ianferrell

Just thought someone else might find this as exciting as I do...

Ken says he's going to add it for the next alpha:

http://msextra.com/forums/viewtopic....78a669#p305155

y8s

iTrader: (8)

I sweet-talked him. Sort of.

The reason it came up was because his new idle duty target table takes into account MAT (or CLT) but not load. I tuned it for no accessories turned on and it worked great. But when I turned on the headlights and hvac blower fan, the load went up and the target duty table was too low.

He said to try the voltage compensation table but the voltage remains fairly constant due to alternator control... i.e. you exchange dips in voltage when accessories are on for load on the motor when the alternator is working harder.

So he said the only way to accommodate that was to feed-forward the alternator load requirements and the best way to do that was to know what the alternator load was going to be.

At first he said it would only be useful for me (and I guess other NB miatas) but he said later that some Neons use it also?

Anyway, it's pretty cool!

Ben

iTrader: (33)

Neon and Miata are 2 commonly megasquirted vehicles that use ECU controlled vreg, but those are not the only applications out there with external, ECU controlled alternators. As we expand the capabilities, it will attract new users with varied applications. Eventually, other applications will become commonplace as well.

Feed-forward thanks to Ken for adding this feature. It will benefit current users and attract new users to the brand as well.

JasonC SBB

A feedforward from system voltage is like a 'D' term on alternator load.
When a sudden electrical load is added the voltage will dip momentarily and the alternator applies a step-load on the engine. The voltage dip thus looks like the derivative of alternator load. So it will help prevent stalling, but the steady-state duty needs to now be larger. In the absence of knowing how much the alternator is putting out (and thus how much load it presents on the engine) and a feedforward table from it, the 'I' term in the main idle feedback loop should take care of the increased steady state idle% needed.