land-speed
[Top] [All Lists]

Re: magneto

To: LGMCAFEE@aol.com, land-speed@autox.team.net
Subject: Re: magneto
From: ardunbill@webtv.net
Date: Sun, 17 Mar 2002 12:54:59 -0500 (EST)
Larry, mags have always been dear to my heart, the basic mystery of how
they could generate accurately timed sparks that increase in voltage
with rpm (up to a point; this is not a serious consideration) and keep
on doing it indefinitely, as a self-contained device.

In thinking more about this discussion, it occurs to me that the
high-tension system in a mag is grounded until the points open, which is
what causes the current to then jump the spark plug gap.  So if one of
your mags fires before the other, it strikes me that the HT current will
just go to ground back up the other mag rather than jump the gap.

You are right, we read that with CD systems, a series of sparks jumps
the gap for some duration of crank rotation.  But I seem to recall that
the same thing happens with a mag to a degree, not sure about that.

There are a lot of things that can be faulty with the magneto to
interfere with its performance, such as internal shorts in the coils,
faulty condensor,  low magnetism in the magnets(which can be
remagnetized or "re-charged" as they sometimes say), and resistance in
the breaker points circuit due to corrosion or burning or corrosion in
the wiring connections.

A simple test on the kitchen table should be able to prove any magneto's
performance.  The Vertex manual says the spark must be able to jump a
5/16" air gap at the terminal end of a one meter long plug cable, when
the mag turns at
100 rpm, for a four-cylinder unit, and 85 for an eight-cylinder.  I
imagine it should also be able to do the same thing when flicked with
your fingers on the drive gear. It is interesting that Vertex says their
racing mag should be able to give the same performance up to 4000 rpm
(8000 crank rpm) for the four-cylinder mag, and 4500 for the six and
eight cylinder mag(9000 crank rpm).

All the old timers seem to agree that with a Vertex, with the cap off,
if you clamp a piece of steel onto the body of the mag and put the end
of it up 1/2" from the end of the coil wire, it should jump a spark
there when you turn the drive end with your fingers.  Don't shock
yourself!!  If it does, the mag is good to go.  If not, it needs the mag
shop.  The famous Harman & Collins Flathead racing magneto of yore will
produce this same amount of spark, in good condition.

The Brit bike makers used to say if your plug wire (using a Lucas mag)
could jump a 1/8" air gap, the ignition was okay.  In working on many
Lucas mags, I have seen some that could generate 1/4", but no more.  Yet
this was enough to produce World Motorcycle Speed Records.  We're told
this represented about 10,000 volts. It may have called for a spark plug
gap as little as .011" in a very high (13 to 1) compression engine,
according to the factory instructions for racing Vincents.  I once had a
Harley Sportster with a Fairbanks-Morse magneto, and it was a super hard
starter, but I never had any experience working on the F-M mags.  The
one on my Harley may have needed attention.

It would be interesting to know how much spark you can get from your mag
on the kitchen table.

I recently interviewed vintage GMC 6 fuel and gas racer J.D. Tone for
BRN, and he shed new light on the use of Vertex mags for me.  I
understand from him that the traditional Vertex is sufficient for every
unblown gas application on the GMC, and works well on the fuel GMC up to
a point. But that when the nitro percentage goes way, way up, the
new-style Vertex with the external coil will give substantially more
power because the higher amperage spark simply allows more of the nitro
to be burned in the cylinder, compared with the original style of
Vertex.  It's not that the old-style Vertex will mis-fire, just that the
new-style will give more power.  
 

Cheers Bill

///  unsubscribe/change address requests to majordomo@autox.team.net  or try
///  http://www.team.net/mailman/listinfo
///  Archives at http://www.team.net/archive/land-speed


<Prev in Thread] Current Thread [Next in Thread>