I'm dealing with a few "issues" with my first JEH.

I work in an electronics shop so I'm slightly fortunate to be able to do some "tests" durning lunch and break.

I'd like to offer a couple things to think about and maybe get w8ye to get more specific.

The design of the JEH TIM4/TIM6 is just like the schematic shown on 5 Bears site as the TIM4.
The basic difference is in the resistor values and HAL Part Number. Jerry also adds an LED for "timing" purposes. I.E. you can see it flash when the magnet passes the sensor.

The function of the circuit is basic.
The magnet turns the sensor on and passes a low (.5v) to a transistor (2N2907A) that turns on the TIP42C. The TIP42C acts sort of like a switch, applying battery voltage directly to the coil. While this voltage is applied, the coil builds a field. When the voltage turns off, the field collapses and causes an induced voltage to be applied to the secondary side of the coil. This collapse is so rapid that it created a very high voltage pulse that is passed to the spark plug. ZAP...Spark.

So, here are the issues to address:
1. Dwell.
As a function of time, the size of the hub holding the wheel is very important in this circuit.
Jerry gives some formula info but it's not going to cover the entire RPM range.
The longer the magnet is under the sensor, the longer the voltage is applied to the coil. Too long, and it burns up, too short and you don't get a hot spark.
Translation: if you set your hub at a diameter good for 5,000 rpm you can burn up your coil at idle and not run very well at higher rpms. If you set it for 2,000 rpm you'll preserve the coil but probably won't get any rpms out of it.
2. Coil damage:
The Modelectric PE coil that I have came with a paper that indicates that 2/1000 seconds (2 ms) is the optimum time for voltage to be applied to the coil.
To quote the paper: "Power applied to the primary winding longer than 2/1000 seconds does not produce a stronger spark. It does waste battery power and more importantly, it causes heating of the coil winding."
I also believe that the TIM6 may be applying too much voltage to the coil. The coil is designed for 3.6-6.0 volt applications. But running it at the lower voltage would certainly lower the current and prolong the life of the coil.
That said, I could not get my TIM3 to fire with 3 NMHi AA cells. (4.1 V)Using my bench power supply at 4.0 volts I got great spark. Converting it to a TIM6 (2 resistors) and powering it with a standard flight pack (4 AA NiCads) at 5.2 V produced spark with a vengence! So, I'll try to address the 3 cell issue another time or perhaps, by fixing issue 1.

So, here is my idea for fixing issue 1. and this is where I need help.
The idea would be to eliminate the change in dwell that happens when RPM changes, or when different hub designs are used. By the way, distance between the sensor and the magnet also effects dwell time.
To eliminate the change, we need to trigger that TIP42C for 2ms every time the magnet passes under the sensor. If we could use the magnet/hal sensor to trigger a single shot IC...the IC creates an output that is .5v or less for 2ms. This output, when applied to the existing 2907a base would provide a more precise/desired result.
Because the circuit would only fire once for each passing of the magnet, it would not "stay fired" if the magnet stayed under the sensor and would not burn up the coil.

I'd bet it can be done with 1 IC and an R-C (resistor-capicitor) network.

Summary
We need a circuit with the following:
.5v input trigger.
Static High (5.2 V) output with a single 2ms .5v low for every trigger.


Tom