Related Posts with Thumbnails

Last Updated: August 20, 2010

DC-CDI schematic (updated)

DC-CDI counterpart of AC-CDI is an ignition analog or digital that uses low voltage external power supply to work. the difference AC-CDI needs external High Voltage COIL no battery configuration whereas DC-CDI needs a battery without HV COIL.
DC-CDI schematic diagram

Looking at the picture on the left, it is a complete schematic diagram of a dc-cdi. Its a 4-pin system, comprising pick-up input, battery +12 volts in, Gnd, and Ignition coil out pins. There is no High Voltage  input pin like those of AC-CDI. If it will be differentiated with the AC-CDI here, you will notice theres a transformer diagram on this schematic.The circuit on the RED BOX comprises the internal High Voltage generator of a DC-CDI, where the 12 volts battery in will be converted to 200-400 volts depending on the design of the inverter. This design uses two transistor dc-ac converter with external on/off circuit to charge and discharge the capacitor BLUE BOX via SCR. the HV converter is controlled by the circuit on GRAY box, that as soon as the pulse conditioning circuit sense the pick-up coil will send triggering signal on both SCR and HV oscillator respectively to turn them on and off at the right time.

CDI or Capacitor Discharge Ignition, from the name itself, capacitor is very critical during its operation, must handle charge and discharge time as well as heat being generated by the whole circuit inside the block box, that is why manufacturer uses special capacitors specifically made for ignitions, and not just like those mylar capacitor found on some electronic circuits. They are rated 400 to 630 volts. ranging from .47uf up to 2.2uf. Value is also so critical in every applications and designs. In this circuit, they use 1uf / 400 volts with HV out of 200 volts will equals to 20 mj.

Capacitor Energy:
Energy stored in capacitor is a function of V and Capacity:  
E (joules)= V^2 * C * 1/2.
For example: for 1uF capacitor, charged at 200 volt, E = 0.02000 J = 20 mJ.
Values near 50 mJ are standard, values higher than 50 mJ are high power

Capacitor value may be critical because it should match with the power of the HV circuit to enable to charge itself at maximum rpm. If capacitor is too large it may not charge enough at high rpm. And if capacitor is too small, spark energy can be lower.


The circuit above with the help of sir Robert Long, carefully put some annotations to fully understand how it works, the necessary waveforms voltage check, and the graph of the advance,

Many thanks to  sir Long.

Annotation Courtesy of Mr. Robert Long


ajxavr said...

great post... is it possible to handle more than 10000rpm??

can i put link to your blog from mine?? thanks before...

LEI said...

possible if at any how we can modify the HV that it can cope with the faster time for the capacitor to charge...

Abu-Hafss said...


I assume you have worked on the multi-spark CDI circuit published in Australian, Silicon-Chip Magazine in 1997.

The basic topology was to convert one pulse from the pickup coil into a burst of (4-8) pulses and then fed into the HV capacitor thru the mosfets.

I was just wondering, if we simply use a 555 to convert a pulse into a burst and feed it to the simplest AC-CDI of Yamaha DT125 would not it work?

Nathaniel Berdan said...

for reliablity issue of the high voltage converter section of any dc-cdi, the moment the scr is trigger to dump the energy of the capacitor to the ignition coil, the HV oscillator must stop momentarily in accordance and or in sync with the scr gate triggering. IF the HV converter keeps on feeding HV to the capacitor the SCR might lock up, producing misfire. If you look at the schematic above, both the gate of the SCR and the HV oscillator is connected to pin 1 of the lm339 IC, by analyzing that circuit. when pin 1 is ready to trigger the gate, the HV converter driving circuit will be turned off momentarily.

I had seen that multi-spark CDI articles many times, i also do found their group at yahoo, only few made it work, and they are having problems with misfire, the high voltage keeps on feeding high voltage even if the scr is already dumping the energy stored in the capacitor to the ignition coil.

Try bench testing such cdi and you will see what i mean.

try visiting this site

All information are there. thank you

Abu-Hafss said...

Yeh, I have that point in my mind. The HV supply can be blocked momentarily with a triac using the inverted signal from the pick-up coil (i.e. the triac will conduct when there is no signal from the pick-up coil).

Here, I am not talking about the actual multi-spark CDI but, the 555-based as mentioned earlier.

Abu-Hafss said...


Last night I posted a comment here but, not yet appeared. I repeat again below:

I dug out an CDI for Yamaha. Here is the picture for reference.

I replicated the same with slight modification. Here is the schematic:

The value of Q1 could not be read, it was damaged during striping.
The red color shows the values I used in my circuit.
The green color shows my addition.
My transformer is slightly bigger.

The unit is working perfectly, except that Q1 is getting hot. Adding a heatsink makes it run normal but, you know very well there are no heatsinks in the CDIs. Do you have any idea, why the power transistor is getting hot?

Abu-Hafss said...

The HV converter is almost identical to the one you have posted in this article, except that there is a slight difference at the network connected to the feedback winding.

By the way, can you tell me the value of the power transistor which is used in your schematic?

Abu-Hafss said...

Also please inform the power rating of 470R resistor connected to the feedback winding.


Nathaniel Berdan said...

what value did u use for the unreadable Q1 Based on your schematic, you use the tip 31c? is this correct? if yes then it is heating up because that type of transistor is not for high speed switching applications..more often, it is used for amplification purposes only..470R is a wirewound type resistor 1 watt...

for Q1 try using high speed type such as C4813...this transistor has an internal damper diode which is an ideal choices for dc-dc converters using transformers..Vcbo-100v, Vceo-100v Ic-7.5amp....

Abu-Hafss said...

Thanks for your swift response.

The CDI had transistor TIP**** but actually it was damaged during striping. Only I could read TIP so, I tried TIP41C. I'll try 2SC4813 and revert.

In the meantime, today, I
1) Removed R2, D3 and C3.
2) Replaced R1 with 470/1W such that one end of R1 is connected to the feedback winding and the other end to the base of Q1.
3) Added a 56V/1W zener - anode at base of Q1 and cathode to the collector.

This combination slightly improved the heating but, 470/1W also getting hot.
Next, I gradually increased the value of R1 upto 10k.
At 10k, it remained cool but Q1 got hot.
In between 2k and 3k thing getting under control but still I am not satisfied with the heating of Q1. May be 2SC4813 would do it.

Further, when I checked the windings of the original transformer, it was as shown in the schematic. However, I doubt I might have made a mistake in counting the turns of feedback. So, I am thinking to reduce the turns to 26-27. What do you say? Can you tell the number of turns for the transformer shown in your article?

Nathaniel Berdan said...

try visiting this page

it is an OEM dc-cdi of my small motorcycle, and the HV generating circuit is somehow similar above..take note that the OEM uses SMD type of component..none of the parts heats up, only getting warm.

how did you get the number of turns of the dc-dc converter transformer? did u dismantle the original transformer? Did you use any sorth of inductance meter to show the original inductance?

The above inductance of the transformer windings are as follows.
primary winding--30 turns (45-50uh)
bias winding--30 turns (45-50uh)
secondary winding--150 turns (1mh)

May i ask your intention of building the CDI above and or replicate your own cdi for what? The cdi has a fixed advance and no complex than those PROGRAMMABLE DC-CDI that uses flyback mode converters like the transmic or gompy cdi. I am just so curious of your intentions.

Abu-Hafss said...

Yes, I dismantled the original transformer.
The inductance shown is for my transformer.

Today, I could get 2SC4242 and another 100V-5A-100MHz (I don't remember its number now). Both got hot immediately.

So finally, I decided to reduce the number of bias winding to 27 turns. The final adjusted inductance (using air gap in the transformer cores) were:

Primary: 22 turns - 85µH
Secondary: 140 turns - 3,560uH
Bais: 27 turns - 120uH

Luckily, it resolved the issue. Now, TIP41C is super cool and the R1 (100R/1W) gets warm. Changing it to 2W made it more warm and also the transistor.

My basic intentions are:
a) to get control of the DC converter
b) to replace AC-CDI of Honda CG125 which looses power when engine gets hot on mountainous road.

I know that these CDIs have fixed advance and also well aware of Transmic, Gompy and Sports Devices.

Thanks for your attention. In future, I plan to work on regulators. If I have any complications, I'll get back to you.

Abu-Hafss said...

Ah yes, one more question please.

Does engine power (cc) has something to do with CDI?
In other words, can a CDI for 70cc be used with 200cc?
I have already used 70cc CDI with 125cc without any problem.

Nathaniel Berdan said...

as far as i am concern, the answer is NOPE. irregardless of displacement, as long as they have the same ignition mapping stored, and the correct timing plus the polarity of the input will thing is not for sure though. A 2 stroke CDI will not work on 4 stroke engine and vice versa.

Abu-Hafss said...

This CDI has driven me into all sort of troubles.

The bench test was ok. It performed perfectly on the three-wheeler for two days. After two day, the engine started missing. When the original CDI was re-installed the engine ran normally.

Kindly refer to my schematic posted on Sep. 23rd.
First I removed, R12 and C9 (in green box) but, no improvement.
Replaced C8 with new one, the output got back to 322V.
But within moments, R5 roasted and output dropped to
I replaced C4 with new one and removed one zener, the output came to about 240V.
But after few moments, it raised to 500V+ and the transformer emitted sharp whistle sound.
Now the situation is that when it powered on, the output climbs more than 500V for a fraction of a second and then drops down to a few volts. I have also replaced Q2 and Q3.

Can you please help me out?

Nathaniel Berdan said...

the only thing i can find questionable within your schematic was the orientation of one of the zener diodes connected at the output of the HV converter back to the oscillator circuit, try to compare it to

OEM shogun HV converter section that can be found here

Nathaniel Berdan said...

Abu-Hafss said...

In the OEM circuit, a 200V uni-directional TVS diode was used. I couldn't get a unidirectional TVS diode so, I used 4 x 82V zeners which later I reduced to 3 x 82V.

Moreover, the OEM circuit did not had 1N4007 (or similar) with cathode facing the cathode of the topmost zener.

I have seen large bulky capacitors in the CDIs for dumping charge into the ignition coil. However, I have been using this capacitor:

If my circuit had any problem, it wouldn't had worked for two days in the vehicle.

By the way, can you tell the value of C6?

Abu-Hafss said...

The HV AC output before the fast diode was about 700-900V when the circuit was operating fine. And now it goes over 1000V. I don't know how UF4007 (rated 1000V) is bearing those 1000V+. And after UF4007, at zeners, it goes to about 500VDC.

Nathaniel Berdan said...

sorry for the late reply, the value of C6 on my OEM original HV generator circuit was 100n, i had to dismantle that SMD capacitor to determine the value using capacitance meter.

Abu-Hafss said...

Sorry for the extra job, and thanks for your cooperation.

I had noticed that the Shogun circuit did not had any configuration synchronized with the triggering pulse to pause the oscillation. Furthermore, that circuit only had a 330R and 100nF for conditioning of the pulse.So I decided to simulate the circuit in LTSpice. Since I didn't knew the inductance of the transformer on OEM Suzuki Shogun, I used the one which you mentioned on Sep. 24th. The simulation revealed that the oscillator stops when the capacitor is charged. As soon as the spark is fired, it starts the oscillation again. That is very smart approach, it should save the power transistor for getting over-heated.

As far as the Yamaha circuit is concerned, I am just confused with the fact that if it could run normally for two days then what could be the problem now? I have almost replaced all the components except the power transistor and the transformer. I am just getting fed up. So , I am just thinking to re-build a fresh circuit for Yamaha and also for Shogun.

Or do you have any other suggestion?

Nathaniel Berdan said...

It is true that the oscillator is being stop by the MCU when the capacitor is dumping the charge to the ignition coil by the SCR, the approach is not actually new, it is how the high voltage converter works on any DC-CDI. To prevent the SCR from locking up during discharge.

Why not try building the transmic DC-CDI that uses pic micro and uses the flyback converter topology instead.

Abu-Hafss said...

Hmm, that means the Shogun's CDI is actually controlled by MCU.
But, what I am saying is that the oscillator stops itself when the capacitor is charged. And it waits for the trigger signal to dump the capacitor's charge to ign. coil. When the charge has dumped the charge, the oscillator resumes.

In the simulation I did not incorporated anything (MCU) to stop the oscillator. If you want I can send you the LTSpice files, you can see it yourself.

Nathaniel Berdan said...

check this link

the updated schematic of my OEM shogun cdi

two outputs from the MCU is controlling the HV oscillator and gate of SCR independently, if i killed one pin there is no output and no spark, don"t rely too much on your spice, work it out on the bench real time. I already remove my original MCU for i am wanting to incorporate the GUMPY cdi that can be configured by a PC software so that it can change the ignition mapping instantly..but up until now i have not able to make it work. It did work with transmic cdi though.

Nathaniel Berdan said...

I did experiment and isolate the HV converter of my cdi, and without any trigger on the oscillator, it does'nt start, so how come in your LTspice start it up, just for curiousity.

ryan said...

Is it possible to trigger this box with a DC pulse? And if so what is the minimum voltage required? I'd like to trigger this box from a programmable module.

Abu-Hafss said...

It cannot be triggered with simple DC voltage however, a square wave or triangular wave pulse can be used to trigger the circuit. And most important, the above circuit is designed to get triggered on a negative pulse. So you have to use a voltage inverter to convert the positive pulse into negative pulse. A -5V pulse would be enough to trigger the circuit.

paman doblang said...

mohon bantuannya,untuk mengubah limiter dalam rangkaian diatas bagaimana caranya? terima kasih.

Nathaniel Berdan said...

@paman doblang

it is impossible to remove the limiter since the cdi above is not a programmable cdi. Thank you

paman doblang said...

CDI ini untuk shogun 125 cocok tidak?

paman doblang said...

CDI ini cocok tidak untuk shogun 125?

Nathaniel Berdan said...

@paman doblang

saya tidak mengujinya pada shogun 125 , tetapi jika u ingin membangun itu , saya pikir itu mungkin bekerja karena tingkat advance adalah 30 derajat sudut

Hoàng đảng said...

Was können Sie Diagramm ic 135 excitrer durch intelligente senden danken

Nathaniel Berdan said...

I do not have the cdi schematic of 135 exciter. Sorry

Hoàng đảng said...

ic-Diagramm, das Sie nicht haben, geben Sie sich nicht

Hoàng đảng said...

Vosz#33 said...

Is there a way to modify or override or even fake the inputs so you may get rid of rpm limits? Thanks

Nathaniel Berdan said...


none, input isn't the concern and got nothing to do with the power limit of any cdi.

vipin upadhyay said...

I have studied that circuit but i like to make a programable dc digital cdi. so plz send me a way to design dc digital cdi.

Abu-Hafss said...

>>>vipin upadhyay

You may try visiting:

Energy Science said...

This circuit seems to require a battery on the bike - is that correct?

How do race bike CDI's work if they have no battery? Is the cap charged only by the magnet on the flywheel? Doesn't seen to be enough to get a cap up enough for these purposes. Is there a diagram of this somewhere that includes how it interfaces with the ignition coil/magneto and the spark plug?

Abu-Hafss said...


Racing Bikes are usually equipped with programmable CDI based on microcontroller, for better performance. If the bike has HV generating coil it will use a AC-CDI which does not need a battery. And those bikes which do not have the HV coil uses a DC-CDI which definitely needs battery. For racing bikes, I have heard they use 9V rechargable battery which is light weight.

For more information and schematics for programmable CDIs (AC or DC) please see post of Feb. 10, 2015.

Nathaniel Berdan said...

energy science A dc-cdi can work even without a battery if the Voltage regulator of the Stator is functioning properly..For race bikes equipped with programmable Cdi, either with AC or DC type do not use large battery to lessen the weight. Mr.Abu-haffs is definitely correct. Some uses the 9V battery since race bikes (I am talking of small bikes here and not those MOTO GP and or WSBK ones)..since they do a limited laps on track and a fully charge 9volts battery is enough.

There is also a hint on using DC-CDI on a race bikes, with a healthy voltage regulator, replacing the battery with a large value of capacitor will work well.

Jack bowman said...

Primary: 22 turns - 85µH
Secondary: 140 turns - 3,560uH
Bias: 27 turns - 120uH

maksudnya bias tu apa ya?

Jack bowman said...

bias yg mana, primari yg mana, secondary yang mana?

Pradeep Kumar said...
This comment has been removed by the author.
Pradeep Kumar said...

Miracle Electronics has a highly capable team of engineers and technicians that have expertise in manufacturing best-in-class and highly safe power transformers in India that are RoHS compliant, Reach compliant and available with CE and UL marked with request.
Transformer manufacturer in India

bis kalamen said...

"Bringing Dead Batteries Back To Life Is Simple!"


"Save Money And NEVER Buy A New Battery Again"

"Bring Nearly Any Dead Battery Back To Life Again"

Over 19,000 people are already using the EZ Battery Reconditioning Method

"Make thousands by purchasing old batteries and selling them as reconditioned ones for massive profits"

And it doesn’t matter if you’re not technical or don’t know the first thing about batteries …because our course is incredibly easy to follow and absolutely anybody can use it.

For more info please click on the following link: "Bringing Dead Batteries Back To Life Is Simple!"

Talk soon.