Dirt Rider . Net Text Version Home
Pages: 1
Lighting Coils - A State of Flux
(Click here to view the original thread with full colors/images)
Posted by: MACE
I need more juice from my WR 250 popcorn machine (2st)lighting coil.
Now I can pay someone $75 to "rewind" it and get about 75 watts at some unstated RPM or I can try to understand the physics and mechanics of lighting coils.....
Well I hate to just buy stuff I can make myself. So lets talk lighting coils.
Anybody out there rewind their own coils. You XR guys just wind empty poles, but anyone actually figure our the combination of wire diameters and lengths to get more juice?
I've got a lot of thoughts on this but before I do a lot of typing, let's see if there is any interest.....
------------------
MACE
One night I was layin' down,
I heard mama 'n papa talkin'
I heard papa tell mama, "you let that boy MOTO,
it's in him, and it got to come out..."
Posted by: cujet
I have done it, with some success. You will need to calculate the number of turns of the wire to determine the voltage. For instance (example only) the stock coil will produce 13.5V at 3200 RPM into a 35W headlight and a 7W tail light. The stock coil has 1756 turns.
Now remove the stock wire and install the next heavier gauge wire. Lets say you can only fit 1300 or so turns. You now have the ability to produce less voltage at 3200 RPM. However as the revs increase so will the voltage and the larger wire carries more current. So you can now run 135W electrical load. As you can see, this setup requires some (maybe not much) addl RPM.
The Ideal setup is to run the correct number of turns to produce the necessary power at a lower RPM. Then, of course the voltage rectifier/regulator prevents overdriving the lights at higher RPMs. Keep in mind, the neater you wrap the wire the more turns you can fit.
I tended to bulge the center of the windings just like the shape of the flywheel. In this manner I could fit the most wire.
Good luck!
Chris
------------------
94 Husky 360 2 stroke, Street legal,(gotta clear my throat now)! Ported engine, FMF pipe, kehin PWK carb, full knobbies, Pro Action Showa suspension, Pro tapers. 50HP
Posted by: MACE
Oh boy, someone wants to play!
Correct me if I'm wrong....
I think the WATTAGE you can generate is only a function of engine speed and magnet power. There is no other source of energy.
The voltage is proportional to the number of turns. So at any given engine speed adding turns will increase voltage and reduce current (amps) with the WATTAGE staying the same. The only way to get more WATTAGE output is to reduce the loss to wire resistance that becomes heat - which is what you did with the larger wire diameter.
One idea to decrease the resistence and still pack a lot of turns on is to use square cross section wire (I found some sources on the web). I like your idea of building up the middle portion. Any chance that the uneven windings could induce reverse flow within the coil?
The other plan is to increase the magnet strength by mounting the poles as close as possible to the magnets or increasing the strength of the magnet.
What if I replaced the stock magnet with one of these monster neodymium mags. http://www.wondermagnet.com/dev/magnet7.html
Does anyone know enough about ignitions to know if they have an internal voltage regulator so that they won't fry if I start pumping ten times the current through the ignition stators with the killer magnet?
Is there anything to gain by increasing or decreasing the number of laminations on the core?
Now where did I put that Physics book?
Posted by: techman
Hey Mace, when you get it figured out let us know. I scratched my head about it some time ago as well. I recall that ultimately the power transmission is limited by flux across the air gap. Hence stronger magnet, shorter gap, wider surfaces at the gap, and better flux return path give better ultimate mech to elec power generation potential. However, I think the factory systems out there are not electrically wound to match that maximum value. Additionally, there's the VI form the power takes. Ignition is big V, low I. Lighting is big I, low V. The heating loss and volumetric packing efficiency in the copper is balanced vs voltage at what rpm (and possibly having to dissipate power in the regulator as well). The gnarliest dc systems I know are low V, high rev, verrry high I model airplane motors. They achieve high power to weight and volume. In our case, we're trying to pack power into volume and weight doesn't bother us much. This is where Cujet has fit different wire into the same volume, trading V for I. His advantage comes from getting the same regulated V after a threshold rpm, but higher I and hence higher output power. In the XR rewinds where they use more poles, they make the wire bigger and get the number of turns back again by having more poles to fit them onto. Hence near same turns but bigger wire (and I capacity) = bigger power out @ 12V. I would bet you that if you found that chapter (in an electric machines text, not physics text) and plugged in the air gap geometry and a flux approximation you would find that the generatable power was real big, not nearly tapped by hundreds of watts. Or it could have been optimised(?). Or maybe at 15,000 rpm you can get huge wattage with the proper coil winding.
The real mystery to me at the moment is how the shunt regulators work - do they open circuit the coils, or do they switch them and ride some type of inductive energy storage and then current decay/release? Or if they don't, is that the real secret to higher power output - make a regulator that isn't dissipative but instead uses switching and inductance for dc-dc power conversion, thus not wasting any output (can you say bright headlights at idle?)
Mace, if you run stronger magnets, say 1.5x, you could increase the ignition coil voltages x1.5 (if the magnetic system doesn't saturate first - ugh, saturation could cause harmonics that could mess with everything) and those voltages might get too close for comfort to the overvoltage ratings of the solid state parts in the ignition. Maybe a bad idea.
Changing laminations could give more cross section to the magnetic circuit and change the flux saturation point, or the amount of operating flux. However, the air gap is the big choker for resultant flux density. It's hard to mess with magnetic systems because of all the interactions. They're also generally optimised to reduce screwup effects like saturation and hysteresis. I'll bet the magnetic design just keeps the iron core path out of saturation. I think that's that turnsxI thing again.
The electrical side is much more manageable. I believe the maximum current x coil turns is related to max flux (?), and given you've attained that, the rpms create the voltage so you get power out proportional to rpm (assuming you have a load that can draw max current at your rpm voltage) and max power at max rpm. Resistive losses in the copper and hysteresis losses in the iron are just sort of incidental bull**** that one has to add in. Both can be overheating sources and power drains. The copper R also affects the loads you can run(R is related to turns and gauge), i.e. where in the spectrum of VI possibilities.
I think Cujet has the practical approach that ought to work for do it yourselfing. Count your old # turns, go down a wire gauge, fit what you can on, and expect more low rpm voltage drop out but higher current(power) at speed. Be as aggressive as you dare.
Yeeehaw! My brain feels alive! But not as alive as during a wheelie!
------------------
Posted by: cujet
Techman, good input. On many of my bikes I should have designed a lighting coil with 25 to 50 percent more laminations. As I had/have the room. My current bike runs a 85W and a 35W headlight at the same time along with the 7/21W tail light. It is fair at idle but great at any speed over idle. I use a shunt type regulator (stock Husky) and I was told it was a zener diode. I find that hard to believe, yet it works well.
The flywheel magnets are about 25mm wide and the lighting coil is around 16mm. There is room for addl laminates and therefore MORE WIRE http://dirtrider.net/ubb2/smile.gif However I really have enough power for my needs as I can run some very bright lights. But isn't more always better?
Chris
------------------
94 Husky 360 2 stroke, Street legal,(gotta clear my throat now)! Ported engine, FMF pipe, kehin PWK carb, full knobbies, Pro Action Showa suspension, Pro tapers. 50HP
Posted by: MACE
I think I remember that the magnetic strength(flux?) is one of those cubic functions. In other words if I cut the gap in half, I get eight times the "flux". This makes me inclined to ovalize the holes on my coil and trial and error them to give minimum clearance to the flywheel. I will have to make some allowance for crank flex and whatnot but I bet Yamaha is very conservative on their clearances for 100% production line clearance.
I read on the Electrex website that you can't really measure the AC output. Any ideas on quantifying the output? I think I'm going to get a Tympanium reg/rectifier (http://www.fourstrokesonly.com/Elecmisc1.html) and then I'll just measure the DC output.
Does the core know the difference if it is laminated? If I wanted to increase the core size could I just add a single thick plate trimmed to the correct periphery? Mild steel or what???
(BTW, I am pursuing this because I decided I couldn't justify adding the eline coil to my three year old bike. To much $ for something I won't be able to transfer to my next bike in a few years....)
Posted by: cujet
The core will saturate at a much higher point if laminates are used, therefore you should use laminates. A solid piece of steel may not give you what you want.
Chris
------------------
94 Husky 360 2 stroke, Street legal,(gotta clear my throat now)! Ported engine, FMF pipe, kehin PWK carb, full knobbies, Pro Action Showa suspension, Pro tapers. 50HP
Posted by: Kramer
Cujet, is ti important that the laminations are seperated by a nonconductive film? I think I remember something like that??
------------------
Kramer
member BRC, AMA
98 WR400 Dualsport!
Posted by: techman
Yeah, they are electrically isolated to make the eddy current paths smaller.
If you add laminations, you get a fatter pole that has more cross sectional area, lowering the reluctance - hence more flux per magnet gauss, which should cause higher voltage in the coils. Hence your lights would be brighter at lower rpms than before. But I think you get higher winding diameter with fatter poles - more winding dia = longer linear piece of wire and higher resistance, a bit of a loss. However, if you kick up the wire dia, you drop the resistance and probably make up for it.
So Cujet,I think you've been boosting your low rpm voltage with extra laminations and making more efficient use of the available magnet power - and presumably you've upped the wire dia to handle the bigger currents your pulling in your load. Yes, a linear type shunt regulator is just a zener, or maybe a bunch of forward diodes stacked in series to add up to 12V (17 x 0.7v = 12V) - I've seen weird looking old regulators from 25 years ago. Nowadays big power single zeners should be available. My mystery still stands - how do the switch-mode based regulators work, and is there dc-dc power conversion going on?
Mace, I think if you drop the air gap, you might take the metal into saturation, causing harmonics and weird spike activity. Cujet's approach also drops the magnetic path "resistance", but actually provides more metal material to handle the higher flux level. Yours just crams it into the same metal pathway, raising the amount of flux per metal. Too much flux per metal is saturation. That's when induced voltages and currents don't behave nicely. You'll get higher voltage at lower revs, but not higher power output as the limit would still be resistively heating your still-small-gauge winding wire.
I'm a big fan of gnarly lights for night riding. BTW, did anyone see the Baja Designs single extra 2" wide angle bulb placed above the main headlight setup on Speedvision tonight? That single bulb is the same one used in the Acerbis DHH setup. The bulb itself comes in 25W, 35W and 50W powers. I've got 3x 50W's and a 35W on my bike. Melts away the night.
------------------
Posted by: MACE
I was kind of hoping that Brittany Spears had some insight on permanent magnet generators but I guess she sticks to quantum physics. http://dirtrider.net/ubb2/wink.gif
Techman, I'm jealous of your lights. I did try to pull up the great discussion we all had a while back on eline coils but it seems to have fallen off the edge of the earth.
I saw the BD lights in Dirt Rider. (I'm also jealous that you get Speedvision - all this envy is bad for my soul...) Aren't those bulbs (MR16 type) the same ones sold at Home Despot and others? I used the Home Despot bulbs for my killer helmet lights I use with my XR. You can get some cheapo "driving lights" with lightweight plastic bodies at some auto departments that use the same bulbs.
Anyway, I think my plan is to...- Install the reg/rect and measure my stock DC output with a 20 watt load.
- Dewind the stock coil, note the wire gauge, length and number of turns.
- Cobble up two or three more laminations and coat them with high temp paint (???suggestions???).
- Rewind with next higher gauge square cross section wire and put on as many turns as I can fit.
- Check air gap and close it up if large.
- Slap it together and recheck output.
- Moto deep into the night.
Posted by: Kramer
Are there any side effects like the possibility of frying your electronics, or ignition? spikes -yikes!
Is computer modeling a practical way to test configurations?? might be cheaper in the long run.
If there's risk of damaging the ignition, etc, maybe you could test it mounted to an elec. motor
------------------
Kramer
member BRC, AMA
98 WR400 Dualsport!
Posted by: techman
Mace, if you want to be scientific, put your load on and also stick an ammeter in series with the dc shunt regulator. At low rpm when your lights dim, there should be no current through the regulator (or very little, just little voltage peaks worth). As you rev it up, current should start flowing. That's wasted output current.
If you really want to know what's going on before and after, disconnect the power from the bike (if possible) or alternately tap right onto the coil output (take out your headlight and tail light bulbs). Now get a big power resistor from a junk shop, or use a big old dc motor and you can measure output at specific rpm'swithout the shunting of the regulator. You'll see how big the V is at certain size load currents, at certain bike rpm's. If you use the dc motor, add a series resistor to damp the motor inductance (it might start ringing with the inductance in the bike stator). WARNING: what can happen here is you can start pulling big loads out of the coil. Try not to go much past the rated wattage in your loading. But, in another way of thinking about it, the wattage is based on the current capability of the wire. So, restated, try not to go much past the rated current (overcurrent = heated, fried stator). What happens here is that the output voltage climbs (if not shunt regulated) and if the load is basicly a resistor, the load currnt drawn climbs too. So here's the oddity: in a zener system, that poor zener sucks up all that extra current and just dissipates 12vxshunt current. A switching regulator, I think, disconnects the coil from the load (?) and doesn't dissipate, and incandescent bulbs just keep on glowing during the off time (PWM) and the battery if present just charges in a PWM fashion too. So you cold get the real story with a scope, to see if you have a zener regulator or a switching one, and see all the V's and I's before and after the big rewind exercise.
Your planned measure V (and I?) with a 20W load and a reg will be influenced a lot by the action of the reg (ie switching or shunt currents involved).
Your best source of laminations is another dead stator - all precut, formed , perfect, right material, coated etc.
On one pole, I'd try a test wind with the new thicker wire and see how close you can get to the original number of turns. Also pay attention to the direction on each pole and connection order - otherwise they'll fight each other and nullify everything. I wouldn't try to close the air gap - things move and bend, and the flux vs air gap is extremely non-linear, like 1/x.
If you've got time, try just the lamination increase first with the same size round wire - a practice run before you do it for real with the good wire. You should see V increased at a specific rev and load by the ratio of pole cross section areas (new bigger vs original), or conversely the same V at that proportionally lower rpm. Then if your #turns drops by a miracle by the same ratio when you upsize the wire, you'll be back to the same V at rpm as the original, but with higher ultimate current capability.
They are MR16 Home Despot bulbs, with the glass cover. There's lots of mfrs and I'm not sure who's better yet. One day I'm going to have to figure out all the above for my E-line coil. Or just go riding http://dirtrider.net/ubb2/wink.gif Maybe when I fix my battery pack I'll get the instruments out.
------------------
Text Version Home
vBulletin Copyright ©2000 - 2009,
Jelsoft Enterprises Limited.
vB Easy Archive Final ©2000 - 2009
- Created by Stefan "Xenon" Kaeser
