I've heard it mentioned on here and elsewhere that it's a good starting point to tune the blm's to 128. I've heard of some more "advanced" tuners aiming more towards 124-120 blm. Is this just so the ecm tends to default to a more rich condition than lean?

I am just curious why this would benifit anything, as I get closer to my optimal tune. It was my understanding that anything between the BLM limits will be corrected to stoich anyways, and will not be reflected in the actual AFR.

Thanks guys

 

Well, if one part of a block has to learn differently than another, that can cause problems. Also, block learn is applied on a slower loop, which can cause throttle response problems if, for examply you idle at 108, and at low RPM high MAP or LV8 it needs 128+, the slow update can cause a lean pop (maybe not a problem on all ECUs, but for some it is). Also, most stock calibrations only allow about 10% Block Learn difference from 128 to be stored on powerdown, which can leave you rich or lean on startup. Also, it takes quite a bit of time in each block to learn it fully, and not all blocks are hit during normal driving, thus when going to new blocks that aren't learned - performance or economy can suffer. Also, during PE, most ecu's don't apply BLFs of less than 128 for warranty sake. Also, when the battery is disconnected, the learn values are lost. That's all I can think of for now, but yeah, in general, I seek 120-126 and call it good.

 

thanks for the reply

Does that also explain why some people set BLM's to 128 and let the integrator do all the work? The faster response time?

Or what about setting a very narrow range for the BLM, and allowing the integrator to do the work? Like, 120-140 or so? So that under normal driving conditions the ECM still has the memory to fall back on?

And then what about taking that one step further and extending the integrator to around 90-180?

Have you ever given these a try? What kind of results would that get?