Fuel System Chaulk Talk - Long Term Fuel Trim

[landtank]

Rick, we never run them for long enough to get that. After getting the trucks running, we start the tuning immediately in baby steps to ensure I do not have to buy the customer a truck.

My 100 series is running for 10k with the same setup as the 80 (different tuning) and no check engine lights. We have another 80 running with a customer that has about the same amount of miles, with no issues either.

The reason for the question is as follows. I've worked on countless systems with feed back loops for various reasons. The one thing common to all of these systems is that you need to disable the feedback circuit prior to calibrating.

If you don't you tend to chase your ass around. The system is adjusting while you are adjusting.

The only way I could see us doing this is to create a fault that puts the ECM into limp mode where the truck runs open loop 100% of the time. That way you can accurately tune to the truck's needs.

I'm sure I'm correct in theory, not sure if this is possible in reality.

 

[sleeoffroad]

Or you could use a SMT7 and do closed loop tuning as well with a wideband o2 input.

Rick, I should have added that this is only necesarry if the truck can't keep up with the fuel demands. So if you monitor closed loop and the truck keep constantly adding fuel, to the extent that LTFT is our of spec, you can use the computer and extra injectors, to add fuel in those fuel map cells to get the LTFT to get back to 0%. The truck will only start backing out the fuel if you are adding to much with the additional tuning.

 

[landtank]

Or you could use a SMT7 and do closed loop tuning as well with a wideband o2 input.

Just to be sure I'm clear, the RR FPR has adjustments that effect the pressure during times when the truck is running in closed loop. But with that loop active you can't accurately set the reg, do to the fact that the ECM is adjusting DC of the injectors from the feedback of the O2 sensor.

Now with the SMT7 this would be exactly the same for it's adjustment unless the closed loop circuit was interupted. While the closed loop is active you might be able to address the ECM shortcomings but you're not able to tune the truck to the point where minimum input from the closed loop circuit is needed to obtain ideal results.

Another words, lets say at point X you need a value of 10. With closed loop active you have a 10. Do you have the abilty with the SMT7 to know how much the ECM is adjusting for it? If you don't have that abilty then you don't know what needs to be done.

Disabling the circuit would make that value appear un altered and you would know exactly what is needed so that when the circuit was activated again there would be no input from the ECM to alter that point as it has already been tuned to the correct value.

The end result at the time of tuning is there would be virtually no difference between active and inactive engine management.

Sorry everyone for the side track

 

[SUMOTOY]

The reason for the question is as follows. I've worked on countless systems with feed back loops for various reasons. The one thing common to all of these systems is that you need to disable the feedback circuit prior to calibrating.

If you don't you tend to chase your ass around. The system is adjusting while you are adjusting.

The only way I could see us doing this is to create a fault that puts the ECM into limp mode where the truck runs open loop 100% of the time. That way you can accurately tune to the truck's needs.

I'm sure I'm correct in theory, not sure if this is possible in reality.

Not sure why you'd want to do that, and it's not correct in theory or practice on OBDII application. Most shops *always* tune with full STFT and LTFT active on OBDII. Rick, what you are looking for is the least amount of correction to LTFT. Set the RR FPR (or put in injectors, whatever), datalog LTFT. If it hovers in the 0 +/-3% range in terms of corretion, you have the fuel dialed in pretty well. If you see 10% swing, you usually see it trending towards negative (installed components cause rich) or positive (installed components cause lean).

Open loop tuning during a closed loop operation isn't going to be accurate at all. As soon as you go back to closed loop, whatever you did will end up causing a change to the baseline fuel setting. A temp input signal can cause a total shift of timing/fuel maps, btst. To understand this better, you can do this testing now. Hook up your 2027 and try changing the fuel, datalog the LTFT. You will get comfortable quickly that the FTU isn't fighting a proper fuel mod, it just dials in the last 10%. However, you can certainly use the FTU to see if you are even close.

You can never get the fuel right for closed loop operation, and you won't for open loop either, because a disable of the FTU, isn't a fuel isolated disable. You are disabling ALL inputs (including load) that address fuel AND timing. "Limp" home mode is also a fuel and timing map. All these maps are based on calculations in the computer. Limp home tables normally ascribes a simple fixed DC to the injectors (usually a rich event over LTFT IME), and timing is dialed back from normal operation. So, if you *tune* in limp home mode, you won't be close when you put the computer back into the equation.

I've tuned a wide variety of OBDII car fueling systems this way. The only open loop 'tuning' I see, is when the FTU is open loop. As I stated in post #1, LTFT active is a good way to evaluate and tune a fuel mod. You will be chasing your tail more trying to be smarter than the 3D tabled maps in the ECU. I understand 0 calibration applied to testing. Here, '0 Calibration' would be tuning with 0 LTFT correction in closed loop operation.

ST

 

[landtank]

Sumotoy, I'm sure this is not a usual proceedure on vehicles but it's a must on everything else I have worked on.

The basic idea is that you can't calibrate an auto adjusting system if it's auto adjusting. Disabling the auto adjusting circuit is the only way to get a clear view of how much calibration is needed and where it's needed.

With the circuit active you only see the circuits short comings and not the need.

My daughters car has had a O2 fault for a while now and it's been driven in limp mode all along. The car drove fine (i just repaired it). Now if there was a wide band sensor on the vehicle you would see exactly where and by how much the ECM had to compensate.

Adjusting the SMT6/RR FPR at this point to get the closest to ideal readings on the wideband would produce the most accurate result.

I'm not saying this is how it is supposed to be done. But take some time and think about it, it is the logical way to acheive the best results.

 

[SUMOTOY]

Sumotoy, I'm sure this is not a usual proceedure on vehicles but it's a must on everything else I have worked on.

The basic idea is that you can't calibrate an auto adjusting system if it's auto adjusting. Disabling the auto adjusting circuit is the only way to get a clear view of how much calibration is needed and where it's needed.

With the circuit active you only see the circuits short comings and not the need.

My daughters car has had a O2 fault for a while now and it's been driven in limp mode all along. The car drove fine (i just repaired it). Now if there was a wide band sensor on the vehicle you would see exactly where and by how much the ECM had to compensate.

Adjusting the SMT6/RR FPR at this point to get the closest to ideal readings on the wideband would produce the most accurate result.

I'm not saying this is how it is supposed to be done. But take some time and think about it, it is the logical way to acheive the best results.

LT:
You have to shift your paradigm a bit. Forget the FTU is even there. Post #1 says use your -0- calibration of fuel based on LTFT values. If LTFT stays at 0, you have no ECU correction to your fueling mods. Isn't that exactly what you are looking for?

Limp home mode ASSUMES values of bad components. You can run in limp home mode in most vehicles indefinitely. That's NOT a baseline for adjusting fuel.

Fuel and timing maps are adjusted based on several inputs, ambient temp, engine temp, throttle position, acceleration, knock, 02 STFT and 02 LTFT. Again, I understand -0- calibration. If you take these inputs out, and intend on dialing in fuel, don't you have to assign these inputs a value? How do you account for lack of timing?

Again, I urge you to put in your 2027 and play with the fuel. You'll better understand my point. The FTU is a given, LTFT value output is the way Mr. T allows you to change the fuel system and make sure it falls into -0- calibration. In fact, by definition it's -0- fuel calibration with an upper and lower control limit of 15%.

You have the logic backwards IMO.

ST

 

[landtank]

so where do you read the LTFT value? And the STFT value as well.

 

[SUMOTOY]

so where do you read the LTFT value? And the STFT value as well.

Depends on which reader you have. You are looking for "Measuring Block Values" Then you should look for:
Front (or) 02 1 Bank 1, and O2 1 Bank 2 = STFT
Rear (or) 02 2 Bank 1, and 02 2 Bank 2 = LTFT

The reader I have for the audis let's you pick datalogging for all four. I tend to ignore STFT, because digitally, it jumps around too much. I have the option to put in four parameters, so I tend to look at LTFT, MAF and injector duration.

A summary clear of codes puts the LTFT value to -0-

Tweek away.

ST

 

[landtank]

Actually it's a lot easier than that. You can monitor the ECM activity through the Diagnostic port in the engine compartment with a digital meter. It breaks down the ECM's efforts into 5 units.

[0 volts] Maximum Lean, reduction of the fuel mixture from the BASIC FUEL CALCULATION

[1.25 volts] Minimum Lean, reduction of the fuel mixture from the BASIC FUEL CALCULATION

[2.5 volts] No correction at all to BASIC FUEL CALCULATION

[3.75 volts] Minimum Rich, increasing the fuel mixture above the BASIC FUEL CALCULATION

[5 volts] Maximum Rich, increasing the fuel mixture above the initial calculation

This is read through the E1 nad Vf pins anytime that the E1 and T1 aren't jumped.

While not as accurate as your reader it serves a purpose as when done calibrating you want the system to be reporting 1.25v minmum lean so you know you have plenty of room to add fuel.

Taking this reading prior to installing a SC or Turbo is a good way to check out you system in general. If it's at maximum rich then you should figure out why before continuing.

One group of guys actually designed a board for monitoring this signal with an led display. 5 leds of different color so they could predict a lean condition prior to it happening.

 

[SUMOTOY]

Actually it's a lot easier than that. You can monitor the ECM activity through the Diagnostic port in the engine compartment with a digital meter. It breaks down the ECM's efforts into 5 units.

Neat stuff, always good to remember that all that cool stuff on an engine is a voltage or interpreted signal to the FTU. However, I believe you'd want to datalog fuel more than pick a point. Many of the OBDII readers can datalog over time, and many can overlap the graphing of some of the output functions. This is good useful data to have in this thread Rick, but I'd highly suggest spending the money on an OBDII reader that can give you more freedom to play. Tweeking the LTFT tables is only part of the game. You might find timing, engine temp, knock, MAF, and injector duration to be helpful here as well.

ST

 

[landtank]

This is good useful data to have in this thread Rick, but I'd highly suggest spending the money on an OBDII reader that can give you more freedom to play. Tweeking the LTFT tables is only part of the game. You might find timing, engine temp, knock, MAF, and injector duration to be helpful here as well.

ST

I've actually worked out a plug and play MAF/Injector upgrade that will put the injector duty cycles back in the stock range and is turbo tuned to boot. But that's for another thread, I've high jacked this one enough I think.

 

[SUMOTOY]

I've actually worked out a plug and play MAF/Injector upgrade that will put the injector duty cycles back in the stock range and is turbo tuned to boot. But that's for another thread, I've high jacked this one enough I think.

Happy to let it go where it may. Another thought regarding your target calibration of 1.25volts..

1.25 volts as a baseline I might disagree with. That translates to 7.5% injector baseline rich 'correction' by the computer. That means you now have 7.5 +7.5/-22.5 window. I think you will find in practice you want to have the upper and lower control limit window equally. Especially in a truck where volumetric efficiency throught the rpm range takes quite a ride.

Interesting to see what you come up with on the MAF and injectors. More interesting to see if it will work, OBDII trickery without the FTU software code can surprise you sometimes.

ST

 

[landtank]

1.25 volts as a baseline I might disagree with. That translates to 7.5% injector baseline rich 'correction' by the computer. That means you now have 7.5 +7.5/-22.5 window. I think you will find in practice you want to have the upper and lower control limit window equally. Especially in a truck where volumetric efficiency throught the rpm range takes quite a ride.

That was the recommendation form that site. The idea was to be sure that there was plenty7 of room to richen the mixture. But then again these are high rpm tuners and probably doesn't apply here.

Interesting to see what you come up with on the MAF and injectors. More interesting to see if it will work, OBDII trickery without the FTU software code can surprise you sometimes.

Not really a big mystery here either. People are bolting the TT Supra MAF to everything. It's a large bore high flow MAF calibrated to 440cc injectors. 4Runner tuners are installing them all the time and they plug right in. However they stop short of the complete install IMO and don't add the matched injectors. They start tweeking their piggybacks to compensate for the small injectors size, and they readily admit it .

I just bought a set of Supra 7mgte injectors last night. $53.00 shipped. Should fit and deliver the needed 440cc flow. And refurbed MAFs are available for $220.00. These will be the first thing I install to shake out the fuel system before the turbo, but not until spring I'm afraid.

This is a pretty good read of what I'm planning on doing, upgrade. It actually is on an older Supra with a Karmen style air meter but the principle is th esame. You might recognize some of the text

 

[Cattledog]

Why not install now?

If the MAF swap and larger injectors are to work they need to work correctly during non boosted conditions too. From my experiance the issue's with larger injectors surface during idle and low load with low rpm's. Typically the injectors just can't be DC short enough by the ECU, thats why I have seen issues beyond 15% increase of injector size. As a note, nobody was swaping/matching MAF units during my time with turbocharging mazdas. This should be interesting.

 

[landtank]

Why not install now?

I'm up to my ass in alligators. I need to get the truck up for it's winter job. Once spring is apon us the MAF and injectors will go in.

Design wise this should be a slam dunk. It took me about 10 hours of research to be able to verify that this combo will most likely function as factory installed equipment. Right now I'm as sure as one could get without actually doing the mod. The info was literally sprinkled over 100s of posts and across many forums. Bits and pieces everywhere.

As far as the idling issue, a 3.0L supra can idle with that setup so our 4.5L should. It's all in the calibration of the MAF to the injector size.

I've also talked with my machinist and he said it won't be a problem.

 

[landtank]

I just bought a used TT Supra MAF used for 300.00.

Cattledog, up for some LC playtime in a few weeks? If I had help I wouldn't might bolting the mod in.

 

[SUMOTOY]

Why not install now?

If the MAF swap and larger injectors are to work they need to work correctly during non boosted conditions too. From my experiance the issue's with larger injectors surface during idle and low load with low rpm's. Typically the injectors just can't be DC short enough by the ECU, thats why I have seen issues beyond 15% increase of injector size. As a note, nobody was swaping/matching MAF units during my time with turbocharging mazdas. This should be interesting.

CD: I absolutely agree. LT, the problem I see with this is a couple fold. First, I don't think you need larger injectors. Second, your readings under vacuum won't be as accurate, because a Supra 3liter motor will suck less air under vacuum than the 1FZFE. I really believe you are opening pandoras box here, regardless of the 4R success. The 4R have motor displacement within 15% of the Supra, hence I believe it will work.

To understand this better, looking at boost airflow vs voltage and vacuum airflow vs voltage on the MAF.

My spreadsheets indicate that the 29lb injector should work on a low boost truck. I believe the real issue for fuel from my measures is in the 0-3psi range. This appears to be something the 2035 can address with the gain function (read: onset of rising rate). I also agree with Christo, that the RR function doesn't need to really rise much at all given the injector size and boost level. It's quite possible that the gain needs to be tweeked, but the RR function will be very small. So small in fact, that smaller injectors with the 2035 might actually work better.

LT: Have you done the rpm data points for airflow and fuel demand based on just calculations? I'm very curious how you came to the conclusion that larger injectors are necessary. Possibly with a stock fuel pressure regulator? If so, doesn't that make them too big when *any* rising rate is applied?

Edit - More: I just added in the duty cycle correction to my LC spreadsheet, and at 8.5psi still come up with less than a 29lb injector at 5000rpm. Again, I'm still stumped as to why go bigger on the MAF and fuel injector. A AFM to MAF conversion might be a good idea... In this scenario, some recognition of load is needed coming onto the boost profile, but larger injectors than 29lb on a 8.5psi motor seems optimistic.

ST

 

[landtank]

I'm not looking for some Nobel prize winning math here, just some simple stuff.

Lets take the SC for instance

NA output is 212HP SC is rated 278, that's a 31% increase in HP

Now to get that you need 31% more air and 31% more fuel

stock injectors are 305cc X 31% = 400cc to stay within the original design DC.

However it has been postulated that the turbo that I'm looking at would produce 330 hp which would be a 55% increase requiring 472cc injectors

Since swapping out injectors requires a MAF swap as well I went with the closest combo which is a 440cc setup. Something down the middle of both.

I'm just repeating exactly what Toyota did when they Turbo'd the Supra.

I'm confident enough to spend the money and time to follow this through.

There seems to be two things here that are obvious:

1. you are hell bent on the fact that your way is the best way and are only looking for support of that opinion.

2. I'm obviously not swayed by your opinion.

The parts will go in as soon as I have time which would obviously need to include removal if I'm wrong. Until then I don't see any point of further argument on it's merit.

Basically, my minds made up.

 

[landtank]

Just got an email that they refunded my money as the part was pulled damaged. Back on the hunt I guess.

And yes, my mind is still made up.

 

[SUMOTOY]

I'm not looking for some Nobel prize winning math here, just some simple stuff.

Lets take the SC for instance

NA output is 212HP SC is rated 278, that's a 31% increase in HP

Now to get that you need 31% more air and 31% more fuel

LT:
Going back to RCeng site for injector sizing, if you plug in the .55 correction factor for boost, and put 330hp as your target, you get a 29lb injector at 43.5psi. You have that in your truck already!

I would be thinking and chasing along the lines of what that means the FTU assigns as the DC of a 29lb injector on a normally aspirated truck. My concern wouldn't be under boost operation, my concern of the bigger injector would be under vacuum operation. I suspect given the stock oversize injector and the fuel pump resistor, the fuel injector DC is already pretty low under vacuum.

ST