Model engine CDI easy and cheap

Home Model Engine Machinist Forum

Help Support Home Model Engine Machinist Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Not to blow my own horn but we published an excellent ignition system in Issue # 34 of Model Engine Builder. David Sage and John Gedde are the designers. It will handle a V-8 up to around 10,000 RPM. Within the Bay Area Engine Modeler's Club, there are probably 30 of those systems operating with no failures (except for the one builder who bound the Hall Effect Sensor and secondary ignition wires together). The circuit turns off after a short time of non operation so the battery doesn't discharge and the coil won't burn out. Most of us use the Ford Coil Over Plug coils because of their reliability and low price (about $5 each in lots of 8). This is not a CD ignition system, it has no spark advance, it is just a very well thought-out circuit. The entire circuit is shown mounted on a battery box for an engine box. The white connectors are temporary as I work out the details of a completely 3D printed distributor for 2-8 cylinder engines.View attachment 110381
I just looked at the Model Engine Builder article and it looks interesting and even possible for us non electronic klutz's. Have you made your own PCB or have you had them made? If you have had someone make a quantity of these boards would you have any available for purchase or would others be interested in purchasing a lot together? I have tried using the pre-drilled universal boards to make TIM6 circuits and they end up looking like a spaghetti factory explosion but they do work.
 
I just looked at the Model Engine Builder article and it looks interesting and even possible for us non electronic klutz's. Have you made your own PCB or have you had them made? If you have had someone make a quantity of these boards would you have any available for purchase or would others be interested in purchasing a lot together? I have tried using the pre-drilled universal boards to make TIM6 circuits and they end up looking like a spaghetti factory explosion but they do work.
I have e-mailed Dave Sage, the board designer to see if he will either order more boards or release the board design files. The boards were constructed in China and are quite nice. I just built the last 3 of the 11 boards I was given by Dave or I could offer a few. Give it a few days to see what, if anything, he will do.
I had considered CNCing some boards and still might do that if all else fails.
 
Hi Foketry,
Didn't get in to the 6 pin versions but I did supply full details of many of the different system that were available including my 4 pin version.
So I have really no idea of what you might have done to try to get a 6 pin version operational.
I'll include the file on the different types here once again as trying to find these things sometimes can take a lot of trowling though posts.

Hi Bluejets
I think the trigger pick up is the same for CDI 4 wires and CDI 6 wires, I probably made some mistakes during the assembly of the Hall sensor circuit which you kindly posted.
Is it possible to check the function of the Hall sensor circuit with the multimeter or or an ascilloscope is required ?
Example :when the Hall sensor is excited by the magnet, what kind of signal should I read in the To CDI Trigger input point ?
Thank you
 
I have e-mailed Dave Sage, the board designer to see if he will either order more boards or release the board design files. The boards were constructed in China and are quite nice. I just built the last 3 of the 11 boards I was given by Dave or I could offer a few. Give it a few days to see what, if anything, he will do.
I had considered CNCing some boards and still might do that if all else fails.
Thanks. That would be great. I tried making my own TIM6 boards and I was sort of successful but I had to do some modifications like jumpers and scaring off connections which should not be connected. I am not sure what would happen with a double sided board.
 
Here is a video of the boards I made up they were simple to machine up I may just print out the stencils try and fine and alternative way to etch them.



 
Thanks. That would be great. I tried making my own TIM6 boards and I was sort of successful but I had to do some modifications like jumpers and scaring off connections which should not be connected. I am not sure what would happen with a double sided board.
I have the board files but I have never had a set of boards made so that may take a while as I learn what and how to do it. If you need a single ignition system, I do have a spare built system. I would sell it if you are interested. Let me know at editor at modelenginebuilder.com.
 
Hello Foketry.
The output of Hall sensor is in a magnet proximity High or Low /depends on sensor type /
so you can check output by common multimeter or LED diode.
Be careful because hall sensors usually have open collector so always connect miniature resistor
2 - 5 kohm to battery Plus pole and to sensor Output.
To check hall sensor just connect multimeter to battery Minus pole and to the sensor Output
you should read on multimeter fast jump from 0V to 5V or vice versa near magnet.
 
Hi Kadora
thanks for the reply, I already have a simple Hall sensor tester with resistance and LEDs, but my problem is different
The CDI scooter receives the impulse from a special pick up trigger, the circuit I made with Hall sensor should replace this pick up, but it doesn't work.
My question is, what kind of signal should the circuit emit to replace the pick up?
How can I check this signal with a multimeter?
In attachment the circuit kindly posted by Bluejets and the pick up trigger
 

Attachments

  • Hall EffectMod per CDI.pdf
    163.3 KB · Views: 585
  • 9951022E-91DE-4654-B4F8-2D92278BDAC7.jpeg
    9951022E-91DE-4654-B4F8-2D92278BDAC7.jpeg
    20.1 KB · Views: 385
Last edited:
Scooter pick up circuit is PROBABLY simple coil which generates minus or negative sinus waves
and this waves trigger CDI. On the other hand Hall sensor can generate only positive impulses.
Signal generated by pick up item can be checked by osciloscope. /Best way/
I can not open your attached file right now because I have brand new computer not having
Adobe reader installed yet.
You need to trigger Scooter cdi with hall sensor am I right?
 
Scooter pick up circuit is PROBABLY simple coil which generates minus or negative sinus waves
and this waves trigger CDI. On the other hand Hall sensor can generate only positive impulses.
Signal generated by pick up item can be checked by osciloscope. /Best way/
I can not open your attached file right now because I have brand new computer not having
Adobe reader installed yet.
You need to trigger Scooter cdi with hall sensor am I right?

Yes, I would like to trigger a CDI Scooter with Hall sensor
The attached file is to do this
 
Now I am able to open pdf files.
Maybe you have low trigger voltage on the input of cdi.
Try to change resistors in this way ---Solder 1k5 resistor instead 220R and 220R instead 1k5
This should rise trigger impulse voltage .
 
Foketry, The Hall-Effect is not the problem here, at least not yet. You need to answer some questions here first before sorting this out.
Start by reading my post about Hall-Effect ignitions https://www.homemodelenginemachinis...ll-engine-ignition-science.31347/#post-325051 then:

1. Is the rotor on the crank swinging a magnet or a steel ring gear with notches or tabs?
2. Does the original pick use a magnet with the Hall-Effect?
2. What is the make of the ignition box?

The first question is the most important to be answered otherwise your just going down a dead end alley.

Ray
 
Foketry, The Hall-Effect is not the problem here, at least not yet. You need to answer some questions here first before sorting this out.
Start by reading my post about Hall-Effect ignitions https://www.homemodelenginemachinis...ll-engine-ignition-science.31347/#post-325051 then:

1. Is the rotor on the crank swinging a magnet or a steel ring gear with notches or tabs?
2. Does the original pick use a magnet with the Hall-Effect?
2. What is the make of the ignition box?

The first question is the most important to be answered otherwise your just going down a dead end alley.

Ray

The pickup reads a steel tab that rotates at a predefined distance , you can see the drawing posted by Bluejets , Cam disc
The original pickup is composed of a coil and a magnet
I do not know the brand of the CDI ignition, generally they are produced in China, mine is DC , 6 wire,
attached file , page 11
my problem is this, i would like to replace the original pickup with a Hall sensor to use the CDI on the engine models
thanks for your interest
 

Attachments

  • Cam Disc Inductive Pickup.pdf
    196.1 KB · Views: 419
  • Scooter CDI units.pdf
    1.3 MB · Views: 470
Last edited:
Now I am able to open pdf files.
Maybe you have low trigger voltage on the input of cdi.
Try to change resistors in this way ---Solder 1k5 resistor instead 220R and 220R instead 1k5
This should rise trigger impulse voltage .

I replaced the 2 resistors as suggested, but it does not work, the led of the hall sensor lights up, but the Cdi does not receive the signal, only with the original pickup everything works
 
The pickup reads a steel tab that rotates at a predefined distance , you can see the drawing posted by Bluejets , Cam disc
The original pickup is composed of a coil and a magnet
I do not know the brand of the CDI ignition, generally they are produced in China, mine is DC , 6 wire,
attached file , page 11
my problem is this, i would like to replace the original pickup with a Hall sensor to use the CDI on the engine models
thanks for your interest

I thought this was the setup, I'm current designing a Krober magneto replacement ignition that works similarly.

Anyway, do you have the part number for the Hall-Effect you are using?

This is important because we need to know the polarity of the Hall-Effect. You will need to place (glue) a small weak magnet behind the Hall-Effect for the Hall-Effect to sense the metal tabs going by. We need to know if we have to glue the North or South pole to the Hall-Effect. Alternately you can buy a Hall-Effect proximity sensor however glueing a small magnet to the Hall-Effect is smaller, cheaper and easier to mount.

Ray
 
Foketry you have probably cdi unit which needs for triggering energy stored in capacitor
so too complicated to trigger this cdi box with hall sensor.
Voltage on trigger capacitor is 70V
Do not waste time and buy hall sensor cdi box used for rc model engines.
 
I thought this was the setup, I'm current designing a Krober magneto replacement ignition that works similarly.

Anyway, do you have the part number for the Hall-Effect you are using?

This is important because we need to know the polarity of the Hall-Effect. You will need to place (glue) a small weak magnet behind the Hall-Effect for the Hall-Effect to sense the metal tabs going by. We need to know if we have to glue the North or South pole to the Hall-Effect. Alternately you can buy a Hall-Effect proximity sensor however glueing a small magnet to the Hall-Effect is smaller, cheaper and easier to mount.

Ray

I use Hall sensor 3144 installed on this circuit , look at the attachment
KY-003 has sensor 3144
 

Attachments

  • Hall EffectMod.pdf
    163.3 KB · Views: 477
Foketry: To begin with Hall-Effects aren't as simple as most people think. They come in different volt & amp ratings, polarities and non-polarity, latch & non-latching, transistor or FET, open collector/drain or open emitter/source, and sensitivity. So as you can see there is a lot of different varieties. You say your KY-003 uses a 3141 Hall-Effect, which is now obsolete and is no longer made and I was not able to get in-depth details on it. However Allegro says it's replacement is the A1104, there is a big difference between the 2 with the biggest being the 3141 is a transistor and the 1104 being a FET. The 3141 has what is known as a open 'collector' and the 1104 having a open 'drain'. In both cases the emitter/source is internally grounded to the GND pin. So when the Hall-Effect turns on it grounds out the signal pin. Now from the info on the KY-003 you posted the circuit shown will not work and if it ever did then it was by fluke. The reason is because the capacitor that goes from the signal to the base of the transistor are both basically at ground level voltage and the cap may only charge up through junction leakage. I simulated the circuit in the PDF and got almost no signal out to the CDI, less than 0.01 volt. I redesigned the circuit so that it outputs positive pulses 11.5 volts (see attachments).

You will need to place a small weak magnet behind the Hall-Effect, the front face is the one with the brand information. You will need to place the south pole against the Hall-Effect. You will need either a compass or a pole detector or do what I do sometimes and use some clear thin tape to hold the magnet against the Hall-Effect. Power up the circuit and then pass a piece of iron pass the front face of the Hall-Effect and see if you get a signal. You may have to adjust the distance between the iron and Hall-Effect to get a signal. If you still don't get one then turn the magnet over and try again. Once you get a signal mark the magnet so you know which way to epoxy the magnet onto the Hall-Effect. Note: use an LED with a 20-40 ma rating.

Ray
 

Attachments

  • Hall_effect Ignition from PDF.pdf
    39 KB · Views: 582
  • Hall_effect Ignition Proper.pdf
    29.1 KB · Views: 528
Foketry: To begin with Hall-Effects aren't as simple as most people think. They come in different volt & amp ratings, polarities and non-polarity, latch & non-latching, transistor or FET, open collector/drain or open emitter/source, and sensitivity. So as you can see there is a lot of different varieties. You say your KY-003 uses a 3141 Hall-Effect, which is now obsolete and is no longer made and I was not able to get in-depth details on it. However Allegro says it's replacement is the A1104, there is a big difference between the 2 with the biggest being the 3141 is a transistor and the 1104 being a FET. The 3141 has what is known as a open 'collector' and the 1104 having a open 'drain'. In both cases the emitter/source is internally grounded to the GND pin. So when the Hall-Effect turns on it grounds out the signal pin. Now from the info on the KY-003 you posted the circuit shown will not work and if it ever did then it was by fluke. The reason is because the capacitor that goes from the signal to the base of the transistor are both basically at ground level voltage and the cap may only charge up through junction leakage. I simulated the circuit in the PDF and got almost no signal out to the CDI, less than 0.01 volt. I redesigned the circuit so that it outputs positive pulses 11.5 volts (see attachments).

You will need to place a small weak magnet behind the Hall-Effect, the front face is the one with the brand information. You will need to place the south pole against the Hall-Effect. You will need either a compass or a pole detector or do what I do sometimes and use some clear thin tape to hold the magnet against the Hall-Effect. Power up the circuit and then pass a piece of iron pass the front face of the Hall-Effect and see if you get a signal. You may have to adjust the distance between the iron and Hall-Effect to get a signal. If you still don't get one then turn the magnet over and try again. Once you get a signal mark the magnet so you know which way to epoxy the magnet onto the Hall-Effect. Note: use an LED with a 20-40 ma rating.

Ray

74Sprint , thanks for your help, in the next day I will try to make this circuit
Very interesting the application of a magnet glued to the Hall sensor, I will also try this
 
Video for those who seem to think it does not work for one reason or another.
Pickup is attached to a motor driving a metal disc with 3 vanes. Magnet (3mm dia) is on opposite side of the vanes.
I have tried with a magnet attached to the rear side of the hall effect and it also works fine in detecting the vanes. Same thing, different tune as they say.
Tacho reading shown is per pulse and tacho tops out at 9,900rpm or pulses in this instance.
After that it starts to show alpha characters.
Initial pulses as can be seen in the video are at around 9,500 mark and later in the video the test bench motor is run up to maximum at over 20,000 pulses.
Previous remote tacho readings have confirmed this.
As can be seen, the Ebay unit works just fine.

https://photos.app.goo.gl/5r4hDb6SNXBTpCeBA
 
Last edited:
Back
Top