T2 can only supply current not sink it, so there is no path to discharge C3 when T2 turns off. Thus you need to add a resistor from T2's source to ground.

 

What is the purpose of C2 and C3? How can they discharge?
When T1 turns on D2 will be toast.

 

The circuit I posted on #10, and the parts C2, C3, D2, D3, they are a double voltage rectifier, and I also lost a Rgs to turn off the mosfet, when you using it in your new circuit then the voltage will be double when the circuit is right and works.

The circuit I drawn as below, you can change the Q1 to N ch, Q2 to P ch mosfet, but the value of resistors that you need to do some adjust as R1 = 100 Ohms, R2 = 4.7K~10K, all the value of resistors will be related to the frequency and flows current when you change the values.

 

D1 will die when FET1 turns on, unless the 400V source impedance is very high!

 

jpanhalt- i forgot to add a resistor between ground and 12v aaround the d3 area. also i have to lookup what a high side config is again, i remember reading somthing about it.

crutschow- yes i see that now, that makes sence thanks.

alec t- i forgot to add an resistor so they can discharge, and the reason for them is so you can turn the frequency inwitch the coil runs at, i belive they are around 8 hz, im not to sure about it. but ive read about them in a few places. also you say d2 will be toast if T1 gets turned on. want to explane that a bit more, im not sure why i thought the diode was good for 1000v. also why would high impedance ac help

scottwang- thank you very much for the drawing and explnation, i may make that circuit if i can make sure it works and no one sees any problems with it. also i have a bunch of 2n23904/6 so il leave it with transistors.

 

Still don't understand the resistance to using a SCR, since thats the way most commercial CDI's are made. It can be triggered by your micro controller with no muss or fuss. That said, here's some more thinking outloud for you

http://www.rcgroups.com/forums/showthread.php?t=1781959

Silicon Chip magazine ran several designs years ago, think there still available on their website. Issue of the mag -Dec 1995, Sept 1997, Jan 2006

A schematic found on Wikipedia

 

i see what your saying and how cdi's use scr's and im not against them. i was just trying to update it to be more mordern and design my own (i know, its silly not knowing abouthing about electronics) honda in the late 80's stoppped using scr's in thier cars ignitions. and now in most all motocross bikes and sportbikes they use transistors. basicily most has moved to TCI ignition. so in a sense im looking to make a TCI ignition that runs off 300 volts oppsded to 12v and a cap for inline. since this is stationary and i can use any power supply i can come up with, the motor deosnt have to make it. i wanted to use over 12, also im not sure why i just get ideas in my head and run with them

ive seen the Silicon Chip magazine cdi's and a whole lot of others(i have probably 25 pdf files with diy cdis on them), i notice they go from using maybe 8 components to some designs upwards of 50. so they can be simple or advanced, but im not looking for any advanced or retard ignition in the cdi its self or a programmable cdi, i just need it to have a constant recharge ready to fire rate, i can account for it in the program on the arduino,

so i guess i want a tci that runs high voltage and has a cap that discharges. im not sure why i want this honestly i just feel it would be cool when showing and telling other people about it, and also then it would be designed by me (well more or less you guys) since ignitions like this arnt common if even exist

ill order some scr's to have around to play with, im not opposed to them. thanksk for the link

 

also you say d2 will be toast if T1 gets turned on. want to explane that a bit more

D2 is the wrong way round!

 

Did you ever made that cdi sir////