Roll Your Own CDI

[Toadmund]

Here ya go.
I haven't used it yet, but if it works as good as it looks, it should be OK.

PS, the plug boot is different than shown, but I don't care.

And I just verified that there is indeed a screw the wire tightens on, just like the stock one.

 

[nightcruiser]

Here ya go.
I haven't used it yet, but if it works as good as it looks, it should be OK.

PS, the plug boot is different than shown, but I don't care.

And I just verified that there is indeed a screw the wire tightens on, just like the stock one.

Thanks for the link and the info, that sounds like it will work out great for me. Let me know how the coil works for you when you get it wired up.
Only thing is that coil is coming all the way direct from china, I am gonna try to search out someone that is selling the same coil in the US if can...

 

[Toadmund]

They are all made in China, but if you want to pay someone for getting one from China, be my guest.
I got this in 12 days.
For $10.00!

 

[nightcruiser]

They are all made in China, but if you want to pay someone for getting one from China, be my guest.
I got this in 12 days.
For $10.00!

I know they are all made in china, just dont want to bother with the international shipping. There seems to be quite a few sources for the same "looking" part here in the US for the same price with free shipping...
Thanks for your input, perhaps this time around I will get the CDI built....

 

[Toadmund]

When I was about 21, I bought some books from Radio Shack and bought some resistors, LEDs, transistors, diodes, capacitors, 555 timer chips and other chips.

I understood some basics, but the transistor confused me, and I knew capacitors charged and discharged and other things and etc and what not.

So here I am almost 23 years later and combined with the Internet and wanting to know how this CDI unit works, I know a whole lot more, and when you understand how each part functions, it is so much easier to know how they work together, how electricity will take the path of least resistance, it will take the 100ohm resistor path, but not the 560, making a switch like a transistor or an SCR take a capacitor off charge, and switch it to discharge, and all that.

I could go on and on, just wanted to say, motorbicycling.com got me soldering again, I made my own CDI, going to hook it up, then I am going to make a better one.
I also wanted to bump this thread, at one time it was much more popular and I want to keep it alive, because it's good stuff and a good learn.

And my bike will run better!

 

[maniac57]

My eyes glazed over long ago but I keep coming back to learn a bit more each time.
This time I learned Toadmund can solder!
Yay Toadmund!
Solder me a ignition Toadmund! Pleeeeeeeeze?

 

[[email protected]]

Hey Everyone

i've built 2 CDI's following the circuit diagrams and parts list supplied by "dracothered" first one i made was to figure it all out and it worked great in short tests but died after a 5km ride. the next one i knew what to do so was easier and on the closest to jag settings did about 50km, but felt less powerful then stock CDI so i changed the jumpers to the more Advanced setting and after about 5km it died. not quite sure what is going on or what components i could check. any advice would be apreciated thanks!

 

[whipaway]

output capacitor should be rated at least 400V.
I think that there are some posts in this thread where the 'scoped output was seen to be in the 300s.

 

[[email protected]]

Thanks ill swap out the capacitors and give it another go!

 

[dracothered]

I noticed that your board is etched on the wrong side, the circuit foil should be on the back of the board. So are you sure it is all wired up correctly?

 

[[email protected]]

It was my first time etching my own board I never thought to have it on the back would have been much easier to solder it all up if I'll be sure to do that on my next one! I'm sure all the components are in the right spot and are firmly soldered in place, as far as I can tell none of the connections are contacting where they shouldn't be. It seems odd to me that it only failed once I changed the jumpers over.

 

[dracothered]

It was my first time etching my own board I never thought to have it on the back would have been much easier to solder it all up if I'll be sure to do that on my next one! I'm sure all the components are in the right spot and are firmly soldered in place, as far as I can tell none of the connections are contacting where they shouldn't be. It seems odd to me that it only failed once I changed the jumpers over.

You may have caused damage to the parts because it is harder to solder them on the top side. The resistors and caps aren't damaged as easy but the other are by the heat.

 

[Huffydavidson]

I saw this reply posting this morning and the first I did was call my FIRST customer and did a follow up call . He told me the bikes never run so good and his long riding time was about an hour and a distance of 25 miles or so stop and go traffic . His is 3 months old.

 

[dracothered]

I saw this reply posting this morning and the first I did was call my FIRST customer and did a follow up call . He told me the bikes never run so good and his long riding time was about an hour and a distance of 25 miles or so stop and go traffic . His is 3 months old.

Is this based off of my drawing? If it is I would say it works good then.

 

[Huffydavidson]

dracothered, I follow the recipe to the tea. Feel free to look at my photo albums and get back to me . I order enclosures and boards made for the enclosure boxes . My boards don't have foil, I tied the parts together by their pig tails. Thank you for helping make a couple bucks. If can ever do ANYTHING for you just pm . I'll Skype if you like.

 

[dracothered]

Thanks for the confirmation that the circuit works well and your welcome about the help.

 

[reapercest]

interesting.

ive had this schematic floating around, and recent events have had me returning, and finally experimenting.

i tore down a magneto, and found the wire soldered to the laminations isnt a centre tap but the start of the entire coil. a major flaw, i feel...

but it returned me to the cdi scheme of things in the search of more get up and go!

five minutes, and the things point to point wired. about the size of my thumb. pretty neat

so, i tried the 33 ohm and 100 trimmer(200 was all i found) and, when hooked up to a handy 6v c90 coil, worked! that blows my mind, that it works on a standard points type induction coil!

then the resistors burnt and fried. heh heh heh.

up graded to the 18 and 82 version. fried

ok, heavier resistors... and besides one really hot resistor, im getting a spark with only 15volts! a multitap transformer is so handy at times. if only i had variable frequency too!

30 volts is nicer im guessing most of my frying issue now is due to using a 1 amp transformer at only 30volts and 50 hz... IR^2 and all that nonsense about duty cycles? they take a lot of current on one cycle!

the stock cdi doesnt do anything, and i dont have a transformer over 30v handy to push it either.

wondering what would happen if i stuck it on mains via an isolator at 240?

anyways. ive noticed OP hasnt been around for some time... thanks all the same dude.

i shall do a run for real tomorrow!


just, some things i noticed went dead ended...

on the tear down, there appeared to be a zener diode in the circuit?

ive been contemplating this avenue before reading all this, so i picked up on it...


but what i just wrote i deleted as i can see now that it would have no effect on the advance curve at all.

so what the bejeesis is it doing in there? it never was mentioned again

oh well. try my hand at depotting.


plans? oh, i got plans...like a twin spark ignition coil, properly potted for all these multi plug heads ...?

a basic dyno in the near future for mod testing.

and such forth nonsense

 

[dracothered]

interesting.

ive had this schematic floating around, and recent events have had me returning, and finally experimenting.

i tore down a magneto, and found the wire soldered to the laminations isnt a centre tap but the start of the entire coil. a major flaw, i feel...

but it returned me to the cdi scheme of things in the search of more get up and go!

five minutes, and the things point to point wired. about the size of my thumb. pretty neat

so, i tried the 33 ohm and 100 trimmer(200 was all i found) and, when hooked up to a handy 6v c90 coil, worked! that blows my mind, that it works on a standard points type induction coil!

then the resistors burnt and fried. heh heh heh.

up graded to the 18 and 82 version. fried

ok, heavier resistors... and besides one really hot resistor, im getting a spark with only 15volts! a multitap transformer is so handy at times. if only i had variable frequency too!

30 volts is nicer im guessing most of my frying issue now is due to using a 1 amp transformer at only 30volts and 50 hz... IR^2 and all that nonsense about duty cycles? they take a lot of current on one cycle!

the stock cdi doesnt do anything, and i dont have a transformer over 30v handy to push it either.

wondering what would happen if i stuck it on mains via an isolator at 240?

anyways. ive noticed OP hasnt been around for some time... thanks all the same dude.

i shall do a run for real tomorrow!


just, some things i noticed went dead ended...

on the tear down, there appeared to be a zener diode in the circuit?

ive been contemplating this avenue before reading all this, so i picked up on it...


but what i just wrote i deleted as i can see now that it would have no effect on the advance curve at all.

so what the bejeesis is it doing in there? it never was mentioned again

oh well. try my hand at depotting.


plans? oh, i got plans...like a twin spark ignition coil, properly potted for all these multi plug heads ...?

a basic dyno in the near future for mod testing.

and such forth nonsense

Go here http://motorbicycling.com/showthread.php?p=418243#post418243 it should give you some more insight on the CDI.

 

[reapercest]

zener to limit voltage to gate on scr. just what i was thinking.

im running over the most basic circuit shown here, that does work, but wondering if theres one slight problem in it. waiting on my batch of scr's to arrive so i can play some more. buy 10 for the same price as one locally. just gotta wait


i do have to say it worked a lot better than i expected. running smoother, revving nicer, generally perkier all round

where the resistor(s) join to the scr and two diodes, somethings telling me its better to have the scr joined only to the 1n5407, and the resistor join only to the 1n4007... im just thinking it would be, it tends to make a more logical "flow" in my mind.... til i test i cant say.

cooking resistors, definitely due to using a mains transformer! works fine on the bike with 1W. in hindsight, one full half wave goes right through those resistors! i note the 82 blows first...

but heres my plan. i like the 12v power supply idea.

so, i want my gate trigger to come from a fixed reference voltage via a zener.

then, it basically doesnt matter how much more current i try to draw from that half of the cycle, as my trigger voltage wont be affected by the load. this means my advance/retard curves are unaffected. does this make sense?

im no electronics guru. i just know how to solder and google datasheets...

of course, i dont expect huge currents, just enough to run my 4W spotlights and the red tail leds!

 

[reapercest]

tried this idea of removing the aforementioned join. resistors not running to junction of bt151 and 1n5407, but directly to base of 1n4007 then to "blue".

doesnt work. some delving into theory of thyristors explains why. im used to tubes that require no current flow. just a voltage present on the grid...

thyristor requires a "sink" on the cathode for gate current to flow to.

the 1n5407 forces current to flow from black through resistors, via the 1n4007. a small voltage feeds off and flows through the gate, and drains to negative which at that time is the blue wire. thyristor then "shorts" the main cap, via 1n5407 and step up coil... so 18r and thyristor gate/cathode are actually in parallel.


but!

1n4007 has approx 0.7volt drop... so the trigger voltage has to be slightly higher to trigger.

lower voltage, maybe a germanium diode, and trigger would be closer to the zero crossing point, or advanced.

same time, using an LED with 1.2v drop would retard...

am i thinking correctly here?

i really need to pull the scope out and start checking this stuff...