The Voltage can be between 6 and 24 Volt, depand the Voltage of the ignigtioncoil.
There is only one problem, if you use 24Volt the timerboard need a extra resistor to reduce the Voltage for the 5Volt regulator.

Rob, I have everything working for the New CDI as a single board....built the RCD version 2.3 programmer and it works fine..loaded the hex file, programmed the pic16f628 chip. installed a TIL111 optocoupler. Installed the PIC chip, powered up with the top jumper removed, got fast mode test just fine. Reinstalled the top jumper, removed the bottom jumper, got the slow speed test function working...unplugged the battery, replaced the lower jumper and powered up again. Tested the CDI with the hall effect and magnet..worked perfectly !! Thanks..

Only one problem afterwards...with all jumpers installed and battery plugged in, it shouldn't be sparking..but it seems like the low speed test jumper is removed and it gives a slow tick....tick...tick...spark on it's own...what should I be looking at, a bad part or is there a problem elsewhere in the programming? Any help would be appreciated... At least it worked perfectly the first power-up..thankful for that... I have used the LP2950CZ 5 volt low dropout regulator which is that same size as the BC557 transistor...it's rated for 800 milliwatt..should be fine for the chip power.. maybe the BT151 scr is leaky??? Had a few off spec parts give me that problem a few years back...anyway, thought you and the rest of the forum members here would have liked to hear a success story for a change.. It worked, and I owe it to you and JMJ for a great project!

John

It looklikes the scr (BT151) live his own life, try a other one.
I know from others, if the Voltgae from the HV-coil is to high, the BT151 can't handle it.
The high Voltage can only when the CDI is running in low RPM - test 300 RPM.
I'll change the low-test in the program to 1000 rpm, maybe this will safe the scr.
Others use a FET or a IGBT, I don't have problems wih the BT151.
IGBT - IRG4BC30K0 or some other N-channel IGBT

Rob, it seems I have found the cause of the unit sparking by itself!!! As I tested it the first time the battery was fully charged fresh from the charger and voltage was 5.6 volts... after first testing was completed, about 20 minutes time had gone by. Well, as it seems, the unit is very power hungry and consumed enough voltage from the pack (4- 700mah nicads) so that the end voltage after the LP2950CZ regulator was below its dropout voltage and the unit was resetting the pic startup. As I see it now, I can't go to 5 cell pak because overvoltage does the same thing as under voltage..creates it's own sparks.

So I have found the cause... now to find the cure! Oh, by the way..I used the New Cdi to run my newly converted weedeater motor today...P E R F E C T !!! Absolutely flawless! Now if I can figure out the voltage drop problem, I'll have a perfect running system.... Everything you have changed over the last 6 months was added to this PCB and I must say, if someone followed this forum and knew how hard we all work...especially you Rob... Thanks a Million !

Here's some pictures of my work over the last 3 weeks!

Hello,

Did you make the toroids or did you buy them? Thats a lot of work.

Charlie

I make them here.. I have a few blanks left out of the 200 I bought.. lots of wire used in past 3 weeks!

All of my previous boards were made without timer boards.. I used the CH Synchro Spark on most of mine for auto advance and also work extremely well. I think I bought one of Terry Grant's first batch of production boards and they are all still working! That says alot for design and attention paid to detail.

I have a PCB file and schematics for my CDI's I make here. Although the parts placement has changed over the years due to different sized boards, all components I use are the same. As for the Spark coil, I also bought some from CH electronics which are so small and light weight, it's hard to believe the spark that you can achieve from them. It's in a black plastic potted form about 1 inch square. Can be seen in one of the photos on the motor. Still looking for the original source of those however.. also pilfered the seperate coil from a McCollough 28 cc weed whacker magneto..also worked well. Zenoah coils work well too. If anyone has a source of coils I'd be interested in looking at them on the website. Found a few for around $20 which work..S&S machine also sells one already wound and ready to wire up. In fact, I believe it is what's inside the square one I use.
The Hall Effects I use are Allegro part# A1104LUA available for around 79 cents each I think, from Jameco.
Honeywell SS443A are ok to use also.

First of all
If JMJ wasn't start his poject, I start never building this CDI.
JMJ is also one of the guys who write the code for the PIC and make the Exelsheet.
Lot of others help us with all kind of solutions to finish this project positive.
Your all a part of the project, even if you only building the CDI and ask for help !

NiCad-cells have a effective Voltage of 1.2Volt, 4 cells = 4.8Volt
This Voltageregulator have a drop of =>0.5Volt, regulator needs 5Volt + 0.5Volt = 5,5Volt minium
There are cells from laptops (Lithium) who have 1.6Volt / cell, 4 cells = 6.6Volt (some powerdrop somewhere - use 7.2Volt)
I test the CDI many times and if the ocsilator runnig smooth, the CDI consummed ~0.5Amp/h - 6.5Volt

Pffff......I see 8 HV-boards, you gone use them all by your self ?

And if someone will try something, http://www.modelbouw.gompy.net/tci/

I would also like to give great credit to JMJ for his major contributions to this project! It takes alot of hard work and time to achieve results with a project like this. Hy hat is off to all who contributed !

The amount of CDI units I make here in a week go to most of the club members of the TCWS. I get calls all the time for boards..mostly I would give them away and also free help with hookup. Just something I do locally for friends. There is always someone in need and I can't refuse the chance to get someone else started..

Years ago I got together with Floyd Carter to produce the RCM ignition module board...which was fun..enjoy working with someone like that..that got me started and have been doing it ever since..

For those interested in making a TCI unit, there is a kit from Velleman....K2543 which is low cost...many sources from suppliers for this kit.. I have one which I built and it works fine for cars, bikes, and is changeable from 12 volts to 6 volts with resistor changes..all info in the kit for doing that. I still think CDI is better!

Rob, I can't resist the TCI project !
Got all the parts here to start except the IGBT...will have one shortly..itching to test this one on my 1973 Harley Sportster !

Here's a TIM-4 TCI hooked up to my Harley Hall effect sensor timing plate which I have been running on the '73 Sportster for a year or so. Used a 6 volt regulator to step down from the 12 volt ignition voltage to 6 volt for power to it. Anything's possible if you have an imagination!!

If you have a spare timerboard, you can use it.
You can also connect the exist timerboard to the Vellemankit.

IGBT's are hard to get here, I ring a lot of shops before I found one (!).......I can't make mistakes with onle IGBT []

Here you can find my bike www.kreidler.gompy.net
Thats why I'm making the TCI, a old ignition isn't in the store anymore.

WOW ! lots of work on that bike...looking good ! [8D]
My Sportster will have new life with a new ignition too.. a picture at present time will have to wait till I empty the garage.. 6 bikes and a 4 wheeler in the way!

Rob.. what regulator chip did you use on your CDI..?? mine is the small to-92 package LP2950cz... looks like yours is the size of the BD151 thyristor...

TO220, but you can also use a smaal one.
I can't get a small version, and allmost every store can sold you the TO220

Intresting http://www.vems.hu/wiki/index.php?page=CdiSupply a specialy the way the make the HV.

What an elaborate setup you have there for testing ! I use a VOM only..and my point to point wiring.. I'm ashamed to show you my work area..but then sometimes simple is best for most people.. My soldering iron is never unplugged!

Rob, what value is C6 on the component overlay? didn't see it on the schematic...is there a substitution for the IRG4BC30k0 transistor?

I choose a IGBT I can maybe buy, but you can take any 600 V / 8 A IGBT (even een FET)
At the time be I use a BUT12A to do the job.
I make space on the PCB for C6, you never know, but I don't think you have to use him.
Into the Vellemankit is also a C (0.22uF).
BTW, you can use the Vellemankit too, connect the output of the timerboard on point 2, remove R1 and R2 and replace a dummy R (0 Ohm) for R3

Attention, you have to make some changes into the C-code if you gone use the TCI !!!
A normal sparkcoil will fire-up for 45% of the time, CDI-fire-uptime is 56us to fire-up the scr.
If your engine runs 16000 rpm, the pulstime wiilbe ~4ms for the hole puls.
If we don't change the fire-uptime, the sparkcoil isn't fire-up long enoufg.
The dutycycle we have to make is ~35 till 45% if we change the delay to 1500us
50Hz = 3000 RPM = 20ms = delay_ms(10) or delay_us(10000)
3000 rpm = delay_ms(10)
6000 rpm = delay_ms(5)
9000 rpm = delay_us(3300)
12000 rpm = delay_us(2500)

ONLY CHANGE THE RED TXT INTO YOUR CODE - DON"T COPY THE CODE BELOWE !!!

-

void test() { //Generate test signal
while (1){
PORTB &= 0xee; // sends a ground to the SCR gate
delay_us(1500); // 56 us delay (Change to 1500)
PORTB |= 0x11; // ground off the SCR
if (PORTB & 0x04) delay_ms(200); //300 rpm
else delay_ms(8); //7500 RPM
}
}

void main() {
const char tspark[]={ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 3, 3, 5, 5, 7, 7, 9, 9, 10, 12, 12, 12, 14, 14, 14, 17, 17, 17, 20, 20, 20, 22, 23, 25, 25, 28, 28, 31, 31, 34, 34, 37, 38, 40, 41, 44, 44, 44, 48, 48, 48, 51, 52, 52, 53, 56, 56, 57, 57, 61, 61, 62, 62, 63, 63, 63,
64, 64, 68, 68, 69, 69, 70, 70, 71, 71, 72, 72, 73, 73, 74, 74, 75, 75, 79, 80, 80, 81, 81, 82, 82, 83, 83, 84, 84, 85, 85, 86, 90, 91, 91, 92, 92, 93, 93, 94, 94, 95, 96, 96, 97, 97, 98, 98, 99, 99, 100, 100, 101, 106, 106, 107, 107, 108, 108, 109, 110, 110, 111, 111, 112, 112, 113, 113, 114, 115, 115, 116, 116, 117, 117, 118, 118, 119, 120, 120, 121, 121, 122, 122, 123, 123, 124, 125, 125, 126, 126, 127, 127, 128, 128, 129, 130, 130, 131, 131, 132, 132, 133, 133, 134, 134, 135, 136, 136, 137, 137, 138, 138, 139, 140, 140, 141, 141, 142};
short int spark=0;
unsigned int tspk, t0,t1,low;
unsigned int ofc; //Holds number of overflows

const unsigned int LowDelay=25; //Delay <1832 RPM, in degrees

const unsigned int ld=360/LowDelay; //Dont change this
TRISB = 0x0E; // set RB1,2,3 for input, all others for output RB0=SCR RB4=Strobe
PORTB = 0x11; // set gate to SCR high i.e pin 6 of PIC +strobe
OPTION_REG = 0x06; // tmr0 SOURCE INT CLOCK, 1:128 PRESCALER
INTCON = 0x00; //Disable interrupts +Resets interrupt flags
TMR2 = 0; //Reset timer2
TMR0 = 0; //Reser timer0
PR2 = 255; //This is how far timer2 counts before overflow
T2CON = 0x07; // start TMR2 counter 1:16

if (PORTB.F3) test(); //If test jumper is on jump to test function

while(1){ //Loop forever
ofc=0; //Reset overflow counter
INTCON = 0x00; //Disable interrupts +Resets interrupt flags
while(PORTB & 0x02){ //Wait for low pulse on Hall switch
if (INTCON&4) { //Overflow
ofc++; //Increase overflow counter
INTCON = 0x00; //Resets interrupt flags
}
spark = 1;
} //North pole of magnet has passed
if(spark){
t0 = TMR0;
TMR0 = 0; //Start new rev count
if(ofc > 0) { //Overflow timer0, we have < 1832 RPM
if ( ofc >= 36) ofc=36; //Min RPM 50 to avoid big calculations
low=ofc*ld; //Simplified calculation, gives about 25 degrees
delay_cyc (low); //extra delay based on overflow counter
delay_cyc (low);
delay_cyc (low);
delay_cyc (low);
delay_cyc (low); //Repeat 16 times - prescaler
delay_cyc (low);
delay_cyc (low);
delay_cyc (low);
delay_cyc (low);
delay_cyc (low);
delay_cyc (low);
delay_cyc (low);
delay_cyc (low);
delay_cyc (low);
delay_cyc (low);
delay_cyc (low);
} //max255 -> 2550 cycles
tspk = tspark[t0]; //use table lookup for spark delay
T2CON = 0x07; //start TMR2 counter 1:16
TMR2= 0; //clear spark timer
while(TMR2 < tspk);//loop until timeout to send spark
PORTB &= 0xee; //sends a ground to the SCR gate
delay_us(1500); // 56 us delay (Change to 1500)
PORTB |= 0x11; //ground off the SCR
spark = 0;
} //if(spark)
} //while(1)
} //main

////////////////////////////////////////////////////////////////////
//Notes
//
// ; At a engine speed of 1200 RPM;
// ;
// ; 1200 / 60 = 20 revs per second
// ; 1 / 20 = 50mS per rev
// ; 1 rev = 360 degrees
// ;
// ; 36 degrees at 1200 RPM = (36 / 360) X 50 = 5mS
// ;
// ; 100uS loop time
// ; 4MHz clock / 4 = 1MHz instruction cycle
// ; cycle time = 1/1MHz = 100uS
// ; 100uS/100uS = 100 cycles
// ; prescaler set to divide by 4
// ; preset TMR0 = 231, when = 0 = 100uS
// ; 231 to 255 = 25 x 4(prescaler) = 100


// -O-
// RA2 o o RA1
// RA3 o o RA0
// RA4 o o RA7
// RA5 o o RA6
// VSS o o VDD
// SCR RB0 o o RB7
// Hall input USART RX RB1 o o RB6
// Test hi/low USART TX RB2 o o RB5
// Test on/off RB3 o o RB4 Strobelight
// -

//History:
//1.0 1832fix not working

This is how the ignition looklike with a 2500us delay, normaly it has to be 4286us

OK...typed changes into c-code..saved it...now to finish the board!
Scope trace would show changes too.. thanks Rob !
Trying to locate my scope program to display signal trace on the computer..now where did I save the program??? Hah !
Must be having a memory lapse in my old age..

Rob..
I noticed one other thing with the timer board..when the magnet is approaching the hall effect it turns on and makes a spark...if that is normal operation, ok.. but to compare it with my systems I make, the spark waits till after the magnet leaves the hall effect. Not a problem really.. will adjust timing position manually the extra 10 degrees of rotation to compensate for the early trigger if needed. Just thought I'd ask..

John