How to detect if stop button on magneto coil is pressed?

[alec_t]

If you have a PIC which can toggle a pin at ~500kHz or more then that could substitute for U1a. However, depending on the (unspecified) inductance of the magneto coil the voltage developed across it at that frequency may be too small for the PIC to sense without external amplification. It is small because of the necessity for a high series resistance to prevent (a) damage to the PIC from the 150V and (b) loading of the coil. Once the signal is amplified the charge pump could be replaced by pulse detection in software.

 

[ccurtis]

You should do an experiment to determine what's the smallest value of resistor you can place in parallel with the stop switch/magneto without significantly affecting the motor operation. Lower value resistances can mean less amplification/parts needed.

 

[Futterama]

I once measured how much current I could draw from the coil while the engine kept running, as far as I remember, it was around 100mA. I don't really know how my multimeter read that, if it was peak or RMS. But it is quite a bit anyway. I would say, without further experiment, it would be perfectly safe to draw 1mA from the coil.

 

[ccurtis]

You can try this. The idea is to adjust the frequency of the oscillator (pic output) so that it equals the resonant frequency of the series LC network of C2 and the magneto inductance. Then you get the amplitude required without amplification. When the stop switch is open, pulses at the oscillator frequency at required amplitude for the PIC input appear at the output. With the stop switch closed (magneto shorted) a solid low level appears at the output. The oscillator frequency (and resonant frequency) should be selected to be considerably higher than the RPM of the motor so that the circuit does not resonate while the motor is running, minimizing the load on the magneto during normal motor operation.

You would have to experiment to determine the resonant frequency since the inductance of the magneto is probably not known, but 10 mH is probably a good starting point. You adjust the oscillator frequency and/or C2 to get the highest amplitude pulses at the output.

 

[alec_t]

.....and here's a simplified version of my earlier circuit. C2 and C4 (high voltage types) replace the 100k resistors and afford protection of other components from the 150V AC. Their use also means that no amplification of the signal across the coil is needed. Simulation shows this should work for any coil inductance from 50uH to 50H, so it's non-critical. A PIC could replace the parts to the left of line A and right of line B or C.

 

[Futterama]

We are getting very close to a final solution, I'm impressed by how you take your time to draw up the whole circuit and do simulations ect.

I will make sure to report back when I have tried the circuits.

 

[Futterama]

alec_t, would it be possible to replace the four 1N5818 with two dual diodes e.g. BAT54S from Fairchild?

Link to BAT54S datasheet: https://www.electro-tech-online.com/custompdfs/2013/05/BAT54S.pdf

 

[alec_t]

Those should be fine.