The Blue,the Black and the Ugly (White)

Discussion in '2-Stroke Engines' started by duivendyk, Sep 2, 2008.

  1. duivendyk

    duivendyk Guest

    There is still rampant confusion about the ignition system of HT engines.This tutorial might be useful reading for folks trying to get their bearings.
    Inside the engines is a flywheel generator that produces a ac (positive& negative going output) pulse,from the generator coil at the appropriate time.
    Three leads are connected to this coil,one end of the winding on it is connected to the Blue lead,this is the output lead that needs to be hooked up to the blue input of the CDI unit (about this later).The Black lead is the other end of the coil,it is what EE's call the "common" or ground,although strictly speaking it does not have to be connected to the engine or bike frame.Then there is the infamous White lead,this is tap on the coil fairly close to the bottom end (the black wire).This produces a much smaller ac pulse and can be used to power a light (if you're careful),the output depends on the engine speed,it's better to leave it alone though,it's a potential source of trouble.
    The CDI unit ,it receives the ac pulse from the generator(blue&black wire) and produces a high voltage pulse to the plug to fire it.It appears to be underdesigned and frequently fails (get some backups).High voltages and moisture don't get along at all.Repackage it in a box and seal it with RTV (only when it is DRY!!.)
    The kill switch is used to kill the input to the CDI unit,this can be done either by shorting the input,that is hooking up the switch between the blue and the black leads or by opening up the input to the CDI, that is hooking up the switch between the blue lead from the engine and the one from the CDI input.You could do both if you want a "secret" disable capability.Some people have expressed concern that they could "fry" the coil with the short,those concerns are misplaced,there is around 300 ohms in the winding and the amount of heating of it by the short circuit current,for the short duration until the engine stops is quite small.
    The coil output at the blue wire is a healthy pulse, a few hundred volts !,under the right conditions it could kill you,treat it with respect.This voltage is at the kill switch no matter how you hook things up.The best way would be to have the switch close to the CDI unit,the CDI box would be a good place for it.The kill switch should be a high quality sealed switch.I am not a fan of using the frame for ground returns,it allways entails multiple questionable frame ground connections for the black wire,at the switch at the CDI unit or at the engine.I would prefer a single frame ground at the CDI or the engine and run the black wire to the switch.If you want to use the series switch configuration the blue connection has to be run from the engine to the switch and then back from the switch to the CDI.In this arrangement the blue leads are carrying pulse current during normal operation.It is advisable to minimize the EMI due to magnetic fields generated by these currents by using a twisted pair.This is why I'm not all that thrilled by this scheme and prefer the shorting method.
    Trouble shooting,invest in a Multimeter, you can get one for around $10,it's money well spent.
    1)Engine has quit, has run previously,reason to suspect ignition.Disconnect blue wire at engine,measure resistance to black wire, should be around 350 Ohms, if open circuit coil bad.If open circuit, operate kill switch (ONLY if shunt mode is used) .If open circuit disappears,coil bad between white and black wire (opened up).This can happen if light output was shorted or overloaded and a portion of the coil has been fried.May still be able to run engine but function of kill switch is reversed.
    2) Coil appears OK,disconnect kill switch at engine or CDI (shunt mode).Test engine,no go, either CDI or fuel supply.If switch in series mode get rid of blue connections to switch and connect blue wire from engine directly to CDI & test engine
    3) Using the white wire,it had best be left alone&taped up,if you feel compelled to use it take at least the precaution of wiring a 1 Amp fuse into the connection,that keeps you from frying the generator coil if you had a short,if you had one you might nor be able to get started but once running it would not necessarily cause the engine to quit,at least not until you had managed to fry the coil
    4) using the white wire to charge a battery,it should be no problem charging a 6 V leadacid battery,the trick will be to keep it from overcharging.If I had more info about the coil resistance beween the white&black wires and had measured open circuit output (ac) voltages at idle&at speed it would not be all that hard for me to come up with a simple circuit that could be built from readily available parts, to take care of the battery charging.
    5) Testing sparkplugs.I have stated before that I think the CDI unit is a marginal proposition especially the output pulse transformer.When testing plugs keep the body of the plug GROUNDED !,to prevent zapping it. Good luck.
     
    Last edited by a moderator: Sep 2, 2008

  2. Drunkskunk

    Drunkskunk Member

    Whats the output voltage of the white wire?
    My Degree is in electrical engineering, and I was thinking it might be possible to regulate the output of the white wire to a battery. a Lead acid battery is fairly tolerant of over-voltage as long as it doesn't have to stay that way.
     
  3. BSA

    BSA Guest

    I would guess that the problem is with the white wire is that it's rated at 6v/3w. What people don't realise is that you have to use a 3 watt bulb. If you use less than 3 watts the voltage will be higher. If you use more than 3 watts it will kill the engine.Voltage is proportional to load.
    BTW if thats nonsense someone please correct me, as i don't do much electriks.
    fortunately it is fairly easy to come by a 6v/3w dynamo lighting set here in england. Thats why the white wire puts out 6v/3w, so that it could be used instead of a dynamo on the old fashioned British style bikes they use in China.

    BSA
     
  4. duivendyk

    duivendyk Guest

    You need to do a number of things:
    1) Keep the average charge current within limits,say 0.25 Amp
    2) Start tapering off the charge current when battery voltage gets above 7.0 V.
    3) Stop charging altogether at around 7.25 Volts
    If you can pull this off with a fairly simple circuit,that anyone can build we're in business.I think there's a good chance I can do this,but as I said I need some info.As an afluent old geezer,I wouldn't go near a HT engine except for kicks maybe.Could be bad for the cardiovascular system but good for the brain.Can any one lurking out there in HT land with a VOM take some readings?
     
  5. arceeguy

    arceeguy Active Member

    Yes sir, you are correct, more load = less volts - less load = more volts. The PR12 (6V .5A = 3W) bulb was "standard equipment" with the chrome bullet style dynamo headlights. If you get a battery powered one (with 2AA cells and a PR2 bulb), just replace the bulb and wire it to the white wire. If your genset has a tail light, I would wire in a couple of super bright red LED's (with suitable dropping resistors) as the incandescent bulb may overload the lighting coil and weaken your spark.
     
    Last edited: Sep 2, 2008
  6. duivendyk

    duivendyk Guest

    The objective is to charge a 6Volt 5Ah SLA battery,then you have all sorts of options like running LED's which cannot run directly on ac.All this palaver about lights is beside the point.That so called 6 V/0.3 A ac signal does not look anywhere near a normal sinus wave ac signal it consists of short much higher than 6 Volt peak voltage ac pulses with nothing much in between,which does not make charging a battery without messing up the spark any easier.All this stuff that you have been talking about is old, old hat and has been chewed over time and again.What I need are some actual measurements,coil resistance between white and black wire,and ac output between white& black at say around 5000 rpm.
     
  7. arceeguy

    arceeguy Active Member

    Since the CDI unit does not have a separate trigger sensor, it probably uses the odd waveform as a trigger to fire the plug. This could be one of the reasons why an excess load causes a "no spark" condition, and not necessarily a reduction in voltage. Maybe something as simple as a full wave bridge feeding the battery might work out, with the resistance in the coil being the current limiter. If the ignition won't fire, just add a series resistor to reduce the load until the CDI will fire. (making sure there is enough tolerance for a discharged battery) In any case, if the lighting coil will only provide 3W at 6V, it doesn't matter how big your battery is because you have to at least replace what you use.
     
  8. duivendyk

    duivendyk Guest

    The CDI unit is a bit of a mystery, it could use the polarity reversal of the ac pulse as a trigger to discharge the storage cap,but I would think an increased ac load like a lightbulb would primarily affect the cap. voltage not the timing as such.A full wave rectifier is awkward to implement you have to float the whole ignition circuit, if you want the battery to have one side grounded.This would probably work OK but might be very confusing to the uninitiated, if they were to find out.Unless someone knowledgeable does a post mortem on a CDI we may never get to the bottom of all of this.
     
    Last edited by a moderator: Sep 2, 2008
  9. duivendyk

    duivendyk Guest

    RC guy,you seem to in the know.Can you do these measurements?.I would like not to overcomplicate things,A voltage regulator chip consumes current and you want to keep the idle current low (less than 1mA).The bicycle dynamos keep the output voltage at higher rpm from rising too much by having using the leakage inductance of the device ,this reactance appears in series with the output and the increasing impedance at higher frequencies (speeds) helps to keep the output from going up too much.Maybe something similar might hopefully be going on with the white wire.
     
  10. BoltsMissing

    BoltsMissing Active Member

    I have a spare CDI I can disect. I already have ground off the plastic coating to some extent, it seems like a metal box, but not sure at this stage cos I was doing another experiment at the time, not related to this topic, but it is there now to do as we wish if need be.

    This is my thoughts, and keep in mind I have not much to do with the practical side of electrics except some school projects in 1971/2, we made our own car battery charger from scratch, this means wounding by hand the layers of varnish coated copper wire, to get the correct voltage output and charge a car battery. It worked. There was a 3 prong thing, called a diode, or it would not work.
    The input power was from the mains, 240 Volts and the casing we made in the sheet metal work lessons btw, but long long time ago.

    What I am trying to say is, there is a formula as to how many windings are in a coil to get x amount of voltage etc. and what size copper wire/s.
    I don't know that formula, it's engineering level, this was just school project stuff, later it was all discontinued.

    I THINK,the HT magneto might be re-wound to suit, that might be step 1
    step 2 is to hook up the CDI to a computer to read it settings. I have read somewhere the Linux language is used due to it's simplicity.
    If we then can reverse-engineer the HT CDI, and change things or read what's there, it may be an advantage to see how it works.
    I know these thing can be done, I have seen it done with commercial black boxes where 2-way radios were used to identify a number.
    If you wanted to change the ID number from vehicle to base, they plug in the wires to the black box in the vehicle, and the numbers were typed in from another control box with a keypad ( digital), and presto.

    The other stuff I have read, is racing cars with electronics and a laptop to adsjust things.
    This is why I say, it can be done by those who know, and we here are only talking about the HT CDI, so I see no problem in this type of info. being made available.
    I hope, the formula be made available to re-wind our own coil and be able to adjust the CDI.
    If we can re-wind our own coil, then perhaps the white-wire can be used without much compromise.
     
    Last edited: Sep 2, 2008
  11. duivendyk

    duivendyk Guest

    You might be able to do a gross dissection of what I take to be deceased CDI unit,that is to identify major organs,like the output pulse transformer,that's the thing that steps up the voltage coming out of the energy storage capacitor and connects to the plug cable, and the storage cap. itself ,probably a good sized round affair with a lead sticking out of it at each end.But you would have a hard time with the nervous system,wiring,IC's etc.
    Frankly I'm kind of leery of the whole thing, not even sure it is a honest to goodness real life CDI!,it's probably not worth while shipping the corpse half way across the planet for my expert scrutiny.My aim is to come up with a HT charging system that people with your sort of background can put together with readily available parts and which of course works well.I have been to Aussiland,but only NSW & Queensland,skipped the Ayers rock, a real tourist magnet.Nice cities, impressive scenery,nice people and very good beer.My brother in law lives near Brisbane.
     
    Last edited by a moderator: Sep 3, 2008
  12. Drunkskunk

    Drunkskunk Member

    I'm thinking this may not be as complicated as it's being made out to be.
    A diode inline with a resistor from the white wire to the battery should provide some measure of charging. Using a zener diode in parallel should protect the battery from overcharging.

    Most likely, the CDI would discharge when the voltage goes low in the AC pulse, instead of going all the way negative. With that being the probable case, the white wire could be used with no fear of disrupting the ignition, as long as it didn't draw too much current.
     
  13. arceeguy

    arceeguy Active Member

    A CDI doesn't have to have any Integrated Circuit Chips. Just a diode, capacitor and an SCR that will trigger on the falling edge waveform of the magneto. A 5 wire pocket bike CDI with automatic spark advance would be nice if you could rig the pickup sensor and trigger magnet on a HT engine. (This means a separate CDI and coil though) I think the CDI on these units are probably very simple with no spark advance built in. I'll have to get a scope on the outputs once I get my Moon Dog barking!
     
  14. duivendyk

    duivendyk Guest

    Yeah,differentiating the negative going portion of the input wave form to derive a trigger pulse for the SCR could do it and provide automatic spark advance as well.I was thinking of using a Zener+ LED(as indicator)to start cutting off a normally saturated series pass transistor (or MosFet) when the battery voltage gets above above 7V.You would lose about 0.3 to 1V with that depending on what device is used.
    It would be very useful to get some idea what the peak to average ratio of the white wire signal is,I bet it's quite high.I have a peak reading Fluke but it just takes any old diode,a 10k resistor or higher and a cap across it with an RC time constant above 100 msec ( 100k with 0.1 micro farads).There is bound to be a lot of EMI electrical trash around.The rectifier could be just a single diode,or a full wave Schottky diode affair.But that gets you into some complications,either you have to float the ignition circuit in order to be able to have one side of the battery grounded, or the battery circuit itself has to effectively have the white wire ac on it.The first option looks to me like by far the lesser evil.
     
    Last edited by a moderator: Sep 3, 2008
  15. BoltsMissing

    BoltsMissing Active Member

  16. BoltsMissing

    BoltsMissing Active Member

    Disection Of HT CDI, during lunch break

    It was "war" !
     

    Attached Files:

    Last edited: Sep 4, 2008
  17. arceeguy

    arceeguy Active Member

    Looks like it is about as much fun as removing a fossil from granite!
     
  18. BoltsMissing

    BoltsMissing Active Member

    Going by the softness of the Copper wire, it became obvious that it don't take much to burn them out.
    The wire felt soft like paint brush hair.
     
  19. relaxxx

    relaxxx Member

    From image 12, looks like nothing more than a stepup transformer inside.
     
  20. BoltsMissing

    BoltsMissing Active Member

    Yes, similar principles weather it was a battery charger, electric train set variable transformer, whatever, coils of wire wound around a bar.

    The point is, the actual composition of the Copper wire.
    Convincing the fact further to use On/Off switch rather the the standard HT kill switch, in my humble opinion.
     
    Last edited: Sep 4, 2008
Loading...