jmj

As some of you aleready know, i have a CDI project on my [link=http://home.online.no/~jon-mj/cdi_ignition.htm]homepage[/link].
Tonight i have made a big update to the pages, so please have a look at it.

Some info of my CDI:
-Runs from a normal 4.8V RX battery
-Programmable advance curve
-Built from easy to find parts
-Integrated strobe light, using a hi intensity LED
-Integrated test function - no need for a signal generator
-The advance control can be separated from the PCB and be used to upgrade other CDI's without advance control
-Low part count
-Can use a homemade coil

The CDI is tested and works, but there is still an issue with noise from the high voltage board reaching the advance board.
This should not be too hard to fix.

This project is totally open source, all the info you need can be downloaded from my pages.
I get a lot of email about this project, so it may be better to use this forum to share problems and ideas.

My homepage is at: [link=http://home.online.no/~jon-mj/cdi_ignition.htm]http://home.online.no/~jon-mj/cdi_ignition.htm[/link]

Regards,
Jon-Magne Johansen

blueberry

Looks like a great project. I am very interested in the outcome.
I don't know much about this stuff but could you use something to filter the noise from the high voltage board?

Pete

jmj

Yes, the next thing to do is to make a filter of some kind.
I have already made some tests, but i will try to keep it simple, so i will have to experiment some more..

Jon-Magne

blueberry

Would any of the products from this company work for filtering?
http://www.wirelessdesignonline.com/...rproducts.html

Pete

jmj

Yes, probably. But i do not want to use 'exotic' components. I will need to make a filter by using diodes/capacitors/resistors.
The whole idea of this project is to keep it simple and low cost.

Jon-Magne

blueberry

I understand, Just wondered as this company is just up the road from me.
I checked out your site, that is some spark on your video!

Pete

diablo_r

couple of questions JMJ, I made the circuit last night but it does not work quite right. On the far right of your schematic there is a resistor and capcitor (C6 and R1) now it says on the diagram to use 4.7uF, but in the parts list it says 4.7nF. I'm just wondering if this is my problem. It makes enormous sparks, but the circuit free runs at about 1Hz, it sits there happily sparking to itself but I'm getting over 600v from the diode bridge[:-]

PS, your toroid pics, is that really 0.1 wire for the 450 turns, it looks a bit thicker? I gave up on 0.1 as my fingers are too big to wind it so used 0.2.


Cheers,
Rich

jmj

C6 is a electrolytic capasitor of 4.7uF/16V. Obviously an error in the parts list, sorry.
If you get 600V, at least the high voltage board is OK. This is a new record, the most i have got is 350V.

If it sparks by itself, it sounds like you have activated the test function, or have a problem with the PIC.
Try removing the wire bridge between the 2 parts of the board, and see if it stops.

The wire in the pictures is thicker than 0.1mm, it is not a picture of my torroid.
I used 0.1mm, but it is not important.

Jon-Magne

diablo_r

Thanks for the reply! I think this might me my problem, I said 600v but thats all my volt meter goes up to[:-] you cans ee the voltage rise as it charges, it hits the 600v limit and JUST afterwards it self discharges, what I think is happeneing is the DC volts are so high it jumps across the thyristor internally, that dicharges the high volt capacitor and it makes the spark at the plug. I did not have the PIC board connected so there was no prob there. I'll change the 4.7nF for a 4.7uF tonight and try that. I was running off a 6v battery.

My home made coil did not work either, you can hear it sparking internally. I'm going to wind another with an insulation layer between each run of secondary.

Cheers,
Rich

diablo_r

sorry, more questions.

In the early schematic there was no resistor in parallel with the high voltage cap, but R11 (10meg) is there now. Also R9 has been added (10k). Could you just let me know the function. I'm guessing R11 is to discharge the cap when the ignition is not turned on, and R9 is to limit the switching current through T1, but maybe I'm wrong!

Cheers,
Rich

Ben_C

I built a high voltage board which worked OK - though my voltmeter recorded between 500V-600V from 6V input. I had a few odd behaviours with the SCR triggering the spark: the spark would trigger, but only start to recharge when I earthed the trigger.

Also, I only triggered a spark when I used a 1.2V power source as the trigger, I tried connecting directly to a Hall sensor but it didn't work. I don't understand the requirements of the Hall trigger circuit to trigger the CDI system. Any clues?

My high voltage coil is wound on an E-core. I would rather spend $200 on a professional CDI than wind another toroid coil! The E-core coil only took about 20 minutes to wind.

Thanks for the informative page, Jon.

Cheers,

Ben

diablo_r

HAHA, me too! have you any pics of your wound E core, I've got loads of E cores and as you say they are very easy to wind.

Cheers,
Rich

jmj

R9 and R11 is not used, and should be left out.
The 7805 and its associated components is not needed, either.
These components is some leftovers from the previous design, and is not on the new PCB.
A new schematic will be uploaded soon, where these components are removed.

Jon-Magne

jmj

Reduce the value of R6 from 1K to 240ohm, and it should be OK.
The SCR need to be triggered quite 'hard' when it is switching high voltages.
This has a bad side effect - a hall device can not source enough current to trigger it directly.
It also place a big load on the output on the PIC, so maybe a transistor buffer will be a good idea.

Jon-Magne

diablo_r

Hi Jon, another question. Changing the capacitor on the LV side fixed the voltage problems, it runs fine now at 380v DC. BUT.... I cannot get the thyristor to trigger the circuit. It works fine if you manually touch a wire to short the anode and cathode of the thyristor, but no matter waht volts I put to the gate I could not get it working. Could you tell me what voltage and current you give the thyristor gate please? I see fr the diagram you have a 470R resistor (R10) and C2 in the line from the PIC to the thyristor. I'm guessing the PIC puts out 5v and the resistor and capacitor are there to limit the volts and current passed to the thyristor gate?

What I'm trying to do it get the CDi part with fixed timing working for this weekend so I can go flying before I borrow a PIC programmer to do the PIC. I want to run the CDI circuit with a hall effect pickup as I already have a kettering circuit running on the engine off a hall sensor.


Thanks again in advance,
Rich

Ben_C

I can't find any info that I can understand on a buffer for the trigger (but then I'm not an electronics guru). Are you referring to using the hall effect sensor to trigger an SCR which then triggers the spark SCR?

Thanks,

Ben

aalbea

To diablo: Probably, if you runned the circuit with too much voltage (more than 400 V), you burnt the SCR. It was quite usual for me during thqe development tests.

Try changing the SCR.

Another important detail is that if you're using a logic signal Hall device, the output will be reversed (e. g. you have output until the magnet goes to the sensor), so the power stage needs to be adapted to work correctly.

Best regards.

diablo_r

Thanks Aablea,

I changed the SCR once I had it running at 380v so it should have been okay. My hall pickup is open circuit when no magnet is near, 0v when the magnet is nearby. I know I'll need an inverter circuit but my question is what volts/amps should you be supplying to the SCR gate to make it switch? Is it really 0.8 volts and 0.2amp max?

Thanks in advance,
Rich

aalbea

I use 5V and the current is limited to about 25 mA with a resistor of 180 ohm and the SCR triggers really good. The voltage at the SCR imput falls to 2Vpp max. Probably it will run OK with less current.

A good and easy way to invert the pulse of a hall device is by using an optocoupler. 4N25 device can do the work, but we're still working on.

Best Regards

Dr Evil

Will this overall reduce weight and increase performance? Please forgive my ignorance of this subject.

diablo_r

well I installed the old SCR that was spark jumping at 600v DC, and used a potentiometer to drop the +5v and adjusted it until it triggered. Seemed to work okay, I built a 30Hz input circuit to emulate the hall sensor and all seems okay, but a bit of inconsistent sparking at 30Hz, I think the pot needs a bit more adjusting, or maybe the SCR is damaged too. I should have some new ones tmorrow so I'll change it anyway. the sparks are impressive!

I just used a transistor to switch the hall froma 0v to a +5v signal, then the pot, then to the gate of the SCR. I put an LED in too to use for static ign timing.

Dr Evil, compared to magneto ign it will cut the weight by about 12- 14 oz I guess, and it should run better as you have fully varaible timing instead of the almost fixed timing of magneto ign. It should also give better starting as you get a spark at any RPM, magnetos usually stuggle below 300rpm which makes hand cranking difficult on some (but not all) magneto motors.

Cheers,
Rich

PS Jon, fixed the xls file, its written for French excel, the dec to hex converter is a different name in the english version[:'(]

Dr Evil

Thanks, Rich Very interesting indeed. I will be attempting htis inthe next year, I can see it now.

skazoo

Diablo, without magneto the engine could spin a little more faster (higher rpm with the same prop). I'm very interested what kind of rpm rise it will be.

jmj

Can you upload a drawing of how this is connected? That should help Ben_C out, se earlier post.

Also, i know the dechex macro is a pain, it is different in my norwegian version, too.
Maybe i will rewrite the sheet, so it can do without the macro.

Please also note the updates to my web, i got the new schematics from my co-developer, Jorge (aalbea), so i have put it all on the web. There is also links to the datasheets of all components used.

Regards,
Jon-Magne

diablo_r

I made my hall to SCR circuit off the top of my head in the workshop last night, I'll do a diagram and post it up over the weekend once I've worked out how to use my circuit layout program! Its basically similar to the TIM4 circuit with a 1k pot on the output and using only 1 transistor.

Also most of the transistors and the SCR I've used are different to your circuit Jon due to what components are easily available here in the UK. I just tried to match spec as close as possible when choosing the alternatives.

Thanks for the update, but one question for you or Jorge. I like the simple idea of using the magneto coil for the spark, I have a box of those at home! From the pics it seems you have 3 wires coming out of the hand wound coil. I only expected 2. Also did you literally just wind 50 turns of 0.5 onto the core that comes out of the magneto centre? its tricky too see in the pictures, if you have another picture or two looking end on or from the other side it may be more clear.

Thanks in advance,
Rich