Confirming the Delayed Lenz Effect

[Farmhand]

I would really like to get a reliable measure or at least a "gauge" of the shaft power before changing test arrangements. Does anyone have an idea I can implement fairly easily to get a benchmark ? I don't need a HP or torque value just a reliable benchmark to gauge against. I'm thinking I'll just hang a given weight from a leather belt over the polished shaft like a friction brake and it should be equal if done the same then I can add weights to load it down and see the most weight it can tolerate hanging over the shaft from a fixed anchor point something like that. But as it heats up the friction value will change. Seems a tricky thing to do reliably but I'll see what I can do. Maybe a turnbuckle and a digital hanging scale for weighing chickens. 

Cheers

P.S Scratch that, I have a simple test load, I found a small KV 1880 - 3 phase model airplane motor, an "outrunner", the outside is what turns with the magnets on it, so I can just put the belt around the outside of the outrunner motor, the pulse motor can hold over 1800 rpm with a 1 Ohm load on one phase (between two wires), I'll try it with two 1 Ohm resistors I guess. If I short a winding it throws the belt immediately. 10 Ohms only loads it a bit. It take a fair bit to turn the three phase motor and the voltage is low but the frequency is relatively high, it'll do for a test load it should remain a stable measure. It produces one volt through a one ohm load so that's neat. It can accelerate under the one Ohm load if kept in the torque range. If I de-tune it too much while the 1 Ohm load is on it suddenly drops speed beyond recovery without removing the load.

Oh and the wave form remains a stable sine wave so measurements are even possible. An improvement in Watts can be found then.



..

[Farmhand]

OK so I shot a video of a 1 Ohm draw down test on the three phase motor as a generator. As It turns out if I switch in cap to double the size of C2 I can drive the generator with two 1 Ohm resistors between the windings but only at 1200 rpm. 

http://www.youtube.com/watch?v=MAEgLr5zXK4

Also I redone the schematic to show the boost converter hardware and the way I intend to employ a second motor coil with a bifilar wound charging coil and switch them in parallel with separate switches. Everything is renamed and I think C3 and C4 could be 220 uF instead of 330 uF. With 50 volts max on the boost capacitor I should be getting nearly 80 volts to switch the coils with at times.  It'll bang along then and really slap the table. 

Cheers

[SeaMonkey]

I like the resonant charging feature of your
setup - it can provide an additional "boost"
to your supply voltages.  Shouldn't there be
a couple of "check valve" isolating diodes in
the feeds to capacitors C3 and C4?

Very nicely done!

[Farmhand]

Hi SeaMonkey, Nice to read you,  Yes you're right, the "de-q-ing" diodes I left out for the sake of less parts, but they are kind of important, they would go between the Boost cap and the MC2 charging coils, or they could go after the inductors but I would put them before, the point is to check (stop) the reverse flow of energy from the two charging caps C3 and C4 back to the supply which is the boost cap C2. However the Q is quite low because of the low frequency, still they should be there.

I made an improvement, for some reason the charging coil works best reverse polarity but placed slightly before the "S" magnet after the one over the motor coil rather than before the "S" magnet that just passed the coil, this makes the coils close together with the inside edges of the cores tips only about 1 inch apart, but Wow what an improvement in speed and torque to how I had it for the load down video, now it can hold at over 2000 rpm with a 1 Ohm load on one phase of the generator and 10 ohms on another phase. Should I make a triangle of load resistors for each value ? 2 ohms is not much load for a 1.3 volts generator but 1 ohm is and a short circuit almost actually stops the generator, it throws the belt, now the belt is slipping when I apply two 1 ohm loads at high speed so I have to make it a tad tighter or increase the pulley friction some other way. Placing the charging coil close to one motor coil doesn't sit well with me so I might go silly with the glue gun again. But anyway, I now know where it likes to go and it means the motor will be able to rotate in only one direction. It wont run well backwards.

The charging coil needs to be reverse polarity and in an advanced position to a 45 degree "S" magnet, I need to look at the currents in the coils again, I setup charging caps so I can look at the differences in the currents when I go from 220 uF to 440 uF on the charging capacitance, using 440 uF the torque at lower speeds is better but it won't run fast, I'm almost convinced there needs to be two charging caps so a larger capacitance can be switched in by the micro if the rpm is below a certain value that would improve start up speed, lower rpm torque and power throughput. Doubling it is too much but it'll make seeing the difference easier.

After the latest coil position adjustment the motor can now reach 2800 rpm at the peak of its input power curve but all the torque is at about 2000 rpm now rather than the 1800 rpm before and with the same input pulse width. I think i could draw the curves fairly close after the loading experiments, it really helps to load the motor down with a real case load, the 1 Ohm resistor it can work into gets a bit hot but it's only a 0.5 Watt resistor. 

I'm almost setup now for the second motor coil. I've got several things I want to do with respect to my own generator coils for the rotor which will be much less lossy than driving a generator especially if the cores slide in and out.  My rotor design means that the cores cannot "catch" on the rotor magnets, worst they can do is to scratch the outer edge of the rotor so they can be made to go right up to the rotor and be shaped for the curve, I intend them to have very big cores about 1.5 inches diameter so the core tips will be curve shaped I might even double the thickness of the rotor and get another 8 magnets to make it better that way. then it would be a 24 mm thick rotor, it really needs to be thrice as wide.    I'm wondering if the motor coils would benefit from larger diameter cores. The motor coil core is fairly small and the coil gets hot after a while.

Cheers

P.S. Now since there are 8 magnets and 4 are souths MC2 can be placed in 4 locations on the rotor in an advanced position of about 5 to 10 degrees before any of the "S" magnets.

SeaMonkey do you think it could be possible the motor coil is handing off some flux directly to the charging coil at some point ? Or maybe sharing cores or something ?

[TinselKoala]

@Farmhand: Very Nice! You are doing a great job of documenting and developing your system. Keep going!

You probably realize that your torque measurement system can provide a known variable resisting drag by externally energizing one of the "extra" windings on the brushless outrunner with some variable DC. You've made a simple dyno system; I wonder if it's possible to calibrate it in standard units somehow.