set the blm min too high or low
same question for max value

 

Depends upon the definition of too high and too low.

Setting the min and max BLM to 128 locks the BLM cells to 128. Setting the min and max to 120 will lock the BLM cells to 120 which will affect the fueling (remove fuel).

Setting the min to 0 and the max to 255 will allow a full range of fueling correction, maybe too full of a range.

Setting the min greater then 128 and the max less then 128 will cause the ECM to become confused and melt down like a Darlek with a coat over it.

RBob.

 

Is that a good thing to set BLM max and min to 128?

 

so if you set the min to 0 and the max to 255 what type of problems can come up

 

i think some guys lock the blm and just tune off the int. i never understood why and i think there is a conflict of opinion on this. maybe some one can expand. i tune off blm currently and dont look at int. maybe i should, dont know. i never changed the range. what ever was there day one in bin is still there as fror a min/max blm/int. i also amusing narrow average. i was recently told to use all wide, narrow, 10 samples and compare. i try to keep it simple till knowledge grows.

 

If the fuel calibration is correct, and everything is working as it should, there'd be no problem.

But, imagine if you suddenly developed a header leak, or perhaps had a sensor go bad...now the ECM, in trying to compensate, is free to REALLY mess up your fueling, based on the bad data. With the stock BLM limits (e.g. 108-160) the car would probably at least still run, albeit somewhat poorly.

 

There's more than one way to skin a cat. I wouldn't really call it a conflict; just different techniques for doing things.

Once you have the fuel tables close, if you're trying to REALLY nail things down, locking the BLM can be useful, in that there's only one fuel correction going on, rather than two.

For example, you might have BLM of 124 (removing fuel) but the INT is 132 (adding fuel). So what's the net correction you need to make? Taking away or the other (by locking it at 128) just simplifies things. And INT updates quicker than BLM.

Might give it a try sometime, and just see if it works for you. Not saying that you HAVE to do things that way, though.

(Lately I've been tuning entirely in Open Loop, using a WB, so I don't use either BLM or INT. As Grumpy might say, not to be attempted without adult supervision.)

 

i remember some heated discussions a while back. in fact on another forum it was said running open loop idle to manage a large cam is not ever wise but that i disagree. thank you for the explanation. i was not aware that the int could be doing just the opposite of current blm settings? i guess that would be possible if air temps dropped a lot quickly. my blms are somewhat stable but occasionally one gets way out of wack. today one popped up blm 153 at 50 map/1200 rpm. narrow average. i know from prevoius dtalogs it was never that lean. if i disable blm and run of int would not i see more wiild fluctuaions as is not the int more volatile so to speak?

 

One thing that can potentially happen is an oscillation about a boundary if the A/F differs by a large amount (i.e. bad calibration). I can't remember, though, if GM's Integrator resets to 128 when entering a different cell or not. Does it? Someone please remind me.

 

Some do, some don't, some have an option bit to enable/disable that functionality.

RBob.

 

So, if the Int resets when entering a new block and, for example say the system is just rich, and INT goes down, and then that leaning out reduces the power and increases the MAP, and that crosses a boundary and resets, and then the richness comes back and brings back the power and lowers the MAP and then the INT goes down, and leans it out a little and the the MAP goes up, and it crosses a boundary, etc.. This is mostly a problem at idle if the system is allowed to switch cells while in idle (throttle closed, car not moving). We designed our system to have learning and found it to be quite difficult to avoid this without forcing the idle block larger, and adding hysteresis to cells, and increasing the delay between block learn updates.
I'm not trying to discourage you (anyone) from messing with those numbers, just discussing what could *possibly* happen, if conditions were met. Actually I'd like to ENCOURAGE you to try different things and find out what they do, and then report your findings here!
Good luck

 

I've been in a cave.

Whats a "Darlek"?

Have I got one in my ECM?

 

From the British SiFi series Dr Who. Had a habit of getting a tad crazy and blowing up whenever confused. Of course I spelled it incorrectly, drop the 'r' out: Dalek.

RBob.