Hi.

I'm reading confusing things and need some guidance.

This is Bosch ME9 related on a BMW k1300s motorbike. The system is alpha-n with closed loop (under low throttle conditions) from a narrowband o2 sensor. No MAP or MAF.

It is generally said that these, and most other bikes now, are horrible to drive at low throttle or cruise, and it is because they are programmed to run very lean in these conditions.

I see some people showing diagrams and talking about a "requested/target AFR" table (on s1000rr which has the same BMSKP ECU). But I also see somebody else saying this is impossible, because as we know, the narrow-band o2 sensor can only say "that's 14:7".

If the supposed "target afr" table says "15:1" or lambda 1.02, well how does it know? If the answer is that "it doesn't know, but it has the engine parameters modeled and can make a best guess", then this must mean the o2 sensor & closed loop feedback can only be utilised/sampled when the bike is not doing its normal thing - i.e. the engine must have to periodically run through all of its "desired afr" values until it reads back an actual afr of 14.7:1, at which point it would know the offset/correction factor. but to do this, it would have to hit 14.7:1 to just do the check/test/feedback bit, and if supposedly the engine is mapped to run at ~15:1 then.. well it can't be doing both can it.

So what am I missing? Is this concept of "desired afr" tables purely a fantasy and a misunderstanding by people who are altering other tables?

Someone help me out ;-) It seems to be an accepted fact that vehicles, and particularly bikes like this, run lean, at least at cruise. If the o2 sensor cannot give feedback at this point, then how is that even achieved?

Am I right to be looking for a 'target afr' table. Is it true that I cannot just increase a fuel injection quantity table, because the fuel trims will compensate for the added fuel?

if anyone is interested, some .bins from various bikes are here: http://www.internetsomething.com/bmskp/bins
The one in question, mainly, is the K1300S one. That's my bike.

I have a Zeitronix ZT2 kit here, with the black box logger, and the LCD screen, but I haven't fitted it yet.

I'm waiting for a spare bmw o2 plug to come & a big resistor so that I can do a plug-n-play install, i.e. put it in place of the narrowband, but be able to swap back easily.
I'm not sure there's enough space to have both fitted, and I have no tig gas, nor am I confident that I'd actually do a good job of welding the bung in
I'm also waiting for an Arduino compatible OBD thingy so that I can log from the OBD port, which will save me doing much hacking of the bike's wiring, and also allow me to log CL/OL status.

At this stage I'm just trying to get my head around the theory and possibilities, and identify the purpose of the maps that I have found in the .bin

You would want to have both installed, not swap back and forth between the two. You want the ecu to be under normal operating conditions, while you monitor things. Not sure if this is what you meant.

That is exactly what happens.  It's how it's done on pretty much every system.  It is only open loop on idle and cruise in most cases.  The Audi S4 ME7 that people on here tune works in exactly the same way

Rick

It will almost certainly target lambda 1 when cruising. This historically is considered lean, especially for an "camy" engine with lots of overlap and poor low speed combustion.

Rick