On Sat, 30 Apr 2016 16:40:11 -0700, wasaol wrote:

On Saturday, April 30, 2016 at 7:35:03 PM UTC+1, Johnny B Good wrote:

====snip====


The battery powered Kettering ignition system has to be designed to
provide enough spark energy at the minimum engine rpm limit and

In my original 'vanity/proof reading' scan, I *had* used the correct
"maximum" word in the above statement. However, even I managed to get
sufficiently confused as to change it to the word "minimum". Apologies
everyone (I do see other posters' justification for criticising my often
overly long (and flowery) postings. :-(

consequently is over-specified for startup and tick-over where, despite
the CB points spark quenching capacitor, most of the surplus to
requirements energy lands up counter-productively eroding the CB points
in rather short order necessitating frequent filing and burnishing of
the contacts themselves, along with a gap adjustment to recalibrate the
ignition timing[1].


====snip====


Thanks for the informative reply. You gave some useful info that hadn't
occurred to me concerning cylinder pressure affecting things. As a
fellow ex-tinkerer and owner of several British twins (1953 BSA Golden
Flash, highly customised, plus a couple of Royal Enfield twins and one
350cc Triumph twin) I rebuilt engines on three occasions, so I expected
that a simple little petrol mower would be a piece of cake to
understand,
but I am baffled by both the way the carb works and also the ignition
system. I can see that the main flywheel-like thing has a smooth shiny
surface that passes within 1mm of something covered in plastic that is
about 2" diameter and an inch thiock from which the HT lead goes to the
spark plug. That must be the coil, is it? I presume that there is no
contact-breaker. And this would also mean there is no condenser, yes?

I suspect this must a modern hybrid version of the magneto system where
the functions of a self contained magneto were split into a seperate
generator (permanent magnet rotor keyed directly onto the engine
crankshaft) feeding ac current to what would otherwise be a conventional
ignition coil and CB arrangement normally powered from a battery. The
keying of the rotor of the PM alternator guaranteeing that the current
would always be interrupted at the peak of the generator's waveform.

The generator would also be used to provide electrical power to
headlights and instrumentation (speedo and main beam warning) lamps,
usually from seperate generator coil pairs (it was common practice to
split the generator output into two sources (using either 50:50 or 34:66
ratios) and combine them according to load demands using auxiliary switch
contacts on the side/head light selector switch as a crude charging
control in a battery set up or else to avoid burning out sidlelights when
the extra power wasn't required in a batteryless system (typical of
mopeds and some models of motorbikes, usually models designed for
scrambling events or bush trail riding where the battery would be an
unnecessary luxury and a liability).

In this case, the flywheel generator is most likely powering a
contactless "transistor assisted" ignition module circuit and coil hidden
under that plastic cover. The ignition timing will most likely be
triggered by a magnet and pickup coil sensor (avoids mechanical wear and
tear, hence the 'benefit' of the additional 'electronics' which might
literally be emulating the original magneto idea, replacing the CB points
with a high voltage switching transistor interrupting the appropriate
polarity peak current of the ac output taken directly from the generator
winding or, a little more sophisticated, triggering a capacitor into
discharging some 400v into the low voltage winding of the ignition coil
to generate the necessary 30 odd KV spark pulse almost regardless of any
leakage (damp or fouled plug insulation).

In this case (Capacitor Discharge Ignition) the ignition coil is only
being used as a step up transformer, not as in the original case, for its
inductive properties to generate the necessary 400v pulse *and* its step
up transformer action which relied on there not being any leakage in the
HT side to compromise its ability to rapidly turn off the 3 or 4 amps of
current flowing in the inductance of the coil's primary winding fast
enough to produce the required 400v pulse of back EMF.

Whilst a CDI has a bit more complexity than a naive "Transistor
Assisted" ignition circuit, it does offer the benefit of less erosion on
the spark plug points, much greater immunity to HT leakage current issues
and a more powerful spark to allow improved fuel economy and engine
performance. If you're going to start adding additional electronic parts
(points of failure) to the basic ignition circuit, you may as well "Be
hung for a sheep as for a lamb". :-)

I'd guess what is actually hidden away under that plastic cover rather
depends on the vintage of that B&S engine. If it's using 21st century
technology, there's every chance there's a transistor or 3 involved.
Earlier than that and it's likely to just consist of nothing more
sophisticated than a set of CB points and a 500v 100nF capacitor.

If it's using 21st century technology, the two most common ways, aside
from a mechanical CB points set, were either magnetic or optical
triggering. Since optical triggering was popular on aftermarket
transistorised ignition upgrade kits for 20th century motorcars, the most
likely choice for 'by design' ignition systems built into the later
electronically ignited petrol engines was 'magnetic' triggering which may
be nothing more than a sensing coil embedded into that plastic cover with
a tiny triggering magnet embedded into the flywheel (although the
generator magnets *could* be used for this function, a seperate tiny
magnet offers more precise timing - this may even simply be an extra bit
of iron to divert a fraction of the magnetic flux from one of the
generator magnets rather than an actual seperate tiny magnet).

Careful probing with a small steel rod might reveal the presence of such
a triggering magnetic field 'hot spot' on the inside of the outer rim of
the flywheel. If such a 'hot spot' is detected, you have your answer as
to why there's no sign of a contact breaker.


The carb has no air leaks. It has a rubber bulbous primer that I am
instructed to push three times. I *think* this sucks petrol up out of
the tank sufficiently to fill what appears to be a sort of floatless
float chamber. Its' hard to see how this carb works by looking at it.
All my bikes had Amal monoblock carbs AFAICR, and I could see exactly
how they worked. Anyway, this is a moot point since by replacing the
carb entirely, with no improvement, I think it's safe to assume it's not
a carb or fuel problem.

I will try cranking it over with the plug out in the dark and see how
bright the spark looks within the next 24 hours. If said spark looks
feeble, I will try buying a new plug. If that makes no diff I will try
buying a new coil (or at least, the thing I suspect is the coil)! It
doesn't remotely resemble a motorcycle coil that I've seen (However, my
motorcycle familiarity is limited to 1950-1970 British bikes and
1990-2006 Harley Davidsons. Oh, and more recently, a Honda 125 scooter,
which was IMO, the most practical machine of the lot!!

With regard to testing the sparking efficacy of the ignition system, you
really need to use an air gap of at least quarter of an inch, preferably
a 10mm spark gap at standard atmospheric pressure. The breakdown voltage
per millimetre is roughly proportional to pressure.

The voltage required to jump the the typical 0.8 to 1.00 mm spark plug
gap is only a tenth or less than that needed in actual service. BTW, if
the handbook specifies somewhere close to a 1.5mm gap, you can be
reasonably be sure of the use of a CDI module.

A simple high voltage power transistor switch substitute for the less
elegant electronic version of the Kettering CB points arrangement as used
by the Suzuki GSF600 digital ignitor module - a microprocessor controlled
ignition module let down by the weird choice of electronic CB points over
the technologically correct CDI system - I was *so very not impressed!*,
is just as prone to HT leakage current issues as the traditional CB/
condensor/ignition coil system of yester-millenium.

If you have a spare sacrificial spark plug to hand, you can bend the
earth electrode away from the tip to regap it out to 7 or 8 mm purely for
use as a spark gap tester for ignition systems in general (attach the HT
lead and rest the plug body on any handy metalwork electrically bonded to
the engine if not the engine itself).

Also, a handy insulator such as a piece of thin paxolin or perspex sheet
which can be gently wedged in the spark plug gap to force the spark to
bend around the obstruction for a total spark path length of circa 8 to
10 mm is a non destructive way to check the actual spark plug's
insulation performance after removing it from the engine and letting it
rest in contact with the engine as previously suggested for the test
spark gap plug.

The point is that even a failing ignition system can easily produce
sparks in open air over a 1mm gap leaving you to try and judge the
quality by intensity of the spark alone. Testing with a circa 8 to 10 mm
gap removes such uncertainty from out of the equation.

--
Johnny B Good