non-zero corrective factors KMR are plotted as a function of engine load. Specifically, the corrective factor KMR increases with the engine load. The corrective factor KMR is derived by referring to this table. After the 108, the program advances to the step 107."

This is the Nissan ECCS system from the US Patent office, which applies to Altima, Sentra, Stanza, 300zx, 240sx.

Those are corrective values applied to the basic injection quantity in step 111-112 are they not? I see the references to the 'table' but that's not used to calc the fuel, it is used to determine if the car is accelerating/cruising or decelerating is what is looks like to me. The rest of the document repeatedly states the "first time enrichment" when the detected throttle position is greater than a predetermined value.

[KATwo40]

After step 107, it goes to step 111, then 112, which, after looking at the formula again, seems to use KMR and KAC (both the corrective factors found in either lookup table) to determine final injection pulse.

Also, Edub (and a couple others on the ecu2 forum) have found that, regardless of airflow, the ecu referenced the exact same plot on the map for a given throttle angle and rpm.

He traced at both 7psi and 14psi.

[Edub1]

I wanted to hold off on this while studying other Nissan patents.

The patent we are discussing outlines a system for correcting the following.

When the throttle plate is opened by some moderate angle a significant change in flow takes place but when it is opened farther, a less significant change in flow takes place. The change is not linear and causes a rich condition. In order to get around this problem, Nissan put in place the check systems to weigh air flow agaist throttle angle and add correction values when neccessary. That's all it does.

The patent never mentioned the main fuel map because it only reffers to the "checking" technology, but from the bulk of the discussion it is fairly clear that the ECU uses air flow, K and RPM to determine a base pulse width for the full range of operation. This base is farther modified by things live battery voltage, temp, or the other afformentioned check values which are going to be set null for most purposes. The fuel table we are all concerned with is just another modifying factor. The ECU uses throttle angle and RPM to look up this last modifier.

The important thing to understand is that the Air flow, K and RPM work as a formula that runs in the background to maintain a stoich TP. Our fuel map simply comes in after the fact to throw a modifier into the equation for a given throttle angle and RPM.

This is why piggybacks work, because when we have a*b*c = d we can change any variable a,b or c and have a directly proportional change in d. IIRC, this is the transitive property of multiplication. And this makes a global change without any affect on map location or scaling because the map is not part of the base TP.

Also, it seems that many people use the term "load" to reffer to throttle angle or air flow, which are somewhat connected.

Anyway, in the end our fuel map is just a map of modifiers tossed into the standard TP equation and I believe timing is a simple lookup based on throttle Vs RPM. Other Nissan patents have suggested this is so with refference to this being a generic meathod of the prior art found in other Japanese patents.

As for my TP values. I did have some trouble when modifying them. When experimenting with the values to the right in the boost areas, I was unable to affect any change regardless of how much I altered them.

[Bernard]

After step 107, it goes to step 111, then 112, which, after looking at the formula again, seems to use KMR and KAC (both the corrective factors found in either lookup table) to determine final injection pulse.

They are correction factors to the basic injection value (tp)... At least that's what the patent says. At times the factors are set to zero, but still go through the same equation. How much weight do you think they have in the overall scheme of things?

Has anyone used the safc with the tps signal to correct for larger injectors?

[KATwo40]

So, basically, our lookup table is the equivalent of an Alpha N system?

Sweet.

Thx Edub for adding info on this. In the end, yes, MAF voltage is a major determinate of TP, but TP is NOT the X axis in our mis-labeled lookup tables.

I think I'll do some playing with my TP scaling this week. I've got a nasty little idle issue that's been present ever since I reflashed for the larger injectors, despite the AFR's being dead on. I'm blaming one of two things: Latency is incorrect, OR TP scaling is wack.

[KATwo40]

After step 107, it goes to step 111, then 112, which, after looking at the formula again, seems to use KMR and KAC (both the corrective factors found in either lookup table) to determine final injection pulse.

They are correction factors to the basic injection value (tp)... At least that's what the patent says. At times the factors are set to zero, but still go through the same equation. How much weight do you think they have in the overall scheme of things?

Has anyone used the safc with the tps signal to correct for larger injectors?

They have a great role, as they are used in the final calculation of the ending pulsewidth (Si). However, as we've agreed, and as the patent says, they're sometimes null, at which point they don't do squat in the final injection calculation.

Either way, no matter how it's interpreted, I see no way to dispute the lookup table being anything more than an alpha N system (RPM vs TA), rather than being a true load system (RPM vs TP), since findings indicate that map trace in unaffected by changes in TP scaling and/or boost increase, but rather by TA, according to Edub's tracing.

[Bernard]

ok then. I suggest you get another verification of the maptracing from another source before deciding that things are what you thing they are. I'll leave it at that.

[KATwo40]

ok then. I suggest you get another verification of the maptracing from another source before deciding that things are what you thing they are. I'll leave it at that.

What map tracing have you done to solidify your opinion? If you've done any, why not contribute to this thread instead of playing the same old song?

[eazye2000]

What map tracing have you done to solidify your opinion? If you've done any, why not contribute to this thread instead of playing the same old song?

[Bernard]

What map tracing have you done to solidify your opinion? If you've done any, why not contribute to this thread instead of playing the same old song?

For gods sake man, it says it in plain english in FSM .....how stubborn can you be. I really don't want to argue with you, if you think that's how the ecu works, then fine, have fun with it, but don't claim that it's rock solid proof just because.

I have several megs of datalogs. That's why I was so interested in what you guys found, I see no proof of anything so I'll let you tune your alpha n nissan ecu in peace.

[KATwo40]

What are you talking about with the FSM?

I'm not trying to argue with you here. I'm hoping at some point you'll say, "Ok, here's a shot of a maptrace at X airflow/boost, and here's a shot of another maptrace at 2X airflow/boost. See how they read different areas of the map?"

I'm not saying the entire ECU is an alpha N, but rather that I'm interpreting at least ONE form of the lookup table to be. Even at that, the ECU looks at MAF voltage later to calculate TP.

I've never once said "just because" in any of my delving into this subject. In fact, every time I've had an objection, or question, I've provided quoted literature from the Nissan patent writeup.

Just tryin' to get to the bottom of it, so as to not be blowin' up motors.