Stator winding patterns? (includes smoke let out of wires)

[stinkwheel]

Ok, full story. I've built a Belarussian Minsk 125. As standard they run off a 12v dynamo with an electromechanical regulator. They are famously rubbish so i didn't bother with it.

I wanted a powerdynamo from a German company called Vape but they stopped making them for this bike.

I found a Ukranian company (Vape UA) making kits for them ordered and fitted one.

The kit has a replacement rotor and stator with a little self-contained AC CDI ignition module and a single-phase star-configuration charging system. The ignition part of it works absolutely fine.

The top coil is the ignition one and it runs off a hall effect sensor, all good.
https://lh3.googleusercontent.com/pw/AP1GczMUkQlqwT1zN3lnRvzgnQxr4BhLHRNror-btQDRVRA6aKSe1qZkR6rzp3G4jLGxxgZKXEpcypmxS0t49fOwsujDl6GIumWDPi0E1PWybuF5aGwtbMzAkU84YRAFrxmFAza_dBKIei-DXcC4AX21bTkF=w1547-h870-s-no

The charging part of it has been sketchy from day 1. Tried it with LEDs and they were super flickery, tried a variety of smoothing capacitors which were no help with this. Went back to incandescents but it was struggling with these too and cooked the electronic flasher relay and was popping bulbs on occasion.

Replaced the universal 4-wire Chinese reg/rec it came with with a FH020 mosfet. This seemed fine when tested. Lasted about 5 miles then popped a load of bulbs. Failed to high voltage.

Fitted another one with a big capacitor in paralell to the output. This also seemed to be working ok (not great but the lights worked) but the lighting died after about 10 miles. On returning home, I checked the stator, it was open circuit. Visual check confirmed distressing burning smell and lots of scorched plastic and wires.

https://lh3.googleusercontent.com/pw/AP1GczP4dUJNG5HMAoAFLThDr0a3eImvTuQL3iCH94ow96voATrBYX5zXDzmHFCFLr8Ce7SGc6y8oxJxvbjhE1wdM0wRtEW4soeUmz3BSShEiByHGKPiRkdmNek46UlRk3EwIML1NIpPiOrjNFrNdNB7Pgwb=w1547-h870-s-no

I set about dismantling it, to see how it was configured as much as anything. Two things I learned. 1) It's a standard "universal" Chinese stator bolted to a custom made backing plate. 2) The wires aren't copper, they are aluminium with a copper-ish coating!
https://lh3.googleusercontent.com/pw/AP1GczNjHccYsBj9XDaupmoXSCJaYHEsAn1x0atwQxm8Y2tsP6GvXtarrLFSARJ6sT1TcBarfae7QfrqOZ-They68kSwcHu8EYvMiz3O1GXiKvN-jg8zUgFOHnWt_NLVMsH2JrF1-xxOSQxk-hxiTUGIl2Xm=w1547-h870-s-no

I've ordered another one with unmelted bobbins, £20 on amazon. It's my intention to rewind it with proper enamelled copper wire.

Now the question. Do I wind it how it was or do I do something different? It doesn't seem quite correct to my amateur understanding of how these things work.

Here's what I have. The top coil at 12 o' clock is a very finely wound high-voltage coil to power the CDI. I'm leaving this as-is, seems to work fine. The remaining 7 coils each have 55 turns of 1.0mm wire. The winding direction is opposite for each adjacent coil as you work round the "star". The two outputs were fed into a standard 4-wire shunt-type reg/rec.
https://lh3.googleusercontent.com/pw/AP1GczPRNjl1iLIB_02Z_J4rdQOLlRQRTc3lvZr_G3l81N0o3DjM-lDnvaR7mtuuymmxivwNkYiwxLm4QyPVQiQ7kxQWN02rUgTnb6ZQIMFr_UdaIL7SGlc1DNnanhwkQVzLOfn5BYZAJQWL00SFmIw_GNog=w592-h870-s-no

Now, one thing I noticed is the opposite coils in this setup are wound in the same direction when viewed from the top of the coil. Is this right? I thought this would give you weird overlapping field effects and cancel out a lot of the voltage? I thought opposite coils were supposed to be wound in opposite directions?

I double checked with a compass and the rotor is absolutely vanilla. Eight equally spaced magnets alternating in pole orientation as you go round.

One thought was to miss out the 6 o' clock pole but carry on with the pattern so the opposite poles are wound in opposite directions.

Another thought was to re-configure it with two wires wound onto every other pole and miss one out so each wire has two opposite poles wound in opposite directions plus a third pole equally spaces between them. pair one side so I have three outputs to feed into my 5-wire reg/rec. As per the following diagram:
https://lh3.googleusercontent.com/pw/AP1GczMYp0YyB0F3z5nhEJ9m3M1tO5ISjUWIV7kZhHCXHuJVOtou0eHiw3_1B-z5wM-MCwO10iN3KWcPl8Sd8k399beaUh9UIn7Y_mBK7Kdcs7bqXFbkhmcijOFxqrq7-kvqUQEVAr9gbRq0joJKazJugbDN=w581-h870-s-no

Another thought was to wind it as standard but with a take-off exactly halfway leaving me three outputs to feed into the five wire reg/rec (broadly as suggested by WD forte in my build thread). As below:
https://lh3.googleusercontent.com/pw/AP1GczM-tzgbCe9FNLvFJvhPIOVk0zSLaB21c_x7ocfghDJqGC3ZJaXrxV7XhP9T7tLe5i3NBjrfajF3kLvWIHdE46oMOn74PJaf1n4GwdqH7bV8ThdO83sk1dYCcTf_ilMlPB2o_ebycyiAzJA12MOndnt5=w610-h870-s-no

Thoughts/questions/ridicule?
____________________
“Rule one: Always stick around for one more drink. That's when things happen. That's when you find out everything you want to know.”
I did the 2010 Round Britain Rally on my 350 Bullet. 89 landmarks, 3 months, 9,500 miles.

[Easy-X]

Copper coloured aluminium?! Clever chaps, those Chinese Except Aluminium has more resistivity than copper which makes you wonder if that led to the coils getting cooked. Probably means you can use less than the original 1mm diameter in copper - less resistivity and better at conducting away heat. Shame you have the ignition coil up there at 12 o'clock otherwise it'd be easy to wire up a typical three-phase generator

I'd say it's probably worth picturing it in terms of how you'd wire a DC electric motor with regards to coil wiring/arrangement. IIRC there were some bikes that the generator doubled as starter motor. Thing is "dc brushless motor" mostly throws up three-phase designs. I did turn up this:

https://i.stack.imgur.com/ip5PT.jpg
(source: https://electronics.stackexchange.com/questions/365796/how-do-single-phase-bldc-motors-start-in-proper-direction)

And if I'm reading the diagram correctly one would wire it up clockwise, anti-clockwise, clockwise, anti-clockwise... with the bottom of the first coil coil connecting to the top of the next coil.
____________________
Husqvarna Vitpilen 401, Yamaha XSR700, Honda Rebel, Yamaha DT175, Suzuki SV650 (loan) Fazer 600, Keeway Superlight 125, 50cc turd scooter

[jeffyjeff]

[stinkwheel]

[Robby]

I can't help with the detail on which way to wind the wire, but I am looking at this from a general fault finding perspective. Step 1, establish a good baseline and eliminate all known problems and causes for concern.

The known problem is that it let out the magic smoke. The causes for concern are aluminium wire, and you putting non-standard loads on the system (LED bulbs instead of incandescents, capacitors all over the place).

So my fault finding preference to find a baseline is to make it shit but standard. Replace the aluminium wire with copper, but stick with the shunt reg/rec and incandescent bulbs. Get enough enamelled wire to re-wind the stator several times, you'll probably cock it up or want to mess with it in the future.

My alternate suggestion, that I would consider, is to see if a generic pitbike setup would work and redo all the electrics from scratch using that. If it fits, it means that replacement parts are very quickly available and cheap.

[stinkwheel]

[stinkwheel]

Any reason I can't fit a temporary fuse in one of the outputs for testing? For example, I don't know if my reg/rec is still good and they aren't simple to test except by plugging them in.

The enamelled copper wire is rated to 1.21A so if I fitted a 1A automotive fuse in one of the stator outputs, this should prevent it going full meltdown if there's a fault?
____________________
“Rule one: Always stick around for one more drink. That's when things happen. That's when you find out everything you want to know.”
I did the 2010 Round Britain Rally on my 350 Bullet. 89 landmarks, 3 months, 9,500 miles.

[WD Forte]

1mm copper wire is rated well above 1.2A
more like 10A but 5A would be safer useage

3x 5A phases would give you a 180W stator @ 12V

This subject can get very complicated and confusing
Star wound? Delta wound? etc so going back to basics is
a good start.
I usually start with paper and pencils to sketch out the plan
and follow that
One thing to note and you no doubt agree on
is you don't need to wire it for max powah
you have a working AC CDI so all you need is simple, modest
and reliable supply of DC for leds n shit

Those early CG type stators above are common as muck
One big fat source coil to feed AC to the AC CDI box
the other windings split into AC feed for lights and the rest often rectified for other stuff.

My first thoughts
( modest supply for leds, dont overwind them, you wont need anywhere near the current incandescent lights would require )
would be to try 6 coils Y wound into 2 phases running into a 4 wire RR.
I'd use 40 turns of 18 gauge/1mm copper wire( magnet wire)
per bobbin
one of phase tail going to the RR
The other joined to the free tail of the other phase and insulated from ground.

I'll guess this would give you around 100-120W
bit you only need say less then 50W
to run the leds and excess power will have to be burnt off
by the RR so don't over wind.

One plus is its easy to get at for modding/upgrades if any were needed
____________________
bikers smell of wee

[stinkwheel]

[WD Forte]

Typo?
Many 250w stators use 18ga/1.2m copper wire and they must be running up to 250W/14V = 18A ish
Thats around around 6A each per phase.
I've used 1.2/1mm copper wire for stators

You've no starter suck the battery dry on cold mornings
so unless you want gadgets like heated grips and ferkin great spotlights
you just need enough to run the leds and put an amp or two into the battery.

Put out too much and the RR has to burn it off remember
guesstimates:
10W max for an LED headlamp, 20W if stupid bright
2w each for the other blubs
lets say 40W max for lights and 10W to maintain the little battery
50W @14V = 3.5A


you could wind 2 sets of 3 into 4 pin RR for 2 phase
or
3 sets of 2 into a 5 pin RR for 3 phase
and hopefully get a smoother less choppy output

If the magnets are at 90 degree intervals I'd want the windings to be at 120 degree or at least not 90 or 180 degree intervals to avoid big peaks and troughs in the output.
____________________
bikers smell of wee

[stinkwheel]

Ok, here's what I came up with. Slightly numb routing to avoid crossing the high voltage ignition coil 1 at any point.

https://www.bikechatforums.com/files/windings.png

Now, polarity/directionality. If I pick a single point of rotation with this configuration, every clockwise winding has a South magnet passing over it and every anticlockwise rotation has a North magnet passing over it at the same time. This situation reverses every 45 degrees of rotation. So that's what I want?
____________________
“Rule one: Always stick around for one more drink. That's when things happen. That's when you find out everything you want to know.”
I did the 2010 Round Britain Rally on my 350 Bullet. 89 landmarks, 3 months, 9,500 miles.

[jeffyjeff]

If indeed you intend to connect a regulator-rectifier at the locations of the black dots, you would obtain an unbalanced three-phase output, I believe, as shown in diagram A below.

https://live.staticflickr.com/65535/53613197309_110703c6be.jpg

Here is an idea: Wire every fourth coil in opposite orientation to obtain a true three phase output (albeit the legs would not be 120 degrees apart due to the configuration of the windings).

Coil 2 CCW - Coil 5 CW Line 1
Coil 3 CW - Coil 6 CCW Line 2
Coil 4 CCW - Coil 7 CW Line 3
This would result in an output shown in diagram B.

But you could get a slightly more balanced output by wiring as follows to result in an ouput shown in diagram C.

Coil 2 CCW - Coil 5 CW Line 1
Coil 3 CW - Coil 6 CCW Line 2
Coil 5 CW - Coil 8 CCW Line 3

In the end, any three-phase configuration you wire will leave one coil unwound, reducing output of the stator. If the project were mine, I think I would wire the stator in series for a single phase output, and see if I could source a solid state single-phase regulator-rectifier somewhere.
____________________
History shows again and again how nature points out the folly of men - BOC

[jeffyjeff]

Let's try that again.

https://live.staticflickr.com/65535/53613525739_af098ccc2a_n.jpg
____________________
History shows again and again how nature points out the folly of men - BOC

[stinkwheel]

Well, I've gone for the kind of star-shaped 2-phase thing I pictured above. 40 windings per coil as suggested by WD Forte.

Took 3 1/2 hours from start to finish and used wire left over from the last alternator rewire I did. Even managed to rattle the bottom of the toolbox for some sleeving.

New stator £17 from Amazon. Don't worry about the hole orientation, I have one upside down. It turns out this one was wired differently again. It was odd and I'm not sure if or how it would work. Basically two phase tied to earth but one phase only had a single coil, coil 2, then into a common then another 6 coils in series then earth. Anyway, I just wanted it for the ignition coil and the backing assembly.


https://lh3.googleusercontent.com/pw/AP1GczPM1ohU7rfacUuwoYRBbsXMkW8wfqiTxrboV3jFdHv-hKAQ1FBqA2ZRRbRlYPFhgs8yt4y7KUKwo-TNiMD2ubNcDm3Yc9aJ_AQz8WSI0DGcc1IB-DL4zZgUy74RF7Pyq2lYZhU8loNZMtcislvB4EQV=w1547-h870-s-no

First set of 3 coils wound, commoned into the second and made a start on the second set of 3 coils.
https://lh3.googleusercontent.com/pw/AP1GczPYTDsSi9DNIom-a48OP8a8YatqwjzzEyV-gWSzUKB_fF51wzht__AHMRKNL6kuiZ2QL0c_y2YtD9hAxySPK5HnkeCn1MxTlPxygWSxSCncYGo1QKk560tl6Hd3FwZV4Fxn5lw8ODau0s7uwXgZmsUG=w1547-h870-s-no

Finished winding.
https://lh3.googleusercontent.com/pw/AP1GczP0ChOVRXeDFSSg3lPOSZSulunETN4gZVTtvSo6UD28ZU-D8hkj2OfE2_M3QFDOiyIYfaTVvUSbU3PvS04bRRcow0oQNHzlbhD8gJkksP11hOktqQYoR4DtOikYRQ_Cz8V7228dCImUSAvAEmNhiQGt=w1547-h870-s-no

External wiring attached, routed and sleeved. Resistance tests check out from the end of the wires so 490 ohm for the ignition coil to earth. Open circuit to earth between my two yellows and a very low resistance between the two yellows of about 0.3 ohms (at the limit of the accuracy of my meter) which is about right for roughly 15m of 1.0mm copper wire.
https://lh3.googleusercontent.com/pw/AP1GczPXIW7ax5fnOI2vAgmBX19FAKm7E7nHUvaYkOUC7eTTaKGnsuy46uqrUQPR_s0gb2lRXTbGKtKgycsZo6_w1ElX8CA9XoENjy0lr0nHqysASeKC7SSq06aDFwRV2jW8woqTMKi1bEs2qmuR8wsi5P99=w1547-h870-s-no

I'll attach it to the bike and do some tests tomorrow. Not a good time of day to be revving up a 2-stroke just now.
____________________
“Rule one: Always stick around for one more drink. That's when things happen. That's when you find out everything you want to know.”
I did the 2010 Round Britain Rally on my 350 Bullet. 89 landmarks, 3 months, 9,500 miles.

[jeffyjeff]

Good on you. Will be interesting to hear how it works.
____________________
History shows again and again how nature points out the folly of men - BOC

[rpsmith79]

Ah ha, a fellow DMCer is see Stinkwheel

Thought it looked familiar when scrolling through FB this morning
____________________
Current Bike: Honda CG125 ES4 // Honda CB600FS Hornet // Triumph Street Triple R

[stinkwheel]

[rpsmith79]

[stinkwheel]

It works.

Fired right up and generates between 12and 30V AC.

With the zennor type reg/rec it lights 45W of incandescents and works the horn. That's the original Chinese reg/rec it came with and it's putting out 16v with or without a battery so that needs to go in the bin.

The mosfet reg/rec is not at all happy though. It sits at 15v at idle with either a battery or a capacitor connected but flickers and drops to 10v when you rev the engine.

I'm still not convinced this isn't ignition noise interfering with it. My digital mutimeter goes haywire when you rev the engine, even if it's not connected. I found instructions for a sparx reg/rec and they were very definate on the necessity of using a supressor cap.

I have epoxy on the coils curing just now but I'll try it again tomorrow with a resistor cap. It doesn't damage the wiring, it can cause the plug electrode to nibble away but that's not going to happen in a short run.
____________________
“Rule one: Always stick around for one more drink. That's when things happen. That's when you find out everything you want to know.”
I did the 2010 Round Britain Rally on my 350 Bullet. 89 landmarks, 3 months, 9,500 miles.

[jeffyjeff]

Just to be visual, I think this graphic is an approximate illustration of the wave form you are sending to the Mosfet regulator. Not to scale, but illustrative of the way your windings are wound and connected. Might explain the erratic reaction of the Mosfet R/R.

https://live.staticflickr.com/65535/53615544411_02cd4af3f8_m.jpg
____________________
History shows again and again how nature points out the folly of men - BOC

[stinkwheel]

[jeffyjeff]

Consider that maybe you just don't have enough wire turns to power your charging system. Grasping at straws here, perhaps, but look at a typical 3 phase stator; it has 18 coils arranged in three groups of six. I'll wager that each of those coils has more than 40 turns wound around it.

But for the sake of illustration, suppose that each coil has 40 turns, as does your stator. That would equal 6 X 40 = 240 turns, time 3 phases, comes to 720 turns of wire on the stator. Your stator, as wired, has 240 turns split between two phases.

I'm thinking that it is a possibility that your reg/rec is working fine, and that the stator output is just not enough to support the ignition system as you rev the engine up. Higher RPMs demand more ignition pulses per second. Maybe the voltage is dropping down to 10V because the reg/rec is maxing output to the battery, and that stator output is insufficient.

Don't know why the voltage would drop as low as 10V, however. Seems that the battery would pick up the slack before voltage dropped that low.

Did you count how many turns were wound around each coil when you took it apart?
____________________
History shows again and again how nature points out the folly of men - BOC

[stinkwheel]

The ignition and the charging system are totally seperate. All the bit I fiddled with does is run the lights.

I'm trying a different reg/rec. The mosfet obviously doesn't like it and the zennor one it came with is pinned to 16V, which may be a large part of why the stator failed. Also explains why it was exploding LEDs. I've ordered a podtronic one with an integral capacitor.

Brief testing with the overvoltage one shows at anything above low idle, it will run the lights and sound the horn both at once. That's good enough.

I put a supressor cap on to let me connect it through my digital ammeter so I can check what it's drawing (mainly to check there isn't a short somewhere). The supressor cap stopper the meter going haywire if it's anywhere within 2 feet of the bike when the engines running (even if not connected to anything) but I'll need to use one with a smaller resistor so i don't cook plugs. That's on order.

It draws 3.5A with the lights on, maximum of 6A with the lights on, brakes on and sounding the horn. That'll be lower if I can get it to light LEDs reliably. I'd really like to run an LED headlight, just for the better light, a 35/35W tungsten filament doesn't really cut it.

For reference, this dinky thing is the one on my enfield which will power all the lights and keep a 4Ah battery fully charged.
https://lh3.googleusercontent.com/pw/AP1GczPkRddp5FvvZQeamUbC5H8lhZA0YvotBSMyz00hffLb1cpAVD0BBkeXLGeMdjCGsL3P-Tr4jsfE2xWOWEMet5yEzDLi8IlpBMRG5ctaHyuww1FsGPmRmQfv73GAezsPYthGOwH3J7jqURMj-QU0L6of=w1160-h870-s-no
____________________
“Rule one: Always stick around for one more drink. That's when things happen. That's when you find out everything you want to know.”
I did the 2010 Round Britain Rally on my 350 Bullet. 89 landmarks, 3 months, 9,500 miles.

[jeffyjeff]

Thanks for posting this interesting project. If I am so fortunate as to ride either of my motorcycles long enough to require stator replacement, I am inspired to try rewinding it myself. I hope your repair works out for you in the long term.
____________________
History shows again and again how nature points out the folly of men - BOC

[blurredman]

And you thought you'd save hassle by not using the original
____________________
CBT: 12/06/10, Theory: 22/09/10, Module 1: 09/11/10, Module 2: 19/01/11
Past: 1991 Honda CG125BR-J, 1992 (1980) Honda XL125S, 1996 Kawasaki GPZ500S.
Current: 1973 MZ ES250/2 - 17k. , 1979 Suzuki TS185ER - 9k, 1981 Honda CX500B - 91k, 1987 MZ ETZ250 (295cc) - 38k, 1989 MZ ETZ251 - 49k