crappy chinese components(specifically magneto)

Discussion in 'Electrical' started by rogersverobeach1, Feb 17, 2010.

  1. rogersverobeach1

    rogersverobeach1 New Member

    I've asked if anyone knows of some afermarket magneto's but no reply.
    This bike will be the death of me.i've had a Grubee for about 3 months.
    I put alot of miles on it,but already gone through four mags,two clutch cables,carb cables, and two carbs.

    This is obviosly a flaw in the engineering.It burns in the same place every time.Anybody with a little common sense and some knowledge of electricity and mathematics can tell you if you have a coil of wire(magneto)and electricity is traveling through it,when it comes to two 90o turns there is going to be a great amount of resistance and heat therefore burning the cheap crapp wire being used.this is at the last lenghth of wire where it is grounded to core.I'm tired of having to resolder this thing.the latest mag lasted maybe two miles.the third lasted a couple hundred.:veryangry:

  2. arceeguy

    arceeguy Active Member

    While I haven't seen a huge percentage of magneto failures, it does seem to be a weak spot - along with the CDI/Coil module.

    The only "real" cure is having a motor rebuilding shop rewind the magneto. It seems that many of the components used in HT engines are sourced from the same manufacturer and just assembled by different "manufacturers".
  3. Fabian

    Fabian Well-Known Member

    It's surprising how much variation there is between manufacturers of these engines.

    I've had a hellish time with connecting rod bearing failures, yet others have no such issues.
    In my case the electrical system has been the most reliable part of the engine assembly with 4,500 kilometers (2800 miles) on the original magneto and CDI.
    Yes i've replaced engines but swapped over the original electrics to test their longevity.

  4. AussieSteve

    AussieSteve Active Member

    Electrics/electronics is my favourite subject and was my profession for 20 years, so I'll add my 2c worth here. (Try and stop me!)

    After seeing how much $@#&*% I just wrote below, better make that $2 worth.

    1. A bend in a wire will not increase it's resistance. (If you doubt that, grab a low-ohms meter and measure a wire, then bend it and test again.) There will be no difference.

    2. Rewinding won't really help. To fit the maggy in the available space and supply the same voltage and current, it will need to be wound with the same gauge wire, with the same number of turns, to use the original core. The result will be the same coil. (I have a coil-winding machine and the knowledge, but wouldn't waste my time.)

    3. It is not typical to go through so many magnetos. Something else appears to be wrong. (4 in 3 months is horrendous.)

    4. You'd find that the cheap crapp wire is exactly the same as that used by most manufacturers. It will be copper wire, coated in enamel or polyurethane, with a very precise diameter to allow it to be used on coil-winding machines. They're programmable for very precise wire thicknesses or they wouldn't work. Even my home-made one is adjustable to within 18 micrometres. (18/1000mm)
    What causes the failure is an overload. The spot where the wire fails is inevitably the same in most cases. Where the wire leaves the main coil, but before the solder blob, the wire can no longer sink excess heat into the main mass of the coil, so it becomes the weak link and often goes open-circuit with excessive loading. I've read a lot on this subject on this site and it all fits.

    Do you have anything connected to the white (kill-switch) wire besides a kill-switch?
    How is your kill-switch connected? (white to ground or blue to ground.)
    If you're not using your white wire for anything, is it well insulated so that it can't touch electrical ground anywhere? Also, even a partial short between blue and white could cause magnetos to fail, so check your wiring carefully.

    Almost forgot - the other thing to look at is your waterproofing. It's possible that moisture is getting into the magneto.
    Silicon where the wires come out of the casing helps.
    Also, it's not a bad idea to use sealant on the magneto cover.

    Sometimes too much emphasis is placed on the components instead of the nut behind the wheel. Not calling you a nut - just thought it sounded good. I'm going through a second childhood, I think.

    ... Steve
    Last edited: Feb 18, 2010
  5. Fabian

    Fabian Well-Known Member

    Funny about that Steve - prior to reading your post, it seems i've followed your exact instructions:

    the other thing to look at is your waterproofing. It's possible that moisture is getting into the magneto.
    Silicon where the wires come out of the casing helps.
    Also, it's not a bad idea to use sealant on the magneto cover.
  6. Fabian

    Fabian Well-Known Member

    I've also slapped silicone sealant onto the black and blue wires where they enter the CDI, furthermore slapping silicon sealant around the outside of the sparkplug wire where it screws into the CDI.

  7. AussieSteve

    AussieSteve Active Member

    Yep, I put silicon around the HT lead junction on my CDI too. That's probably even more important than the black and blue wires, due to the higher voltage, (between 5,000V and 15,000V) and the chance of arcing with dust and moisture.
    Despite all the hype for fancy leads and caps, I'm still using the stock copper lead and cap and there are no problems. Nor should there be. We don't need low-noise carbon leads or other stuff. Nothing conducts better than copper, for our purpose and if you keep the lead and cap free of dust build-up, you shouldn't have any trouble.
    N.B. If you fit a radio or television to your bike, you might want to consider the low-noise option.

    With your mileage, I'll assume that you've adopted the recommended wiring. ie white to kill switch (only), blue to CDI. Definitely no kill-switch on blue.

    ... Steve
    Last edited: Feb 18, 2010
  8. Fabian

    Fabian Well-Known Member

    Yes, white wire goes to kill switch, and it makes sense as from what i understand, the white wire is low voltage, preventing high voltage arcing across the kill wire switch.

    Strange thing is, i was advised by another vendor that the correct method is to splice the kill switch into the blue wire and run it to ground.

  9. AussieSteve

    AussieSteve Active Member

    Yeah, mate, I've read that advice too, but don't follow it. There's a very good reason why your magnetos don't fail. You wire everything correctly.
    It's not merely a voltage issue. You'd probably measure similar voltages on both the blue and white wires, open-circuit, in the neighbourhood of 6-12V probably. Under load, though, such as a short-circuit from a kill-switch, the white wire winding would have a higher series resistance due to thinner wire and probably looser coupling to the core, so that not much current flows under that short-circuit condition.
    The blue wire is attached to a coil with a higher current potential, (thicker wire and better magnetic coupling) and can't handle a short in the same way.

    I've been thinking about another electrical issue - running a small dynamo/generator from the jackshaft to charge a small 6V/12V 5-7Ah SLA battery for lights etc.
    I'll let you know if I come up with anything decent. It's weeks away, but who knows?
    It would only be useable with a SBP shift kit jackshaft, though, because anything else would be too hard to implement. If I do come up with anything decent, I probably should have a chat to Paul and Jim at SBP.

    ... Steve
    Last edited: Feb 18, 2010
  10. wildemere

    wildemere Member

    Bending a conducter changes resistance, thats how loadcells and other transducers work. Not much of a change in R or revelance to the HT coil issue though.
  11. AussieSteve

    AussieSteve Active Member

    True, but not measurably in copper, with normal instruments.
    I have a bit of experience with load cells.
    A typical load cell is made of metal, but there isn't enough change in resistance under load to use. Thin wafers, usually 2 or 4, are attached to the sides of the bar and their resistance varies with load, (bending). I can't remember exactly what the wafer is made of, but it's not copper.
    Even then, a very carefully balanced wheatstone bridge circuit and amplifier must be used to accurately measure the very minute change in R.

    Other transducers commonly use a piezo-electric element.

    So, (to all intents and purposes), the resistance of a wire doesn't change when it is bent.

    ... Steve
    Last edited: Feb 18, 2010
  12. Fabian

    Fabian Well-Known Member

    I've been thinking about another electrical issue - running a small dynamo/generator from the jackshaft to charge a small 6V/12V 5-7Ah SLA battery for lights etc.
    I'll let you know if I come up with anything decent. It's weeks away, but who knows?
    It would only be useable with a SBP shift kit jackshaft, though, because anything else would be too hard to implement. If I do come up with anything decent, I probably should have a chat to Paul and Jim at SBP.

    Steve we think alike.
    I've been planning an idea to use a 2.7kw DC motor running off the chainwheel side of the jackshaft to even out the stresses once power is supplied to the DC motor.
    My concept is to have the 11T jackshaft gear machined to accept an additional toothed belt or 9 speed chain sprocket and locate the motor towards the top of the A-frame where the seat post tube connects to the top tube and the down tubes.

    Photo's of the motor attached.
    I'd have to rig up some sort of electrical speed controller and a method of allowing the motor to reverse operate by supplying current to a 12 volt SLA battery.

  13. AussieSteve

    AussieSteve Active Member

    Do you mean an e-bike, or a hybrid?
    If a hybrid, don't forget that whatever power is generated is robbed from the HT engine.
    ie If you generate 1kW, it will rob a bit over 1kW of power from the HT engine.
    I designed a regenerative controller a couple of years ago, but it won't drive a 2.7kW motor, only a couple of hundred watts.

    I'm thinking of using a 12V, 120W DC motor, to produce about 30W, I hope.
    (No-one would miss 30W of engine power.)
    I started a thread on this, in the electrics section.
    If enough people are interested, I'll look for commercially available motors/generators before I start, but if I'm only making a 1-off, I have a suitable motor here. (At least, I hope it will produce a usable amount of power, even at idle.)

    I was thinking of running a belt from the centre of the jackshaft, to the motor/generator mounted directly below the jackshaft, then a bit of electronics, of course.

    On many bikes, there's enough room under the engine, in the 'V', for a small battery.

    I just checked out the pic of that motor - it's not a standard DC motor, it's brushless.
    A regenerative system might be hard to accomplish.

    ... Steve
  14. Fabian

    Fabian Well-Known Member

    I guess you could call it a hybrid system.

    I intend for the piston engine and the electric motor to power the jackshaft system at the same time and also to use each on their own, but allowing the piston engine to generate electrical power, recharging the battery.

    The thought of 1.5kw of petrol power (maybe more with the new Pirate Cycles stage 3 engine) and at the same time slamming in 2.7kw of electric power really appeals to me.

  15. Fabian

    Fabian Well-Known Member

    The method of location i intend to use.

    Attached Files:

  16. wildemere

    wildemere Member

    How are you going to gear the high speed (probably over 4000rpm) brushless motor to the slow speed (around 600rpm) jack shaft?

    Also that motor is 36v and 60A, any thoughts on the controller and power supply setup?
  17. Fabian

    Fabian Well-Known Member

    No thoughts on the technical intricacies as yet.

    Step 1 will be to bolt the motor into the frame - you've got to start somewhere and once it's attached to the frame you're compelled to go forwards and complete the project.
    Step 2 will be working out the mechanical drive system to the jackshaft.
    Step 3 will be leaving the set up of the electrical system to a friend of mine who is an electronics engineer.
  18. AussieSteve

    AussieSteve Active Member

    2.7kW of electric would be nice, but not practical. Besides wildemere's observations, have you thought much about batteries and range? (You won't fit the battery(ies) under the motor in the 'V'.)

    Speed control of the brushless motor will be another big issue. My e-bike has (had) PAS intelligent pedal-assist and even that was totally inadequate at pedalling speeds, so I eventually disabled it.
    Matching HT speed to e-motor speed will be another big tripping point. (A hall-effect sensor throttle coupled to a cable throttle, or electronic (PAS-type) control, either won't be easy. As an ex- electronic designer, I'd pass this one up.)

    My e-bike hub-motor is 200W and has a 36V, 10Ah batt - about 5.6A current drawn at 200W. It only has a range of 32km. (Have a look on my profile page album to see the size of that battery, then transpose up to your requirements. Possibly a small trailer full of Lithium batteries.)
    Imagine the size of battery you'll need to get any range with a 36V, 60A, 2.7kW motor, even used in conjunction with the 2-stroke.

    (Sorry to rain on your parade. As I said, it would be nice to have a 2.7kW electric, but.....)

    ... Steve
    Last edited: Feb 19, 2010
  19. Fabian

    Fabian Well-Known Member

    I never said i intend to use the electric motor at 100% duty cycle.

    The electric motor is more intended as a charging system to top up the batteries (to be housed in the trailer) and for powering an automotive lighting system.

    My intention is to run the piston engine at 100% duty cycle.
    When going downhill, i'll hold full throttle and let the system generate power.
    When riding on the flat, i'll use the piston engine on it's own and when going uphill, the electric motor supplies additional "burst power".

    Last edited: Feb 19, 2010
  20. AussieSteve

    AussieSteve Active Member

    The biggest problem there is that you'll still need a battery big enough to supply up to 60A current. That's a small car battery in lead-acid or a big bank of Lithium cells. 60A is one **** of a lot of current. Your wires would need to be almost as thick as those going to a car battery.
    Then there's the brushless controller...
    (Duty-cycle on an electric motor is slightly different and will vary depending on speed, as well as the 'duty cycle' you speak of, ie how long the motor is used for. If you're giving it a bit then you will draw 60A.)

    I'm honestly not trying to upset you or start an argument, but it really isn't practical on an MB.

    The censorship is over the top, I only said h e l l.

    ... Steve
    Last edited: Feb 19, 2010