True Flex Fuel Tune

[K4fxd]

@Rolls

None of the flex spark modifiers are defined in the mustang tune like they are in the F150. IE 79239, 79263, and several others. This is in a 2017 tune.

Are these not defined by your software or are they missing from the strategy's

Thanks.

Sponsored

 

[mejohn50]

This is where using a flex sensor is going to be 100x better than the inferred ethanol. Run any mix your want. We can add in a purge delay as well so it lags the signal by the fuel consumption just like the oem system also.

A purge delay would be great. We’d need some ability to configure it though. Going from ethanol to pump gas is a potentially dangerous switch. Even with a sensor there’s a lag from when the fuel hits the sensor to when it purges all the way to the DI pump (on gen 3). I’ve considered putting the sensor between the fuel rail and DI pump to eliminate as much of a lag as possible, but a purge delay would be a great thing to have.

 

[WildHorse]

Going from ethanol to pump gas is a potentially dangerous switch.

100 fold when it's inferred.

 

[K4fxd]

Going from ethanol to pump gas is a potentially dangerous switch.

Mike has come up with a way to make it much safer. I'm trying to get the Flex spark modifiers added to the gen 2's.

Fingers crossed.

 

[engineermike]

100 fold when it's inferred.

The ford logic isn’t that bad once you realize how it works and log a few transitions.

 

[Rolls]

A purge delay would be great. We’d need some ability to configure it though. Going from ethanol to pump gas is a potentially dangerous switch. Even with a sensor there’s a lag from when the fuel hits the sensor to when it purges all the way to the DI pump (on gen 3). I’ve considered putting the sensor between the fuel rail and DI pump to eliminate as much of a lag as possible, but a purge delay would be a great thing to have.

Right now we have a simple filter constant however lagging the system based on fuel consumption using the built on scalars I think will be the best option.

Once we have a few cars testing the logic out we will continue to take feedback and improve it.

Something as simple as adjusting the filter constant based on fuel consumption since a change > x % and a different filter constant in the e85 - > 93 vs 93 - > e85 direction may be suitable.

We could also couple this with a timer for lagging the spark blending by a simple time value when a greater than y % ethanol change is seen.

The goal is you will be able to take a 1000hp boosted car with e85, fill it with 93, floor it out the gas station with no knock and acceptable trims. We can make this a possibility without a doubt.

All easily implemented in our custom firmware once we get some feedback on how to performs.

 

[Rolls]

@Rolls

None of the flex spark modifiers are defined in the mustang tune like they are in the F150. IE 79239, 79263, and several others. This is in a 2017 tune.

Are these not defined by your software or are they missing from the strategy's

Thanks.

Can you post on our forum with the f150 strategy/osid, then a 15-17 cal where they are missing. Most likely they changed the resolution and hence the IDs in the later models.

We should be able to locate them for you.

 

[K4fxd]

I started a support ticket.

Thanks

 

[shogun32]

The goal is you will be able to take a 1000hp boosted car with e85, fill it with 93, floor it out the gas station with no knock and acceptable trims

Except who drives like that. And are we getting close to perfect being the enemy of good?

Because once you achieve that capability and advertise it, some dumb ass is gonna do it. And if they blow the motor they'll come looking for you. If instead the directive is drive like grand ma for 10 minutes, but the car is ready in 3 minutes, nobody does anything stupid and you avoid an encounter

 

[K4fxd]

the tables in pcmtec are auf73806 and 73838. It would be cool if you can add that logic using customos.

@Rolls

Is this possible? The above tables are from a gen 3 and I would like to have them in the gen 2 strategy's

 

[K4fxd]

Except who drives like that. And are we getting close to perfect being the enemy of good?

Actually with the Ford system you can go WOT right after filling up. I don't reccomend it but the ECU operates very fast. Also right after a fill up you are still on the fuel previously in the tank. It takes a few miles of driving to purge the old fuel from the lines.

 

[junits15]

Except who drives like that. And are we getting close to perfect being the enemy of good?

Because once you achieve that capability and advertise it, some dumb ass is gonna do it. And if they blow the motor they'll come looking for you. If instead the directive is drive like grand ma for 10 minutes, but the car is ready in 3 minutes, nobody does anything stupid and you avoid an encounter

You need that level of capability to ensure safety in all scenarios.

It’s a corner case, sure most people don’t drive like that, but the current system has a hole in its capabilities. Mike is working to improve the inferred system, but the current established flex method widely in use has inherent risk right after the fuel is switched. If you don’t have a PC to monitor inferred ethanol content you’re at risk, that’s too much work.
The goal of the work done here either by PCMtec to sell a sensor based system or by mike and k4fxd to improve the inferred system is to eliminate this risk. The system should be safe all the time, willingly leaving a 5-10 minute time period in where the system isn’t safe is bad engineering.

Tldr: most people don’t drive like that but if someone does it’s dangerous, so we want to eliminate the danger.

 

[engineermike]

If you don’t have a PC to monitor inferred ethanol content you’re at risk, that’s too much risk.

You can actually use the stock AFR gauge to do that. It only shows the learned stoichiometry modified by either DFCO or PE. At steady state or normal acceleration, 9.8=85% learned and 14.1=10% learned, and it’s real time.

 

[junits15]

You can actually use the stock AFR gauge to do that. It only shows the learned stoichiometry modified by either DFCO or PE. At steady state or normal acceleration, 9.8=85% learned and 14.1=10% learned, and it’s real time.

True, I forgot about that, that's a perfect method.

 

[engineermike]

Re: ford’s f-150 strategy, the way ford adds timing is interesting. They mainly just add it at low load and speed. Once you get to high load and high rpm, they add nothing to the borderline tables even at 85% ethanol. My guess is at high speed and load they figure all the knock advance is in by then and you’re at mbt anyway.

Sponsored