GN injector offset

[Carl Ijames]

> > It has both injector offset ($34C1) and pump compensation ($34D3) vs. 
> > battery voltage tables.  The pump comp table is a multiplier but is all
> > ones.  The inj offset is in msec and gets added to the calculated ipw:
> > 
> > Volts   Offset, msec
> >  0.0     7.78, N=E*32.768 so 7.78=$FF
> >  1.6     7.78
> >  3.2     7.78
> >  4.8     7.78
> >  6.4     7.78
> >  8.0     3.3
> >  9.6     2.0
> > 11.2     1.5
> > 12.8     1.07
> > 14.4     0.82
> > 16.0     0.64
> > 17.6     0.46
> > 19.2     0.46
> > 20.8     0.46
> > 22.4     0.46
> > 24.0     0.46
> > 25.5     0.46
> 
> So, if I see see 2.0ms at idle then say about .945 (13.6 V) is to open
> and the other 1.055 is for the fuel, correct?

Umm, I don't know but I'll theorize some :-).  Qualitatively, I think that
it's safe to say that the injectors will snap open faster at higher
voltages, and with a saturated driver the higher the voltage the higher
the current and thus the bigger the magnetic field that has to collapse
before the injector closes so the slower it will close.  I have no
quantitative idea how much those two times vary, but my impression from
reading is that at nominal voltage the times are about the same.  To me,
that means that if the offset at 12.8 V is 1.07 msec and the total
injector pulse width is 2 msec then the injector will start flowing 1.07
msec after the beginning of the pulse, and continue flowing about 1.07
msec after the end of the pulse, for a total flow time of 2 msec.  At a
higher voltge, if the decrease in open time matches the increase in close
time, the net fuel delivered would not change.  Since the offsets fall
with increasing voltage I think that the open time must not be decreasing
as much as the close time is increasing, so for a 2 msec IPW more fuel
will flow (please verify my logic here :-)).  All of this assumes that the
intent is to normalize the amount of delivered fuel, not just to
compensate for the delay between the beginning of the injector pusle and
the start of fuel flow. 

-- 
Regards,
Carl Ijames     ijames at netaxs.com



----------------------------------------------------------------------------
To unsubscribe from gmecm, send "unsubscribe gmecm" (without the quotes)
in the body of a message (not the subject) to majordomo at lists.diy-efi.org