Ignition Problem

[garfield at pilgrimhouse.com]

On Fri, 05 Jun 1998 12:57:00 +1200, Simon Quested
<questeds at whio.lincoln.ac.nz> wrote:

>So I connected works fluke scopemeter up to the -ve of the coil and 
>this is what I saw.....
>http://www.lincoln.ac.nz/ccb/techs/simon/netpics/coilpuls.jpg

[posted and emailed]

That looks distinctly like the primary voltage waveform of a
dwell-limiting style module, driving the primary. You can see the low
side driver pulling the -Coil (that's I presume your -ve signal)
terminal down to near gnd, then you can see the not-so-tight current
limiter start to reign in the coil current (voltage starts creeping up
toward +12), and then that second jump is when it goes into a harder
current limit mode (I presume once the coil is saturated); a little more
time and the coil fires. As lumpy as the current limiting looks, it
still looks like the low-side driver and timing signal to it is working
right, and the current limit derived primary waveform seems pretty
uniform, so that looks OK.

BUT, on all of the "shorties" there, you can actually see that something
is pulling the coil primary back down toward gnd, and in two of the
cases, actually does make it pulling the primary all the way back to
gnd, with even a little undershoot below gnd. This looks like avalance
breakdown. This is gonna put the brakes on the field collapse, and
probably severely limit the HV in the secondary. You should be able to
also trace the secondary waveform with a couple winds of wire around the
plug wires, and when one of these shorties get's really arrested all the
way back to ground, you should notice a very crummy HV spike, I would
think.

If I had to GUESS, I'd say something is breaking down intermittently as
the primary voltage fly's back as the field collapses, either the
module's output driver transistor, OR the protection diode, OR, since
this signal (don't ask me WHY, it should be well clipped I'd think
before going anywhere else) goes to your EFI stuff, the receiver on that
end should ALSO be suspect as breaking down and avalanching toward gnd.

>Is there any way that I can check the ignition module?

Say more about the construction of this gizmo. The "transistorized IGN",
does it indeed have a separate module? Seems your diagnostic difficulty
is you can't remove the EFI attachment to this -ve signal or whatever it
is, without killing the engine, so it's sorta like the problem of
multiple drivers on a bus; who's clobbering the bus? Only, in this case
it's multiple loads, one of which may be breaking down and glitching the
coil primary back to gnd. I s'pose you could try buffering the signal
feed to the EFI unit, to exhonerate it. If it's got a proper GM module
inside the IGN box, something musta really clobbered it, cuz the GM
modules all have a hefty transorb attached to their output darlingtons.

The other thing I would ask is, has this setup EVER idled OK? If not,
I'd kinda tend to suspect the breakdown occuring elsewhere than the IGN
drivers, cuz they shouldn't be weak right outta the box. If that EFI
unit wasn't meant to see the peaks on a primary IGN coil, which can be
pretty good sized (is that an actual 10v/div I'm seeing, then the
protection diode when it's getting to work is at around 60V), but it
looks like alot of the time something is breaking down at around 30-40V,
then it could have been breaking down intermittently from the beginning,
and gradually getting worse as it heads toward toast-heaven.

>My efi unit gets the injector fire signal from the -ve of the coil and 
>I'm wondering if the efi unit is causing the problem

It's a possibility, I would think. But ya know how this armchair
diagnostics can go; you probably will find it's something TOTALLY
different and unrelated. Sure looks like breakdown to me, tho.

Just me dos centavos, amigo.

Gar