Model engine CDI easy and cheap

[paulc]

G'day Ray,
Bluejets circuit is designed to allow a Hall sensor to take the place of an induction coil for triggering the DC 12 volt CDI. So the small Hall driven interface has to emulate a"sine wave" as a opposed to the Hall sensors square wave, right?

 

[paulc]

The KY-003 Hall Magnetic Sensor will work with up to 24v and already has a pull-up resistor and LED.

Ray

Oh so the resistor on the KY-003 is a pull up? Its role is important in the circuit and is not just for the LED?

 

[Claudio Masetto]

Ok after looking at the specs. of the KY-003 I realise that it does work by grounding the signal terminal when it senses a magnet. I was thinking of something else.
You need to put a 1K pullup resistor from the +ve terminal to the output terminal of the sensor.
You also need a 1K pullup resistor from the base of the transistor to +ve. Without this resistor the base will be floating because of the capacitor.
The 1.5K and 220 ohm resistors on the collector can be removed and replaced with one 1k resistor and the output taken from the junction of this resistor and the collector of the transistor.
I have quickly assembled this circuit on a test board and tried it by flashing a signal wire to ground. Using a 12volt supply I get a nice sharp 10volt pulse(observed on a scope) whenever I ground the wire.
Make sure your transistor is ok and that you have the terminals the right way round.

 

[paulc]

Hi, I've noticed that you are using the Arduino hall effect sensor module. That module outputs a +5V signal when a magnet is detected and so will not work with that circuit because it has a PNP transistor which requires a negative signal to the base for it to conduct.

No its not a KY-003, its just a raw hall sensor potted in a plastic case. When the magnet is near the signal wire goes to earth. When the magnet is away the signal wire goes open. It has no resistor or LED like the KY-003.

 

[Claudio Masetto]

You need a pullup resistor from the +ve to the signal wire of the hall sensor and the other mods in my post.
I will try to post a diagram later tonight.

 

[mayhugh1]

Paul,
Your schematic shows a .1uF cap in series with the base of your transistor, but I believe the cap you're actually using is .01uF. That cap will reduce the width of the trigger pulse by 10X. - Terry

 

[paulc]

Paul,
Your schematic shows a .1uF cap in series with the base of your transistor, but I believe the cap you're actually using is .01uF. That cap will reduce the width of the trigger pulse by 10X. - Terry

Hi, I just checked the cap it is correct, 0.1uF the paper work also reflects the same.
I'll try the pull up resistor across the +ve and the signal wire of the hall sensor and see if that helps.

 

[paulc]

Ok after looking at the specs. of the KY-003 I realise that it does work by grounding the signal terminal when it senses a magnet. I was thinking of something else.
You need to put a 1K pullup resistor from the +ve terminal to the output terminal of the sensor.
You also need a 1K pullup resistor from the base of the transistor to +ve. Without this resistor the base will be floating because of the capacitor.
The 1.5K and 220 ohm resistors on the collector can be removed and replaced with one 1k resistor and the output taken from the junction of this resistor and the collector of the transistor.
I have quickly assembled this circuit on a test board and tried it by flashing a signal wire to ground. Using a 12volt supply I get a nice sharp 10volt pulse(observed on a scope) whenever I ground the wire.
Make sure your transistor is ok and that you have the terminals the right way round.

I tried all the above in various stages all the way up to all suggestion being implemented.

I replaced the PNP BC558 transistor to be sure. I learned how to test it and it seems to do what its supposed to.
I added a 1k resistor across the +ve and switch wire then added an LED to the Hall sensor so its the same as a ky-003.
I added an LED from the collector of the transistor to ground so I can see it switching with the hall sensor.
I removed the 1k5 and 220R resistors for a single 1k resistor and connected the CDI trigger wire between the cathode leg and that resistor.

NOTE. When I hold the magnet at the hall sensor on the LED on the hall lights up. While the LED at the transistor gives a short flicker and then goes off even if I hold the magnet in place. I would think this is the correct operation.

The CDI continues to test fine when on the induction coil trigger. Just wont work when on the hall sensor or points.

Im happy to keep plugging away. But I am at a loss.

 

[Claudio Masetto]

You need to add a pullup resistor from the +ve to the signal terminal of the hall sensor. The resistor and led which you have put there is ok to show that the sensor is working but is not acting as a proper pullup. You can leave the resistor and led and add the pullup as I have shown in the diagram. You also need to add another pullup on the other side of the capacitor as shown on my diagram otherwise the base of the transistor is floating.
You must remove that led across the transistor. That will interfere with the proper operation of the transistor.
And last, all you need is a 1K resistor from the collector to the ground terminal to extract a pulse from.

 

[paulc]

Before the LED I did try the pull up resistor across the hall sensor +ve and signal out and made no difference. But I will try your schematic and remove the LED across the transistor too and do the other things as well.
thanks

 

[paulc]

I’d imagine the trigger voltage produced by the original pick coil would be quite high. Is it possible that the 12 volt circuit here is too low in voltage to trigger?
I speculate that perhaps Bluejets CDI maybe designed differently that mine, the Chinese will keep changing designs to make production more efficient.

 

[paulc]

You need to add a pullup resistor from the +ve to the signal terminal of the hall sensor. The resistor and led which you have put there is ok to show that the sensor is working but is not acting as a proper pull-up. You can leave the resistor and led and add the pullup as I have shown in the diagram. You also need to add another pullup on the other side of the capacitor as shown on my diagram otherwise the base of the transistor is floating.
You must remove that led across the transistor. That will interfere with the proper operation of the transistor.
And last, all you need is a 1K resistor from the collector to the ground terminal to extract a pulse from.

Claudio, See the attached pic, I followed your schematic . All the resistors shown are 1k. I temporarily used a grounded LED to test the cathode leg of the transistor and with each pass of the magnet over the hall sensor, sure enough the LED gave a "pulse". But still no spark!

I have two of these DC 4 pin CDI and both work fine on the original induction coil pick up.

As bluejets is also based in Australia I wish I could send my two CDI modules to him to see if they would work on his parts. If they do not then at least we know that not all CDI are the same and if they do work then I know its at my end!

 

[Claudio Masetto]

Hi Paulc, it is possible that the CDI module might require a high pulse to trigger it. I have only looked at the pulse on my ocilloscope as I do not have a CDI module and it produces a nice clean 10v pulse. Maybe this is not enough.
Do you know if it is possible to obtain a schematic of the internal circuit of the CDI module?
Whereabouts in Australia are you?

 

[paulc]

Hi Paulc, it is possible that the CDI module might require a high pulse to trigger it. I have only looked at the pulse on my ocilloscope as I do not have a CDI module and it produces a nice clean 10v pulse. Maybe this is not enough.
Do you know if it is possible to obtain a schematic of the internal circuit of the CDI module?
Whereabouts in Australia are you?

Claudio, unfortunately I dont have the inner workings of the CDI, it is of course ebay Chinese and god knows what they have inside them and Im sure like any budget manufacturer they will alter the design to suit their parts inventory and although the packaging looks the same and serves the same function, the subtle construction differences may be the reason why Bluejets can get his CDI to fire where as others can not. This is just my hypothesis.

 

[paulc]

Success! (In a round about sort of way).

So I was using the coil ripped out of an old 24 volt relay as an induction coil for the CDI trigger by passing a magnet passed the induction coil and inducing the desired spike (emf) which the SCR is looking for to trigger, and of course worked like a charm. Good sparks.

Yes I too suspected that at least for my CDI unit, I needed a high pulse voltage, more in line with what the induction coil would be emitting.

Then I thought, well what would happen if I used a simple hall circuit, via the cathode of the BC558 transistor to send a pulse to the external induction coil while the coil was still connected to ground and the trigger of the CDI? The coil would be energize and on release of the voltage spike. Well it worked! Then I tested the unit on the lathe at 2,000 rpm with a 9 slot interrupter.... 18,000 sparks per minute using the Hall sensor as a trigger. No problem and with a 5 mm plug gap!

I then replaced the rather bulky 24 volt induction relay coil with the coil from a very small 12 volt reed relay. It still worked and at 18,000 rpm.

Oh and using a timing strobe light I noticed the unit has a crude advance / retard. At below about 700 rpm the unit is fully retarded any rpm above 700 the unit jumps to about 25 degrees advance, there is no apparent curve , just advance on or advance off . I would suspect this is to aid safe kick starting on a bike or quad bike.

I can use this!!!

 

[paulc]

Here's a video of the set up and it running. (as of writing the video was still uploading)

 

[Foketry]

Hi Paulc, it is possible that the CDI module might require a high pulse to trigger it. I have only looked at the pulse on my ocilloscope as I do not have a CDI module and it produces a nice clean 10v pulse. Maybe this is not enough.
Do you know if it is possible to obtain a schematic of the internal circuit of the CDI module?
Whereabouts in Australia are you?

here the diagram of the CDI 4 wires, in the same file there are also 5 and 6 wires, I hope it is useful

 

[Claudio Masetto]

Hi Paulc, good to see that you have it working. I believe that Tornitore45 in post 6 of this thread has an explanation of the trigger pulse required for the CDI module. He says that it needs to see a negative pulse to arm the unit before the positive pulse. He says that you can use a simple differentiator for this. I have redrawn the schematic and you can try it if you like. Note that I have changed the position of the hall effect terminals on my diagram.

 

[Foketry]

Here's a video of the set up and it running. (as of writing the video was still uploading)

Excellent job !
Could you also indicate the values of the resistances, the type of Hall sensor, the connections with the CDI and with the relay coil ?
I would like to replicate your schema, could you write the values on the sheet I saw in the video, thanks

 

[paulc]

Hall sensor is the same as most any used here on the forums for all ignitions. 3144 but Ive used several others. The Hall signal wire just goes to ground when the magnet is near and goes open when is not. There is no pull-up resistor used on my Hall sensor. The LED you see in my video is just for indication. It can be deleted.

All the resistors shown are 1k
The transistor is BC558 (PNP)
The induction "booster coil" is just a 12 volt reed switch relay. Im only taking advantage of the the tiny coil from it. No need to chop it up as the whole relay package is tiny and secure. But you could use any coil from any 12-24 volt relay. I just used the reed style relay as it is tiny!
The induction boost coil is wired to ground and the other side of the same coil has two wires on the same pin, one to the CDI trigger and the other to the cathode of the BC558.

Every time the hall sensor is goes to ground, the BC558 energizes the boost coil, then when the hall sensor opens the booster coil collapses and triggers the SCR in the CDI.

Wiring to the CDI is unchanged. trigger (input from booster coil), ground, +12volt, HT coil out.