Has anyone found good justification for having different target lambda curve per gear? Or has anyone found that a good single lambda curve just needs to be made to be followed by each gear. I have tended to do the later but I have also tended to allow the actual verse target readings vary a bit per gear. In this case the lower gears tend to run a bit lean while higher gears run a bit richer. Which ever justification you think is better please try to provide some aspect of significance to your justification such as measured power or torque, drive-ability, better egts, etc, or that common fallback, just your opinion. Thanks.

 

it’s going to accelerate through the rpm range a lot faster in a lower gear is why.
.

 

It's really about thermal management. With modern ecus you can go closed loop based on EGT within lambda limits.

 

Depending what you're doing with the car, I wouldn't mind seeing a slightly richer lambda target at high RPM or high gear. Unless you own a trailer and a spare engine I think the main priority for tuning an engine (rotary or otherwise) is reliability over power.
If you haven't already seen this video from HP Academy, it's great.
Not mentioned in his video, I would add a fuel pressure sensor to any car with a standalone ECU. Personally I would get a fuel pressure sensor before an EGT sensor. And if you've got a lot of money in the engine and turbo, I would probably also add two more pre-turbo O2 sensors to watch fuel balance. I'm surprised that's not more common in the rotary community, a couple of extra widebands is cheap compared to removing and rebuilding an engine.

 

If you're going WOT in 4th gear continuously, you're going to be pushing the engine harder then banging through 1st or 2nd. The turbo(s) don't even fully spool at times in those lower gears.

 

I asked about this in the PFC FCTweak group and the only answer I got was that wideband sensors can't withstand the heat of pre-turbo location in a rotary... But I think they can live there long enough to get a well rounded tune and enough miles of driving, monitoring and adjusting if needed. Then, when all is said and done just remove them and leave the normal downstream one.

 

If you can get 95% the benefit from thermocouples that last years, why bother. I would argue thermal control on high load is the primary objective for 99% of users anyway, ok at idle and cruise you may get some marginal gain inside the noise that temperature doesn't tell, but if you are trying to decipher that you probably want to be programming your own primary secondary split, using the secondary runners throttles and combustion pressure logging anyway.

 

I'll say this. Even though most wideband systems sold by the aftermarket are using the same family of Bosch sensors (LSU 4.2 , LSU 4.9, or newer) some systems tend to burn out sensors more often than others (*cough*Innovate*cough*). I suspect it is due to the heater control. For what it's worth, I've used the Bosch commercial grade wideband systems (thousands of dollars, sold by their ETAS division to commerical applications only) and could withstand up to 950C (1750F) or more without burning anything out.