Hey guys, I'm looking for some dyno charts showing the air/fuel ratios before and after various modifications like intakes, exhausts, etc. The few R56 charts I've seen so far look very strange, and I'm wondering if there is a correlation between wacky air/fuel curves and bolt ons.

Wacky air/fuel curves could be lead to inconsistent or less than optimal engine performance so its worth looking in to!

Thanks,

T

 

Don't tell me you guys bolt on air intake and exhaust mods without checking air/fuel ratios? You're scaring me.

 

The R56 is using a Bosch LSU semi-wideband O2 sensor. If you know how to convert milliamps to AFR then you're already a step up. Nevertheless, the datalogging I've done shows the same strange sawtoothing pattern whether the engine is stock or has some bolt-ons [ECU mods excluded].

 

if it doesn't blow it's within spec, correct?

 

The R56 MCS runs in closed loop mode all the time due to its wide-band O2 sensor. The only way to change the AFR is to either use a piggyback or flash the ECU with new AFR targets.

 

that would be sweet as user tunable... like a cobb piece.

 

But does it use the wideband or maf for fueling? It just seems strange BMW would program saw-blade style air fuel ratios...

Converting Ma to air/fuel is pretty easy, 0ma = 14.7, -.6ma is around 12:1.

 

 

PS. Are there any posted air/fuel ratios with the currently available Unichip piggyback? Still the saw-blade?

 

here was one of mine, not my best run but you get the idea.
Michael_Gary.pdf

 

That shows that the richest was a 13.0 and under the extreme loads of the dyno it was a lean as 15.2! For the compression and boost, I'd be scared to see it hit 15 during wide open throttle. The '03 Cobras with 8psi and 8.5:1 or so compression were running an overly rich 10.5 to 11 AFR. That's pretty extreme and they are nowhere near as efficient as the MCS, but I'd press more towards 13-13.5 AFR with reasonable timing. It looks like you'd be asking for a piston to melt or detonation. The knock sensors should help, but I wouldn't risk my engine on it. The Chevys I worked on with the 8100 vortec would melt the knock sensor harnesses off the engine. That's a different engine, but look at how hot that turbo gets especially since you're pushing it so hard. You could easily melt down something and the lean condition will only make it hotter.

I'm not trying to bad mouth your setup or the tuning, but I wouldn't feel safe with that. I'm not the Gods gift to automobiles, but I am ASE certified and I've seen more than most with many more years of wrench turning under their belt. Good luck with your car.

 

But I don't see the air/fuel curve?

 

the thick black line on the graph?

 

Two words: Direct Injection. The whole game has changed; put away your old thoughts of needing to run painfully rich to cool the cylinders and pistons. No offense however your ASE certification will not help much in terms of tuning in this new realm of GDI setups. Also, most likely that AFR plot was done with a tailpipe sniffer, which means the exhaust is filtered through TWO catalytic converters, corrupting the true AFR dramatically.

 

Direct injection or not, if you lean it out that much you run the risk of melting a piston or blowing holes in it. Direct injection isn't a miracle drug, it's just more efficient.

 

You can bad mouth ASE certs all you want. I won't get into arguments on who has the most training or who has more experience. I hung up my wrenches as a profession, but I still keep current. Now I'm a helicopter pilot and our engines suffer from the same limitations. If the combustion section gets too hot, things start melting. You can't push 20psi of boost in a 10.5:1 compression engine with a lean mixture without creating excessive heat. Lean is lean regardless of how the fuel gets there. 15:1 AFR is lean under load for a carb car, TBI, multiport, or direct injection.

 

My issue isn't with the air/fuel ratios, depending on the boost level 14:1 can be OK with direct injection. I'm wondering about the saw-blade graphs. Attached is a stock 335i (twin turbo, direct injection) dyno with air/fuel taken at the tailpipe through 4 cats (2 on each bank). No number of cats is going to make the air/fuel go way rich / way lean / way rich / way lean. They would effect the curve consistently.

 

--->58driver: I'm sorry I didn't mean to bad mouth ASE in particular. You must keep in mind the O2 was sniffed through TWO cats. In that case, using rule-of-thumb, his AFR's are wandering from 12:1 to 14:1, which is about right for peak torque of 13.0:1. Yes I'd like to see a tighter control around peak-torque, but the stock ECU programming doesn't allow for this.

You're correct about "lean is lean", however the level of danger induced from being lean is much more severe in a carb vs. PFI vs. GDI. GDI setups are far superior in terms of chemical quench and allow the engine to run closer to peak-torque AFR all the time versus a sloppier and hotter fuel delivery method. Like I said, the game has changed, and it's ok to run high boost with high compression with [relatively] low octane with textbook AFR's [i.e. "peak torque" and get away with it on a GDI setup.


There's no graph attached. I'm not saying the cats will create the sawtooth, however it's common knowledge they skew the tailpipe sniffer results, like you said. The sawtooth pattern is a function of the R56 ECU's programming; unless you can hack the ECU and reflash it new AFR targets that's not going to change.

Peace,
Ryan

 

Attachment didn't seem to work, so here is a link to a dyno from our website. Same scenario though.

I don't believe the saw-tooth is a function of the factory ECU targets... But maybe if more people can come up with air/fuel dyno charts we can come to that conclusion?

 

Can't be right, it says 6:1, and 3:1 at the high end?

 

Come on Terry, don't let me lose faith in you! READ the graph; the AFR don't have their own scale printed, but the min/max/avg are listed numerically in the lower summary. From there you can extrapolate the scale.

Look at that right-hand scaling again; could an AFR possibly have a negative number? LOL Who knows what that scale is, but it's clearly not power, torque, or AFR.


I'm sorry you don't want to believe the factory AFR targets, but datalogging doesn't lie [unless you're using a piggyback]. Sawtooth is the factory pattern, no matter how stupid or nonsensical it may be.

 

My response is to "MotorMouth" who says the AFR curve is the black line.

RE: the sawtooth, are there any other stock dyno charts you can cite to establish this as "the" factory curve? Not that I don't want to believe you, but it seems pretty fantastic.

 

I thought I had some in my galleries online, but it turns out I don't. Look up ficcion, he posted a stock R56 dyno that had AFR on there. Same sawtooth. If I were in your shoes I wouldn't believe me either; I'd get a stock R56 and datalog it myself!

 

Still En Route. :(

 

Detonation is much less likely with the GDI and it's still really sinking in just how good the system is. I haven't played with a ton of cat vs. non-cat AFR's but from what little I've seen they were never that lean in either case. As I mentioned before, the Cobras (with cats) show around 11:1 stock. That's why a chip made huge gains on those cars. My carbed Mustang without cats had a rock solid 13.5 or so. It's funny that my BG Speed Demon's AFR looked like it was drawn with a straight edge compared to the Mini's roller coaster. That kind of disturbs me.