True Flex Fuel Tune

[Mikepol2]

OK so this thread has made me feel uneasy about running my Livernois flex tune. Am I understanding correctly that it is PROBABLY set up to be safe, and therefore not realize much gain? Because the car feels noticeably more responsive. Or, that being the case, am I playing with fire by running a tune that gives E gains but is probably so inconsistent that I'm in danger of engine damage?

Sponsored

 

[junits15]

OK so this thread has made me feel uneasy about running my Livernois flex tune. Am I understanding correctly that it is PROBABLY set up to be safe, and therefore not realize much gain? Because the car feels noticeably more responsive. Or, that being the case, am I playing with fire by running a tune that gives E gains but is probably so inconsistent that I'm in danger of engine damage?

Its safe, the offerings from major tuners you can get now are safe. They just tune around the downsides at the expense of power. That's why a lot of them offer two tunes and the dedicated e85 tune makes more power.

I only tune my car but I have never seen it off more than 5% after learning.
Long terms have turned on.

Yeah and that surprises me. I'm also self tuned, and I believe that yours is working, but mine absolutely isn't. I'm using the values from an F150 with some minor tweaks to extend the range. If I fill with ethanol or gas (whichever is the opposite of what's in the tank) and quickly get into a situation where the car is in its learning range a lot it will mature right as the trims start to creep up or down. Then it's locked at the wrong number with large absolute LTFTs until I KAM reset or refill. Not a big deal for fueling, but can be risky if learning from ethanol to gasoline, the timing can be too high for the entire tank of fuel. I saw this situation happen about 2 weeks ago while experimenting. I've seen it get stuck "actively learning" until I reflashed the tune. I've seen it learn down to 15%, and then after filling with premium e10, learn up to 30%.

I believe its working for you, but it absolutely does not work for me. I have to assume its my driving style, but that feels annoying because it should work for all driving styles.

The way I've seen it work consistently is if I buy fuel in the city, and have to creep through traffic before I hit the highway, then it purges the lines and is able to learn better. This finding supports your other post, a return style fuel system would be perfect for an inferred ethanol system.

Either way I don't like that it needs to learn, a sensor would be better for everyone and give me a lot more confidence. I don't like that I need to be careful until its learned correctly, a sensor would allow me to just fil and instantly mash the throttle with little to no risk.

 

[offroadkarter]

I would say to stay away from the proflex, because that thing is expensive and wont be as good as the inferred flex fuel tunes already available. Its more suited to people looking to run flex on a non-flex car for money savings.

That is the equivalent to just taking a gasoline tune and enabling flex fuel with no timing adders. All it does is increase injection based on alcohol percentage. The car can already do that fine, it doesn't fix the main issue which is reliable timing control.

If you put that on, tune it for e85, and then fill with gasoline, you will have uncontrolled knock and potential engine damage.

That same company has another product called FlexLink, no Ford support, but that is what we would need.
https://www.advancedfueldynamics.com/collections/flexlink-1

makes sense, ultimately I agree with you that the end game best would be to have the factory PCM understand a flex fuel sensor. I really like the concept of being able to swap between 93 and E85 just by switching handles at the pump, the idea of having to reflash the PCM everytime I want to swap between E85 or 93 scares me, only takes one wonky situation to brick the computer.

 

[K4fxd]

A sensor would be best but even if it learns 30% off it is still fueling like it is supposed to.

If doing a large change in fuel let it idle for a few minuets.

I think the MAF threshold is 2.4 Lbs a min for 10 seconds to trigger a learn. Just drive on and off the throttle for a few miles

 

[stannypack]

If I fill with ethanol or gas (whichever is the opposite of what's in the tank) and quickly get into a situation where the car is in its learning range a lot it will mature right as the trims start to creep up or down. Then it's locked at the wrong number with large absolute LTFTs until I KAM reset or refill. Not a big deal for fueling, but can be risky if learning from ethanol to gasoline, the timing can be too high for the entire tank of fuel. I saw this situation happen about 2 weeks ago while experimenting. I've seen it get stuck "actively learning" until I reflashed the tune. I've seen it learn down to 15%, and then after filling with premium e10, learn up to 30%.

So if the biggest issue is making sure it learns the e% incorrectly, how do I make sure it learns correctly? After a fill up let it idle, light revs, and babying it for the first 10 mins? How does it learn too quick like someone else previously said or gets "stuck" like you said.

Juggernaut is dialing in my 91 tune atm so there'll probably be a bunch more instructions when it's time to datalog the flex tune but just wondering.

 

[K4fxd]

So if the biggest issue is making sure it learns the e% incorrectly, how do I make sure it learns correctly?

By understanding what it takes to trigger a learn.

CHT must be above 160, MAF must be above 2.4 and below 8Lbs a Min. for 10 seconds time. Ask your tuner if they changed this.

So when doing a major change, like 93 to E85 or the reverse you want enough time for the furl rails to empty the 93 and fill with E85. During this time you don't want the MAF to be between 2.4 and 8 for 10 seconds.

My car steady driving about 45 keeps it above the threshold.

If just adding E to E or 93 to 93 I don't worry about it and it learns close enough. Even if you had a tank of 85% and filled with E53 and it learned fast it will still be under 10% on the fuel trims. 10% is safe.

 

[stang17]

If you have nguage or tuner that has KAM reset function, you can reset KAM after every fillup.

Innovate and Zeitronix make ethanol sensors/guage you can wire in and see the ethanol content in real-time. From there you can look at a/f ratio on dash or nguage, etc along with trims in real time.

 

[junits15]

So if the biggest issue is making sure it learns the e% incorrectly, how do I make sure it learns correctly? After a fill up let it idle, light revs, and babying it for the first 10 mins? How does it learn too quick like someone else previously said or gets "stuck" like you said.

Juggernaut is dialing in my 91 tune atm so there'll probably be a bunch more instructions when it's time to datalog the flex tune but just wondering.

The best way is to understand what triggers the learning, and have an idea of what a realistic ethanol concentration should be.

So if the tank is half full of gasoline and you put in 8 gallons of e66 you should see ~40% (approx.). That way if the thing matures at 20% you know its wrong. An even easier way is to watch LTFT's. They will be 0 when learning and some number when learned, if they're under 10% its close enough. If you see LTFT's above 10% at idle then its likely wrong. That would be with a warm engine, idling for a few minutes, LTFT's that do not eventually settle out.

Its not hard to do, and it works. Plenty of people set it and forget it, but I don't know about that given what I've seen personally. It's not impossible for it to be wrong. I just think that having to do this every time I fill sounds like a headache.

I think the difference between my experience and @K4fxd is that I'm kind of torture testing the system. I fill with ethanol, drive 30 minutes home normally and then check what it landed on, because that how I would like it to work in a perfect world. @K4fxd sounds like he's giving the system time to learn ethanol before he rips on it. In my experience too if you do that, it will be very accurate and repeatable.

 

[junits15]

So if the biggest issue is making sure it learns the e% incorrectly, how do I make sure it learns correctly? After a fill up let it idle, light revs, and babying it for the first 10 mins? How does it learn too quick like someone else previously said or gets "stuck" like you said.

Juggernaut is dialing in my 91 tune atm so there'll probably be a bunch more instructions when it's time to datalog the flex tune but just wondering.

Also if you're just running e85 all the time, you don't need to baby it, you can just let it go and spot check it every now and again. It will pretty much be always "close enough".

This is why I'm torturing the system, because my application would demand that I can rapidly handle large ethanol concentration swings accurately. There's only one e pump within an hour of me, and I can't always get to it. So my E content would be swinging around all the time.

 

[K4fxd]

@junits15

You could raise the time to learn to 20 or even 30 seconds. It's pretty hard to keep 2.4 Lbs second of MAF flow in normal street driving for 20 seconds.

I do think the F150's have additional tables for Flex fuel. I'm trying to learn the actual .bin files so I can see everything.

 

[stannypack]

Also if you're just running e85 all the time, you don't need to baby it, you can just let it go and spot check it every now and again. It will pretty much be always "close enough".

This is why I'm torturing the system, because my application would demand that I can rapidly handle large ethanol concentration swings accurately. There's only one e pump within an hour of me, and I can't always get to it. So my E content would be swinging around all the time.

I was originally going with flex but JPT recommended I get an E85R tune instead since the station near me tested at 80%. And I plan to run a tank of 93 every so often so I'd just flash the flex tune, or also if I get stranded I could just flash right then too.

And I just got my flex tune, time to get the tank to empty so I can fill up with E and datalog

 

[junits15]

@junits15

You could raise the time to learn to 20 or even 30 seconds. It's pretty hard to keep 2.4 Lbs second of MAF flow in normal street driving for 20 seconds.

I do think the F150's have additional tables for Flex fuel. I'm trying to learn the actual .bin files so I can see everything.

Do you find 20 works well? That would probably fix the issue with accuracy, but would take time.

I believes there's more to it than HPT gives us, I think there's likely a delay that we don't have access to. Ford has a patent for this, and in their patent they indicate that the system waits a set amount of time before it starts learning after a fill for the lines to purge.

However, I've seen our system basically immediately starts learning once you put in enough fuel. I'm now wondering if that register is set to 0 and needs to be increased.

If you ever find it it would be an amazing leap for inferred flex on this car. Would essentially fix the major issue with it which is that it learns on fuel stuck in the lines.

 

[K4fxd]

I'm now wondering if that register is set to 0 and needs to be increased.

I believe this also. If I do find it I will have HPT add it to the software.

I have mine set to 15 seconds. Only drawback is sometimes it won't learn for a few drive cycles. But I usually buy from the same station so fill to fill is pretty close in content. When I do a big swing I take care to make it learn correctly.

 

[WildHorse]

All aftermarket tuners use Inferred Flex Fuel. This is a reactionary strategy. I would trust it for any performance use. You absolutely need a ethanol sensor.

The o2 detects a leaner mixture from LAMBDA, relays that info to the PCM and added custom fuel tables adjust based on o2 voltage. Very inaccurate.

 

[K4fxd]

Very inaccurate.

It is accurate enough for Ford to warrant the engines using this logic.

Sponsored