_Shut off_
A spring pulls the wick down into the oil. (no wick = no fire)
At operation a small solenoid pulls it up. (= auto shut off when power failure)
The solenoid/spring arrangement can be in the oil. (antifreeze and lubrication)

_Ignition_
For my grandfathers inventions i once made a hot spark device that can even ignite wood stick by spark alone!
The trick is to make such frequent sparks that the air get very hot, and you will have an arc through air plasma.  Looking very nice BTW.
Make that spark cross the wick, and it wil heat up to oil boiling point and then ignite in one or 2 seconds.

I guess you can use what i used for the prototype: a moped engine ignition coil!
(probably other types will do too, like a lawn mover engine coil)
Normal car ignition coil have too large inductance to be charged easily quickly enough.  On the moped ignition coil i cut off the ends of the core (the part that extends to the magnet wheel) to further reduce inductance.

Then i connected "ground" to +12V, and the wire normally for the comb switch i connected to a high voltage power switch transistor to ground.  Fed it wit a couple kHz and about 90% duty cycle.  A cap about 100nF across the transistor.  Be sure to arrange for quick shut off, so power is not lost.
BU208D NPN such as found in many TV sets vill do good.  Or even better, a mosfet rated min 500V (or more depending on your coil, cap, and gap), and max 2 ohm on resistance.
 Beware of high voltage peaks when transistor shuts off - some hundred volts primary.  Always connect secondary to spark gap, or you will blow the isolation in the coil, or tranny, or...  Also of course use decoupling caps.

"Comb switch": probably wrong word - i mean the switch that is operated by the comb to fire the spark each engine turn.

Protect so it is impossible to touch the ignition wire.  I never tried, but i guess the power is anough to kill a person with a weak heart!

For oscillator you can use the standrad '555, plus suitable discrete drive stage.
Ose an oscilloscope and put in a series resistor to monitor waveform of the current, sou you can see when the coil saturates, and experiment with shorter dead-time.
If you have or make a high voltage probe, than also monitor primary voltage.  Experimenting with dead-time/frequiency and capacitor you can turn on transistor again as soon as voltage dips after spark discharge to optimise.
While experimenting: keep an eye -or rather finger tip- on the coil core temperature!

Careful!

OK... you could use aPIC instead of the 555, but you need more components for IC supply, adjusting, and more drive stage components...

/Morgan

Ian Hooper 18:05 2004-11-25:
>Here's a rather strange one:  I've had the clever notion that I'd like to
>operate my garden lights by means of a PIC-based timer. The catch is that I
>wish to use oil lamps, not the usual 12V bulbs. My dillema occurs in trying
>to devise a method of electronic ignition for an oil lamp; so far my
>experiments have produced unimpressive results by using a simple spark
>ignitor. Has anyone else toyed with such a system and have any ideas of what
>I might try to safely, and consitently light an oil lantern?
>
>thnx
>
>ian
>
>
>_______________________________________________
>http://www.piclist.com PIC/SX FAQ & list archive
>View/change your membership options at
>http://mailman.mit.edu/mailman/listinfo/piclist 

--
Morgan Olsson, Kivik, Sweden

_______________________________________________
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist