Simple CDI failure

[Chriss]

Hi,
i have a project for myself to finish and stumbled into a problem and I can't move forward, so I hope somebody can give me a handheld to make some progress.

The CDI unit gone bad on my 4stroke, one cylinder petrol engine, so I decided to build one very simple one just to start the engine to check if everything is fine.
But for some reason there is spark on the spark plug.

When I attach the ignition coil to a 12VDC and trip one wire by hand then there is se spark on the spark plug.
So I think the ignition coil is ok.

The voltage on the excitor coil is around 5-8VAC RMS depending how fast I pull the starter rope.

The SCR is triggered by the trigger coil.

In the ignition coil when i measure the signal with my DSO i get a signal which is at the beginning around 8VAC but is ringing out.

But whatever I tried to mod on the schematic I cant get a spark.
Any help or suggestion I would really appreciate.

Thank you.

[floobydust]

The exciter coil generates hundreds of volts with a running engine. C1 is usually a 400V part.
If you only get 5-8VAC cranking, it's not enough to charge on the cap to make a decent spark. It is what I 'd expect off the trigger coil.

[Chriss]

I was afraid an answer like yours vos I also suspected that maybe my exciter coil is probably also bad.
But I'm not familiar with these technique and was not sure if that super low voltage can really generate enough voltage to trip the secondary coil in the ignition coil...

Does my schematic looks good enough to use it for a testing purpose maybe?

[floobydust]

There's a few other threads about CDI ignition, I did this schematic for a KX80 and most of the gate-drive complexity is for timing advance and preventing the SCR from retriggering. It was the big challenge to incorporate timing advance and not infringe on other's patents.
But your circuit looks reasonable, if the phasing is correct. Usually the SCR gate has a cap and resistor for noise filtering. Your SCR might be false triggering this keeping the exciter voltage low.

Weak exciter coil output is usually from insulation breakdown giving shorted turns, or a problem with the magnets if they are cracked or damaged.
Metal chips on the trigger pole piece can also give weak trigger pulses.

Check the exciter coil DC resistance and see if it's in spec. and look at the magnets to see if it's a mechanical problem. You could use a resistor voltage divider (say 1/3) and scope the waveforms to see what is going on.

[Chriss]

Thanks for the reply, I appreciate your time.

I tried this:
I took a transformer 220VAC/12VAC and disconnected the original exciter coil and put the 12VAC from the transformer and tried to get a spark on the spark plug by pulling the rope. But still no spark.

I assume, if my minimalistic CDI made by the schematic is ok I should get a spark on the spark plug.

For now I'm not focused on the timing cos I wish to see a spark.

The ignition coil is a simple 3 wire one.
2 wire for the gnd and trigger and the HV wire which is connected to the spark plug.

I also got another ign.coil but still no spark.

I cant see any direction where to turn now.

1. I'm dont trust my design of my CDI
2. I'm not sure what voltage I should expect on the exciting coil.

[floobydust]

The exciter coil puts out hundreds of volts. If you work out the energy in the 1.5uF cap with 12VAC it's to say 17VDC... that is a flea fart. Not enough to make a spark.
Check the kill switch is not contaminated or shorting out as well. Something mechanical is wrong I think.

Suzuki TS185/TS250 CDI Box Replication
See schematic for a cheap chinese CDI module in quads/ATV's, it's not that far off from what you have.

More here:
https://www.eevblog.com/forum/beginners/diode-selection-102979/msg1412985/#msg1412985

[Chriss]

The kill switch is disconnected and not in function, can not affect the working of the exciter coil for now.

My setup is:
The two wires from exciter coil which is located under the magnetic wheel is connected to my CDI like on the schematic.

The TDC position sensor two wires are also connected as in my schematic.

When I measure both signals with my DSO they are well in sync even if I'm not interested right now to be in sync.

Is TIC106 thyristor any good to handle my task?
I think it is, but maybe I'm missing something...

Based on our discussion, I have a conclusion does my DIY CDI circuit should fire a spark if I would have a higher voltage generated by the exciting coil.

Tomorrow I'll be back to my garage and will playing again in this project.

[Chriss]

Guys in the #3 above is a schematic and a C2 on it.
Is that C2 correct polarized in the schematic?

If yes, how will that cap be charged?

The positive side is connected to the one end of the ign. coil.
The negative is connected to the Cathode of the diode.
As Incan understand how it works:

The C2 is charged over the diode before it and untill the scr is not firing.

But in this case the diode will deliver only the positive cycle of the AC signal which is feeding the negative terminal of the C2. In this configuration the C2 can only be discharged but not charged.

Or am I wrong?

[moffy]

For C2, the side connected to D10 will charge to the peak voltage of the exciter. The side of C2 connected to D9 will remain near ground. When  SCR1 is triggered the positive voltage side on C2 gets connected to GND and a -ve voltage appears across the ignition coil and D9, that is the full voltage that C2 was charged to previously. C2 then discharges through the ignition coil, until SCR1 reaches zero current or near enough and turns off and C2 begins to charge again.

[Chriss]

I already took some measurements based on your writings.
And I completely disassembled the whole unit without the original cdi.
The original cdi is destroyed, traces of burnout are visible when I ripped of that black stuff cos it is completely covered with that.

Here are some news:
I could dig out the output stage from the original cdi which was driving the ignition coil.
It is a FQP10N50CF mosfet.

The exciting coil is actually not an exciting coil but a coil which is generating Vpp=12VAC and has a nice sine wave based on the scoping.
Look on the picture.

There is an oil level switch and a TDC sensor which are also connected to the OEM cdi together with the kill switch.

I could also recover and identify two electrolytic caps.
On is 220uF unknown voltage and a second cap of 1000uF/50V.

So, if the mosfet was driving the ignition coil that takes me to the idea does it has something to do with DC voltage.
Bat what?

Probably my cdi design is a bad design for this cos I try to drive the ignition coil with AC voltage.

Does somebody have any idea what setup of a cdi could i make to try to get a spark based on the infos I could dig out of this machinery?

Thanks for any idea.

[mikerj]

Are you sure your "exciter" coil is not for battery charging or the lighting circuit?  Some mopeds us so called "DC" CDI boxes that include a step up circuit to charge the capacitor from the vehicle battery supply.

[floobydust]

What make and model is the vehicle? It does sound like you have the lighting or battery charging coil.
There is no CDI at 12VAC, it's just not enough voltage. There might be a boost converter for that.
Also note the coils (ignition CDI power, ign trigger) are not a sine-wave, usually just a blip and trigger is out of phase with the CDI coil - you want to charge up the big cap, then trigger.

[Chriss]

It is a 1kw gasoline generator.
A friend of me talked to me today and maybe he has a "cdi" unit like mine was on this engine and if he found it he will send photos about for reverse engineering it if possible vos that unit is also bad but they had lucky and could open it.

Btw.
This unit has no battery.

[floobydust]

You can use a heat gun to soften and pry off potting compound. I would check the stator part number if it's also used on larger gensets that have a battery.
There must be a high voltage boost-converter in the module that sounds like it croaked.

[Chriss]

Ok, here I'm with some news.
I got the picture from my friend abhout the "CDI".

His "CDI" is also unusable cos some part's are missing but however, maybe we can figure out something.
Based on the information I got from the engine and the picture I come to a conclusion:

This ignition coil on this engine is a 3 pin coil.
pin 1 and 2 are the primary coil
pin 3 is the HV wire to the spark plug.

The mentioned "CDI" is not a CDI, I don't know what is the correct name of this ignition controller but it is not a CDI.
Based on this I think I have to make a kind of controller with a MOSFET which is switching the primari side of the ignition coil,
depending when the TDCS ( Top Death Center Sensor ) triggers the momentum whebn the piston is on it's top death position.

I also assume I need to rectifie the AC voltage from the "exciting" coil, cos that coil is not a real exciting coil,
it is actually the power source to power.


Description of the PCB from the picture:
Pin 1 is directly connected to the pin 1 on the ignition coil.
Pin 2 is the oil level swithch which is in open stage if the oil level is ok, othervice it is shorted to ground.
Pin 3 is the TDCS sensor. The TDCS is a two wire sensor. One end is on the gnd and the other is on PIN 3.
Pin 4 is connected to the engine housung which is gnd.
Pin 5 is one wire from the "exciting" coil, this is actually a AC voltage source to power the electronic ignition coil controller.
Pin 6 is the kill switch.

The picture from the scope is the signal when I pull the rope and the motor start to rotate and the piston is always in the TD center when I have this signal.

So, I have to collect somehow the TDCS signal and trigger the MOSFET.

Question:
1. Is it enough to collect the positive signal side of the TDCS?
2. any example or idea how to proccess the signal I would really appreciate.

Thank you for any help and/suggestion.