On another forum, one of our esteemed colleagues who use to program for AM said:
"Why do you ask the question about fuel cut-off?? Yes, in gears 2-6 when the throttle is closed, the engine goes into deceleration fuel cut-off, this saves fuel and gives a sports car feel deceleration, unlike some cars which keep fuelling on so that the car 'sails on'. Different approaches; one set of values for sports car, one set for luxury cars. However, in first gear and below about 3000 rpm, the ECU goes from fuel-off to fuel-on. There are driveability issues if the fuel was not reinstated at this speed."
I'm not seeing this in my car and not sure if my mods have prevented an entry condition for DFCO. If I drive my car at steady state above 3000RPM and then close the throttle, my car immediately goes rich for 5-10sec. My wideband O2 records A/F in the 10 to 11 range. I have attached a data log of my narrow band primary O2 sensors and at throttle off you see the sensors going max rich for many seconds. If I have a Deceleration fuel cut off (DFCO) I should go lean! I know some Fords program a second or two of delay but not 10 seconds? Could someone with a stock car and OBD2 data logger try this test and see if it matches mine or not?
TEST: drive above 3000RPM in 2-6 gear at steady throttle then lift your foot and log O2 sensor 1 voltage vs time.

Thanks

 

I think that explains what I felt was bad fuel burping its way through the system on the track. It was a momentary total lack of power as if I had lifted off a lot. It has happened a couple of times.

 

In order to get truly accurate data you need a wideband 02 sensor to get a much more thorough picture. For the same reason, using an AFR gauge on a narrowband O2 sensor is pointless because the gauge will just keep full rich to full lean constantly back & forth erratically.

 

I have a wideband sensor and it shows me going way rich 10:1 or 11:1 for 10 seconds on overrun. I've asked for a datalog from a stock car and since stock only runs narrowband O2 sensors thats the best I can hope for from the community. If, however, someone has a wide band O2 sensor installed, I'd love to see a graph of that on overrun/coast above 3k RPM.
I'd be more than happy to post mine to compare.

 

I know this is super old and most people on this forum probably don't care, but I thought I would note that my car definitely does not DFCO most of the time, as the instant mpg gauge reads a pretty low number after I've lifted off the throttle unless it's 6th gear. This is also why the revs take forever to drop when you rev the engine in neutral or are upshifting. I'm guessing it's because the minimum vehicle speed for DFCO is set to something pretty high like 50mph, as my injectors are back on before 1500rpm in 6th. The shudder when they turn back on is very subtle, barely noticeable.

If you want to get better "city" mpg on the OEM ECU tune, you have to go to neutral when braking or coasting, or you'll be burning quite a lot of fuel on decel as if the car is carbureted. I was kind of wondering why I burned through so much fuel going downhill...now I know.

 

Here are some of the scalars that control DFSO. There are a few subroutines that control it right out of Ford EEC6 programming. If you understand how to modify Ford DFSO in the EEC4/EEC6 then AM is essentially the same. AM is setup to do it only in 6th gear decelerating to generate the misfire profile correction.

 

assembler? ...ouch!

 

@rmrmd1956 This is excellent, thank you so much!

So if I understand this correctly...

Max NOV DFSO should be increased to allow DFSO in lower gears, since N Over V is overall gear ratio * something. Doing some quick arithmetic with the gear ratios on our car, something is between 69.76 and 84.17 since 6th gear is below that threshold and 5th is above. If I increase it to 1038, it'll allow DFCO in all gears, and setting it to something between 650 and 860 will prevent DFCO in 1st only.

The "Load" parameters are referring to air flow/rpm or MAP, to activate DFCO only when specific airflow is low to reduce change in torque. 0xE1C60 and 0xE1C64 are maybe some kind of requested torque bounds for the same thing.

Min_RPM_DFSO (I think this is called DFSRPM by default) is kind of confusing to me, you clearly don't need to be at 7000rpm to get DFSO. I'm looking at a CDAN2 (EEC5) strategy doc and it says this param is used for DFSO in-gear, where rpm - "desired rpm" needs to be greater than this number to enter DFCO, which doesn't seem right. What would make more sense is 7000rpm to be the maximum and 1150rpm to be the minimum rpm for DFCO...is this right?

DSTM5 could be reduced to make it feel more responsive to lifting off the throttle.

There's should be a SHFHYS scalar param somewhere to accompany SHFRPM, seems like this is only used during "shifts" = clutch pressed or neutral. I noticed in neutral, the revs drop below 600rpm before the idle settles, I think decreasing this param would help make it recover to normal idle a little faster, or alternatively, increasing SHFRPM a little.