Finding Performance By Accident

[Pablo]

OK, gentlemen. I think there are a few misconceptions and misnomers regarding the HT ignition.

First, there is no CDI. CDI=C=capacitor, D=discharge, I=ignition. The little black box is a step up coil that is driven from the magneto. This is a either a marketing ploy or just from ignorance.

Are you sure? Have you unpotted one? Why is there a circuit board in there with a capacitor? Why, when checked with an ohmmeter do the low tension leads behave as if there is a diode in the circuit?

2nd, Wire: There are a few things that are going on that probably need to be taken into account. Copper and steel wire that is 6" long has less than a few milliohms of resistance. I think performance differences between them will be minute and probably more to do with HV leakage which is a function of the insulation of the whole system and the surface area of the wire in the insulation.

I agree.

This is a relatively low power system. Any moisture on the plug will burn off in a hurry when it heats up but moisture on the black box coil may cause losses. Dirt on the plug and coil will hold moisture and in severe instances will lead to sparks that convert cases and dirt to carbon which is conductive and there goes your spark energy to ground. Use some silicone electrical sealant on the end of the wires that go into the coil and the spark plug boot to seal out moisture. A cracked coil case would be suspicious too.

Spark power is high voltage, 20,000+ (60KV in cars) volts but very low current, microamps, so it is easy for it to dissipate into any moisture.

This is a great point. Keeping the system clean and even using some silicone grease when threading the wire in the CDI does make a difference.

So copper or steel wire will perform better than carbon wire. Better insulation will keep the spark current IN the wire. Clean plugs on the outside and clean coils help.

There seems to be a wide variation of oem wire with the HT's from your descriptions.

I disagree on the first line. There will be no performance difference with conductor type as long as the wire is well made.

 

[arceeguy]

OK, gentlemen. I think there are a few misconceptions and misnomers regarding the HT ignition.

First, there is no CDI. CDI=C=capacitor, D=discharge, I=ignition. The little black box is a step up coil that is driven from the magneto. This is a either a marketing ploy or just from ignorance.

A real CDI is much more sophisticated and was designed to discharge a capacitor that is charged to about 400 volts though the coil to step up the spark voltage. It is charged by a transistorized 12 to 400 volt inverter. It requires a 12 v supply. It also causes a longer lasting spark. It is also distinct from a transistor ignition which uses a transistor to switch the 12 volts into the coil.......

Does it really matter whether it is a "true" CDI unit, or just a transistorized ignition? You are also making a CDI more sophisticated than it really is. All you need for a magneto driven CDI is the coil, a diode, a capacitor, and an SCR with a few other components. No inverter is needed for a magneto driven system because the magneto coil provides the energy to charge the capacitor directly. As far as "marketing ploy" or "ignorance", most small engines have had pointless ignitions since the eighties, so it isn't anything groundbreaking. It is expected.

If you want marketing, we'll call them ITIMs.
I=Integrated (coil and electronics) T=Transistorized I=Ignition M=Module

 

[VTBikeman]

I must confess, I haven't dissected the ignition coil and it was pointed out to me there is a capacitor and a possibly a diode in there. However, IMHO, it is too small to be an actual CDI. I will do some snooping and get back to you all.
G

 

[VTBikeman]

NOW, I've dissected the thing and here is the poop:

The magneto (generator, alternator) has 2 poles. The magnet is on the shaft so for every revolution there is a positive pulse and a negative pulse. The diode selects the pulse as the piston nears tdc and blocks the pulse that is made as the piston nears bottom dead center. The voltage I measured is about 120 volt peak and probably goes up with RPM. I used a drill to spin the motor with the plug out to measure this.

The capacitor tunes the coil to resonate at a frequency that will give the coil's highest output.

That actually leaves the negative half of the wave unused and if you use a half wave rectifier you should be able to tap some power off that low voltage winding to do something like charge a few batteries without affecting the spark. I didn't measure the output of that winding yet but the published spec is 6 volts or so that could yield a DC peak voltage of about 8.4 volts, not enough to charge the batteries (12v) on my bike directly.I also do not know which polarity you would get. You may wind up with a negative supply (pos ground) which is no big deal unless you've grounded your electric system as neg ground. You could probably still pull 3 watts off that winding without harm as you are using it 1/2 time. Feed a small dc:dc inverter and there's your battery charger.

There is no scr because there is no separate trigger for it. The spark is generated solely by the magneto.-simple and elegant. Could a CDI be built for it with the oem alternator? Probably but you would need a separate detector to trigger the spark and one end of the magneto is grounded so the choice becomes a 1/2 wave rectifier or a voltage doubler to charge the big C. Once you have the cap charged up then the motor would start.

There are some CDI schematics on line that use a microprocessor to advance the spark at higher rpm's but from the curve the advance doesn't kick in until 5K rpm's and now your up to the upper limit of the engine anyway. Has anyone studied spark advance for that motor? Maybe it would help at a lower rpm's.

This leaves me thinking that you could add a coil with a 2 pole core at 90 deg from the oem unit and generate some power off the magnets as they turn in the dead zone of the oem alt.

I will post the schematic soon on a new thread.

 

[Pablo]

VTBikeman - Very cool! Thanks.

 

[arceeguy]

NOW, I've dissected the thing and here is the poop:

The magneto (generator, alternator) has 2 poles. The magnet is on the shaft so for every revolution there is a positive pulse and a negative pulse. The diode selects the pulse as the piston nears tdc and blocks the pulse that is made as the piston nears bottom dead center. The voltage I measured is about 120 volt peak and probably goes up with RPM. I used a drill to spin the motor with the plug out to measure this.

The capacitor tunes the coil to resonate at a frequency that will give the coil's highest output.

That actually leaves the negative half of the wave unused and if you use a half wave rectifier you should be able to tap some power off that low voltage winding to do something like charge a few batteries without affecting the spark. I didn't measure the output of that winding yet but the published spec is 6 volts or so that could yield a DC peak voltage of about 8.4 volts, not enough to charge the batteries (12v) on my bike directly.I also do not know which polarity you would get. You may wind up with a negative supply (pos ground) which is no big deal unless you've grounded your electric system as neg ground. You could probably still pull 3 watts off that winding without harm as you are using it 1/2 time. Feed a small dc:dc inverter and there's your battery charger.

There is no scr because there is no separate trigger for it. The spark is generated solely by the magneto.-simple and elegant. Could a CDI be built for it with the oem alternator? Probably but you would need a separate detector to trigger the spark and one end of the magneto is grounded so the choice becomes a 1/2 wave rectifier or a voltage doubler to charge the big C. Once you have the cap charged up then the motor would start.

There are some CDI schematics on line that use a microprocessor to advance the spark at higher rpm's but from the curve the advance doesn't kick in until 5K rpm's and now your up to the upper limit of the engine anyway. Has anyone studied spark advance for that motor? Maybe it would help at a lower rpm's.

This leaves me thinking that you could add a coil with a 2 pole core at 90 deg from the oem unit and generate some power off the magnets as they turn in the dead zone of the oem alt.

I will post the schematic soon on a new thread.

So theres nothing in the ignition module than a diode, cap and coil? Talk about the simplest ignition system. I figured that there had to be some sort of transistor or SCR switching the output of the magneto to the primary of the ignition coil even if it wasn't a "true CDI". As far as the lack of a separate trigger goes, I theorized that the SCR or transistor was triggering off of the falling edge of the AC waveform generated by the magneto.

My "pet project" is to rewind the magneto to produce 14V, and use the magneto to keep a small 12V lead acid battery charged. I will dump the two wire ignition module in favor of a 12V CDI unit from a pit bike. I have a local generator shop that is willing to rewind the magneto, and he said that he thinks I should be able to get a couple of amps out of it. The pit bike CDI's have ignition advance (actually you physically advance the trigger magnet, and the CDI retards the ignition less and less as RPM's rise) so performance should see a boost, and temperatures may fall.

 

[VTBikeman]

My "pet project" is to rewind the magneto to produce 14V, and use the magneto to keep a small 12V lead acid battery charged. I will dump the two wire ignition module in favor of a 12V CDI unit from a pit bike. I have a local generator shop that is willing to rewind the magneto, and he said that he thinks I should be able to get a couple of amps out of it. The pit bike CDI's have ignition advance (actually you physically advance the trigger magnet, and the CDI retards the ignition less and less as RPM's rise) so performance should see a boost, and temperatures may fall.

Anyone know how fast that motor can run and still stay together?
What will you do for a sensor that is accurate? Photodiode or magnetic pickup?
You may want to just buy an extra armature just in case.
I'm guessing you might get 20 watts out of the generator. It's not much more iron than a small fan or old phonograph motor so that's my point of reference. I could be wrong. My lighting system draws about 12 watts at the highest draw. I don't know how much power your cdi will need.

You could use the current windings to generate HV for the cdi and the LV winding for the logic and skip the battery.

The question is how much potential is in Happy Time engine? 5 HP, 10? Are we better off doing exhaust tuning and gearing?

Inquiring mind wants to know.

 

[arceeguy]

I plan on using a magnet and pickup coil from a "honda style" pit bike/scooter engine, since the CDI unit is meant for that application. There are two types of CDI units available, the "5 pin" type with a 5 pin connector and the "4 pin" type. The 5 pin type is designed to run off of AC from the magneto coil directly, and the 4 pin type is designed to run off of 12V DC. The 4 pin type has an internal DC-DC converter to charge the capacitor. I haven't measured the current draw, but I doubt it draws much current. I've left the ignition switch "on" on my motorcycle and the battery wasn't depleted the next day. The starter/generator guy who will rewind the stator said that he has rewound quite a few small engine/motorcycle stators with lighting circuits and he knows what thickness of wire and how many turns are needed. Even if I only get 20 watts, and the CDI draws 10, that leaves me with 10 watts at 12 volts for lighting. I'm not sure, but some of these CDI units have a 8500 RPM governor - but I'm pretty sure that you don't want to rev a HT engine that high for any extended periods. It seems to me that the HT engine was not designed for high RPM or high performance, but ignition advance will certainly help get the most from it. I have a spare stator coil that will be the "sacrificial lamb" and this will be my winter project. I am having too much fun whizzing around town to take it out of service.

 

[VTBikeman]

Please keep me posted. Remember that the mag pickup needs a small magnetic point to pick the timing pulse from. The generator magnet is too broad to be accurate. A wheel with a slot in it and a led-photo xstor pick up would be the most accurate.

The CDI only draws current to charge the cap to full voltage. When charged there is no significant draw after that. The oscillator in the inverter may even shut off if it is sophisticated enough.

I re-gapped my plug to .035 and that made it run much better. The wire in my kit was copper so no problem there.

 

[VTBikeman]

I said: So copper or steel wire will perform better than carbon wire.

Pablo said: I disagree on the first line. There will be no performance difference with conductor type as long as the wire is well made.

I say: Carbon wire is a long resistor which converts current to heat. There is a loss which is the trade-off for reduced spark plug noise in your AM radio. The resistance damps the coil ringing (oscillation) set up by the spark, shortening the impulse time and the intensity that your spark plug wire radiates the noise.

You can also use shielded spark wire and ground the shield to the frame. In fact you would want to shield the spark coil in a metal box if you like getting the ball game when you ride. This common in military vehicles and high performance engines.