Jim's Cushman Scooter Site

An Electronic Ignition for the Cushman Scooter

By Ben Sloan

The restoration of my 1959 Highlander 721 included major repairs to the original motor. It had already been bored .060 oversize, so obviously had lots of hours on it. The photo below gives an idea of its original condition. After the mechanical work was completed it became apparent that, even with a new coil, capacitor and breaker points, the ignition voltage at cranking speeds was weak. In an effort to keep the motor as nearly original as possible and not add batteries and the like, it was decided to build a capacitive discharge ignition operating only from the original Wico 12 volt magneto.


The Original Cushman Highlander

The basic strategy was to replace the ignition coil with an exciter, or source, coil producing, at speed, about 300 volts DC. This coil generates an output for each of the six magnet passes per revolution regardless of the state of the points and so it will generate more total energy than is associated with the single magnet passing under the ignition coil for the standard inductive ignition. The voltage from this exciter coil is rectified and stored on two capacitors, one for the ignition pulse and another for generating a trigger pulse from the points. The description is for the Wico 12 volt system but the circuit should be applicable to the Wico 6 volt system and many others with appropriate adaptation of the exciter coil.



The Homemade Exciter Coil


The exciter coil form, or spool is made with very thin (1/16") plywood from a hobby store for the ends and the center by cutting a square hole thru a hardwood dowel with a mortising bit. The inside hole does not need to be square but should fit around the stator iron snuggly. There are several pictures of this coil. I used 2400 turns of 31 gauge wire on this form. Winding was done using an old fashion hand drill clamped in a vise to spin the coil form. It is best to roll the wire onto the coil form and unroll it from the source spool rather than take it off over the end of the spool which can twist or snag the wire. After winding and attaching insulated leads, soak the entire coil assembly in some good varnish. The pictures show 4 leads exiting one end. That is because I wasn't sure how many turns I would finally need and I put taps at 1600 and 2000 turns. The insulated leads need to be long enough to reach wherever you want to locate the ignition. Neither side of this coil is grounded.

12 Volt Cushman Stator Plate with new Exciter Coil Installed

Remove and discard the standard ignition capacitor and run insulated wires from the breaker points and a good grounding point on the magneto out to the ignition. My ignition is located under the engine cowl and is essentially invisible when assembled. The new ignition coil, pictured below, is a coil-on-plug Ford product and is available for less than a new Cushman replacement coil. Remove most of the long rubber snout on the coil leaving just enough to get a good grip on the plug wire, as shown in the picture.

Ford Ignition Coil as Purchased, Before Modifications to the Rubber Snout

I used a project PC board from Radio Shack for the assembly. [The shunt regulator, to be discussed, was added later so you will see two boards in the pictures, but all functions can be included on the same board.] Because of the high voltages used, it is a recommended that the resin flux from soldering be carefully cleaned away and several coats of Krylon 'crystal clear' applied to both sides of the board when it is completed and tested. My board has been mounted using industrial strength Velcro which both isolates the board from the metal cowl and allows it to be easily removed.


This is a photograph of the completed ignition system mounted in the top section of the shroud. The heatsink for T2 is shown at the lower left, and the Ford spark coil mounted on a bracket that I made for it is at the lower right.

Schematic Diagram


It is easy to understand the operation of the ignition system if it is broken down into three separate sections.

A) The power source consisting of the exciter, its rectifiers and a 500 ohm isolating resistor R1. The exciter provides A/C voltage from its 2400 turn coil to the full wave bridge [D1 through D4] which rectifies it and, through the 500 ohm resistor charges a 10 UF, 450 volt electrolytic capacitor, C3. The voltage on C3 is used by the pulse forming components to turn on the pulse transistor, T1.

B) The shunt regulator consisting of transistor T2 and diodes D9 and D10.This regulator is used to limit the supply voltage to approximately +300 volts as engine speed increases. Below +300 volts this circuit is inactive.

C) The ignition pulse generator consisting of T1 and C1 and the gate drive components associated the ignition points. Diode D5 and D8 isolate the two storage capacitors and prevents C3, the 10 UF capacitor from being discharged when the ignition transistor fires. The 500 ohm resistor effectively isolates the discharge components from the output impedance of the 2400 turn coil and bridge.

The wave forms from the Wico stator and magnet architecture result in voltage waveforms with a high peak to average ratio, probably 4-5 to one. For the ignition one needs good supply voltage at cranking speed but it must be limited to something the components can handle at high speed. The solution is to use all 2400 turns for good output at cranking speed and limit output at high speed with the shunt regulator. When the voltage from the exciter bridge reaches 300 volts, the Zener diodes, D9 and D10, conduct and raise the gate voltage of T2 which turns the FET on. This transistor has an on resistance of less than 1 ohm, much less than the 500 ohm resistor, so it effectively clamps the power supply output at just about 300 volts.

The same type of power FET is used for the ignition portion of the circuit. The gate drive for this transistor, T1, comes from the 75-300 volts across the 10 UF capacitor. What is needed is a fast rising pulse, but one much less than 300volts. When the points open, current through the 39K resistor causes the voltage on C2 to quickly increase from zero to +10 volts where it is clamped by reference diode D7. This fast rising voltage is coupled to the gate of T1 through C2 and this turns on the discharge transistor T1 for a short time. Resistor R2 then discharges C2 and permits T1 to turn back off without regard to the status of the points. T1 has an internal high current diode protecting it when the voltage across the coil reverses, so no additional components are needed. The discharge of C1 through T1 and the coil primary is what produces the ignition spark.

The current through the points is low with no inductive 'kick' so they should last virtually forever. Timing can be done with an ohmmeter across the points.

The pictures show the two power FETS on commonly available heat sinks and they are mounted in a lot of moving air so heating is not an issue. The ignition transistor dissipation will be low because its duty cycle is very small, but the shunt regulator transistor might get hot during long runs at high speed with no cooling. Note that in the example pictured the shunt regulator is on a separate board below the main board.



Bench Testing the Unit Before Installation

If you have access to electronic test equipment, the unit can be operated on the bench. Apply about 50 volts DC to the input side of the 500 ohm resistor and attach the points input to an open collector NPN driven from a signal generator. The output is best observed with a spark plug. The voltage across the points gives a good indication of basic functionality and should be a rectangular wave form rising from zero to about 10 volts. The spark will be modest at 50 volts DC input, it takes about 75-100 volts DC input across the storage capacitors to generate a good spark. Keep the input pulse frequency low...about 1 pps...to avoid overheating components.

If you do not have test equipment is it still possible to check basic operation off the scooter. Attach the circuit board to the ignition coil with a spark plug installed. To the exciter coil inputs attach the leads from a 24 volt doorbell transformer, furnace thermostat transformer, 25 volt filament transformer (Radio shack 273-1366, $6) or equivalent...up to a maximum secondary voltage of about 100 volts RMS. But DO NOT hook directly to an A/C outlet under any circumstances! Operate the ignition by shorting the points input to ground and observing the spark plug when the points open. With a 25 volt RMS input, the discharge capacitor will have about 35 volts DC on it and produce a rather weak spark. AT 100 volt RMS the spark will jump about a 1/4 inch gap and can be heard in the next room.

Of course the ultimate test is on the engine. After assembly, use a low voltage test light or, preferably, an ohm meter to time the ignition....just as with the original ignition. Then leave the plug out and grounded to bare metal and spin the engine. You should see a spark every revolution.

This ignition seems to work well with a hotter than normal plug using a standard or slightly larger gap. Note: if the flywheel fan sucks water into the inlet it will be sprayed on the electronics, so if you use it under those conditions, the board should probably be enclosed in a plastic project box or located elsewhere! My old Highlander now starts on the second kick every time!

The parts list shows reference numbers for Digikey, but other (mail order) sources are available. The total parts cost of the electronics is about $30-35 dollars and the coil about $14.


Here is how it all came together!

    Top of Page      Return to Home Page

Added 7/31/10