[Diy_efi] PC-laptop based efi

[Marcell Gal]

Hi,

> The difference is that the 332 has a
> header exposed on the side of the chassis for connecting to the drivers.
> With the MS or AVRMS, you have to hardwire in an expansion board.  The
first
> is by design, the latter is a hack.

Ah, I see. The difference is if the header pin connector is on the top of
the board (MSAVR),
or on the edge (efi332). I don't agree that this is the difference between
design and
hack, but you probably know better. The MSAVR (at least v2.2) was designed
to cover a respectable area of uC applications, with engine management also
in mind. The 2 driver FETs that are onboard seem to be a good compromise
to cover the needs of most efi-ers  * and the extension headers
(that can be used for HW-based serial communications or pure signals) leave
lotsof room for people who prefer using their favorite driver circuits.

> All of this doesn't take into account that the TPU controls the drivers on
> the 332, thus you have realtime microcode handling of signals.

the question is what timing precision you can achieve and what is necessary.
Even if you stuff 1000 times the precision and 1000 times processing
power, what good it is for, if you cannot take advantage of it to
 make a _measurably_ better engine operation ?

> Sequential is moot, the real issue is individual PWM current control for
more
> than 2 channels.  This is of more importance than sequence.

Yes, 7 hardware PWM channels (MSAVR) might not be enough for people,
who would want 8 LOW-Z injectors driven sequential (without current-limiting
resistors!).
It could be done from software, but IMHO only a small number of people
concerned.
They can buy Motronic, build the very nice (just laborsome) efi332 or design
their own.
I have yet to see that the overhead of building,
testing and tuning such a system worths the marginal efficiency gains of
seq-inj LOW-Z 8cyl, but maybe for 1000's quantities it does.
People who need this, speak up now. It's possible to put some gates on the
PCB
 to have larger number of HW pwm-channels if the need is there.

      Marcell

* yes, the MS used this appearingly good compromise first, except
the MS onboard flyback is not sufficient to support efficient PWM and fast
injector
 switchoff at the same time, which resulted in much heat and smoke and
  was corrected on MSAVR


_______________________________________________
Diy_efi mailing list
Diy_efi at diy-efi.org
http://www.diy-efi.org/mailman/listinfo/diy_efi