Can anybody explain the prescaler config for timer0?
I would like to change it to 64 in the high rev/high resolution version of the code to get even better accuracy.

This is the line I'm talking about:
OPTION_REG = 0x06; // tmr0 SOURCE INT CLOCK, 1:128 PRESCALER

I want to try 1:64 prescaler, but the manual is not clear enough for an amateur like me..

OPTION_REG = 0x06 = 128
OPTION_REG = 0x05 = 64
OPTION_REG = 0x04 = 32
OPTION_REG = 0x03 = 16
OPTION_REG = 0x02 = 8
OPTION_REG = 0x01 = 4
OPTION_REG = 0x00 = 2

Thank you !!!!

Did you check the excel sheet I emailed you? It has the delay compensation incorporated etc..

Will this excel sheet be posted?

Just athome but I don't look into the sheet jet only a quick look.

<span id="result_box" class="short_text"><span style="background-color: rgb(255, 255, 255);" title="Ti mi&eacute;rt nem haszn&aacute;l?">Why do not you use? F</span></span><span class="short_text" id="result_box"><span title="Application note" style="background-color: rgb(255, 255, 255);">igure2.3. or figure4.5.+ </span></span>Precedent<span class="short_text" id="result_box"><span title="Application note" style="background-color: rgb(255, 255, 255);"> photos.Regards nyemi.
</span></span>

That is only theorie, not realistic.
I don't have find any schematic like the figures 2.3 or 4.5

I see you found a schematic, but thoose schematics are all for AC-Voltage coming from trhe HV-coil into the engine.
Ground isn't ground as we use ground from a batterie.

there is some interesting application notes in there all the same....thanks

Thank you kernet.

<span id="result_box" class="short_text"><span style="background-color: rgb(255, 255, 255);" title="Ez okoz h&aacute;l&aacute;l">This cause of death SCR!!?? </span></span><span class="short_text" id="result_box"><span title="J&oacute;l &eacute;rteni.">
</span></span>

Rob those circuits have a diode on one side which you would get DC from wouldnt you



nyemi i need to try some more protection for the SCR you are correct.....especially when experimenting

brett

If you are talking about the two diodes in series then yes that would rectify the ac voltage.

Hi Brett,

Two diodes in series make a half wave rectifier. Not very efficient. 4 diodes in the normal pattern make a full wave bridge rectifier. From My experiments the full wave rectifier is most efficient.

Here is an example of code I wrote and tested that gives good resolution and very low rpm cotrol.
The expression in bold text makes it possible to read an endless array with delay values. Code delay with this solution is 80us.
The limitation is the amount of memory. I'm running timer0 at 1:64frequency, and when I generated values down to 1000rpm, the memory wasfull.
However, on low RPM engines, there is no need to run timer0 faster than1:128, so this could actually give a good result for all engine typeswhen it comes to ignition accuracy.
The best thing would be if it was possible to rescale the rpm positionson low rpm. Now there are almost infinite number of values for low rpm.The resolution is actually rediculous on low rpm's..
Creating the enormous amount of values in Excel also demands some reworking of the sheet..

I also tried a multiplication factor on the number of tspark[255] times a constant from owerflows without enlarging the array, but I could not get it to work..

Well, just food for thought

while(1){ //Loop forever
ofc=0; // Reset overflow teller
INTCON = 0x00; // Disable interrupts + Resets interrupt flags
while(PORTB &amp; 0x02){ //Wait for low pulse on Hall switch
if (INTCON&amp;4) { // Overflow
ofc++; //Increase overflow counter
INTCON = 0x00; // Resets interrupt flags
} // End IF
spark = 1;
} //North pole of magnet has passed
if(spark){
t0 = TMR0;
TMR0 = 0; //Start new rev count

if(ofc &gt; 0) t0=t0+(ofc*255); //
tspk = tspark[t0]; //use table lookup for spark delay

T1CON = 0x05; // start TMR1 counter 1:1 using internal clock 1MHz
TMR1H = 0; // Reset timer1 high
TMR1L = 0; // Reset timer1 high

do //start of timing loop
{Hi(t1)=TMR1H; // read timer1 Low bits and add to timer variable
Lo(t1)=TMR1L; } // read timer1 High bits and add to timer variable
while(t1 &lt; tspk); // loop until timeout to send spark

PORTB &amp;= 0xee; //sends a ground to the SCR gate
Delay_us(1650); // Calculated delay
PORTB |= 0x11; //ground off the SCR
spark = 0;
} // end if(spark)
} // end while(1)

Charlie
if you only use half the diodes or even just one to rectify the AC wont that give you a higher output voltage as the peaks would still
be there alittle ?


thanks for helping sam, i havnt got a timing board and power board together yet for testing,



brett

Hi Brett,


NO. With a half wave rectifier (2 diodes) you still have some ripple and that is bad because that is current that is not being used. If you want to really increase the voltage try a voltage doubler.

OK

With help from gr8flyer I think I have finalized the circuit I want to build. I'm going to keep the transformer secondary voltage around 400 volts, but all the components on the power board will be rated at 600 volts or better except the 1N4007 (1000 volts). I eliminated everything except what is needed to fly an airplane. The goal is for easy starts at 300 RPM and advance up to 10,000 RPM, even though the engines seldom see 8500 RPM. Shielding is still a concern but I wont really know until i get it built. I notice the Chinese CDI's are fully shielded, but the Zenoah ignitions have the coil, spark plug wire and cap unshielded. We'll get it figured out.

gr8flyer is helping me with coil. Once I have that I'll finalize the PCB layout and start etching. I'll post progress as I go.

Al Z

Al, is this the braided shield you needed? Oh by the way, that's your toroid in the picture..

John

<span class="short_text" id="result_box"><span title="L&aacute;tni fejleszt program,&uacute;j &ouml;tlet.">Software develop a new idea(PWM-CDI). </span><span title="K&eacute;t link:">Two links:1. http://btelectronic.sweb.cz/cdi2.htm 2. http://veyrat.perso.neuf.fr/allumage_electronique.htm
</span></span>

Speaking to a TV technician today says we should not be using a SCR for switching but a transistor, there are heaps of good transistors in old TV sets hi voltage and current

he said with a SCR the Gate has to be discontented and maybe even the Anode or cathode to stop a lock up, where as with a transistor it wont lock on the same,

Also i think what ever you put into the circuit is what you can get out , to a degree

my circuit draws 100-150 ma's ....... this is why i think i have a weak looking spark even below 10,000 rpm

Even the commercial units draw between 350 and 600 ma's ( http://www.ch-ignitions.com/faq.html )

An intense spark is performance as far as i am aware and a weak spark is what i already had with the flywheel and magnet setup

the motor can run fine with a weak spark But will run alot better with a full powerful spark or even twin plug setup as on my microlight.

http://www.youtube.com/watch?v=QpevOS0VQmk


Brett

Awesome!

Interesting stuff.

I was particularly interested in the second link since it is an RC airplane application.

Some comments & observations (food for thought):

1. Uses a zener for on-board voltage regulation.

2. He's bringing the Hall sensor (I assume it's a Hall) into the RB1 pin of the PIC so obviously the programming is quite different than the circuit in our discussion.

3. He's using a 100 volt, 1.5 amp surface mount MOSFET for the oscillator.

4. He also uses a surface mount bridge rectifier (no part number specified). Surface mounting makes for neater layout but PIA soldering.

5. Uses a thyristor to fire the plug - no component specified.

6. No RFI shielding!????[>:]

7. No feedback winding on transformer required with this circuit.

8. Current consumption did not translate well to English, here it is:
15 mA - 0 RPM
220 mA - 3000 RPM
270 mA - 6000 RPM

Just a quick look at the other link I noticed they use two MOSFET's in the oscillator and a thyristor to fire the plug. I guess a thyristor is the right device for this application.

Al Z

Software support two links:1.www.micro-examples.com/public/microex-navig/doc/097-pwm-calculator 2.www.mikroe.com/en/books/pic-books/mikroc/ch3/#c3v7

I'm not sitting stil, this is wat I want for the new CDI.
I hope soon the guys will go to program

Rob that looks good, what hz did you make it oscillate at ? 3khz is a nice number

i was thinking of using a 555 timer for switching the primary coil, can the pic be programed to do the same i guess ???


and i like the IGBT driver haha , all moden day heavy duty igniter coil's use this transistor

Thanks nyemi for the links........ what language is it ???

brett