Confirming the Delayed Lenz Effect

[Farmhand]

OK so my tests had mixed results, taking out the diodes D4 and D6 is good in my set up with the charging circuit, it increases low speed start up torque by prolonging the currents through the coils, but for high speed running close to the pulse width maximum limit it limits the top speed of the motor because of overlapping currents I think.

The other diodes are all necessary, also with my setup because of the charging coil "MC2" then L1 is obsolete, MC2 does the job in my set up that L1 would need to do if there was no charging circuit/coil.

Anyway I went to a rotor with 8 all north out facing magnets, I get a maximum switching frequency of 320 Hz for 2400 rpm, at high speed or under load the input power is much the same as before because with 8 magnets the Pulse Width is much narrower. the start up torque is much better and the magnet spacing now means that my 40 degree charging coil is slightly retarded in timing to push the magnet the main motor coil just pushed and slightly advanced to pull it so I went with a push pull set up by reversing the charging coil polarity/connections so it now makes a south pole and pulls the magnet that the main coil just pushed,
it works really well and I can spin my fan blade at the same speed with 3 Watts less, now I can spin it at 1500 rpm with just under 10 Watts while getting 300 mA charge current into another battery. With a few hundred Hz the charge output is becoming useful at 80 Hz it doesn't do much.

I just use the battery in series with the cap C3 or in series with the charging coil. It's also really quiet now but it does make a nice hum.   

I think this is the rotor set up I'll do gen tests with.

Cheers.

[MileHigh]

Farmhand:

Perhaps you will wow them at the next pulse motor build off!

Here is a software fantasy build:  You have some Hall sensors monitoring the rotor magnet passes at two positions.  The Hall sensor outputs trigger the programmable timer inputs on the microcontroller.  With software you measure the rotor speed and you can derive the rotor position to know when to fire the main drive coil.  When you fire the main drive coil you also light up some LEDs to get a strobe effect to illuminate vertical white lines painted on the passing rotor magnets.

Then with "left" and "right" keyboard/keypad commands you make fine adjustments to the start and stop angles for firing the coil.  You get visual feedback with the LED strobe.  So you can find the "sweetest spot" for maximum RPM at a given average current consumption by tweaking the drive pulse and monitoring the RPM.  Someone that knows how to code software control systems incorporating feedback loops could write a system so the tweaking of the timing is done automatically.

Here is a cool clip that shows the principle in action, the LED hard drive clock:

http://www.youtube.com/watch?v=K1asNB0te0o

The author:

The code I wrote myself, as I wanted to be able to add new types of displays and patterns to it.  It does not assume the drive spins at a fixed rate, but times each revolution and calculates the delays needed to flash the display correctly.  I use a timer interrupt to decide when to turn on or off the LEDs.  I had used a loop comparing the timer value and it worked, sort of.  Every few seconds it would glitch and the hands would quiver annoyingly.  I was never sure what caused it, but switching to timer based interrupts made it rock solid.  It's fun to slow the disk down with a finger and watch it still maintain the image until the disk goes too slow for the 16 bit timer to handle and starts overflowing.

If you had a rotor with four magnets on it, they will not all trigger the hall sensor at exactly the same relative angle.  Also, it may be difficult to have the four rotor magnets exactly 90 degrees apart from each other.  So what you can do is time each Hall sensor tick to tick with the microcontroller's built in timers.  Then in software you could average the last four tick-to-tick times to get your RPM calculation.  Or you could do a running average of the last 8 timing intervals, etc.  This is a software low-pass filter to get a more stable RPM measurement.

MileHigh

Bonus clip:  http://www.youtube.com/watch?v=Md78sOI1-r8

[Farmhand]

When are these build off's ? What you propose is interesting but I lack the programming skills, and I went with picaxe stuff when I decided to learn a bit, most seem to use Arduino or whatever they are called (spelling). But yes I like to challenge myself and try to learn new things. I haven't learned how to count revs yet even.

I had another brainstorm and made a real big improvement, this coil placement and rotor magnet arrangement is the best by far and I've tried several methods to get more torque and start up power. I've got 8 magnets N-S-N-S ect.  I made the charging coil produce a south pole at the top and placed the charging coil so that it starts out pushing a south magnet face then as the motor speeds up rather than get out of tune "timing wise" it goes from pushing a south to attracting a north, now it's more powerful again, it can now spin my fan blade at over 2000 rpm easily and speeds up pretty quick for a pulse motor. With the fan blade on it can go from 1200 rpm to 2200 rpm in about 15 seconds. With 20 Watt's input it can blow a lot of air. At low speed it's push-push and at high speed it's push-pull.

I'm making miniature generators from shaded pole motors by cutting the frame to separate the poles and drilling the rotor (squirrel cage) and inserting magnets, I tested one last night I made form a small microwave fan motor, it worked well but the wire is too thin, I rewind it. I've got an extra big one with a two inch long squirrel cage I want to convert, it has thick wire.

My motor in the previous arrangement spun the little generator up to 2200 rpm with a belt , up geared. I ran it for a couple of hours experimenting. I got 27 volts RMS out of the little generator.

Anyway I was thinking of ways the motor could self regulate it's own speed. It doesn't need to be timed on the fly now except for start up.

My pulse processing is all done in hardware, I don't need a micro to run it, just a CD4047  the sensor and some other parts.

Cheers

[TinselKoala]

What MH is proposing is relatively trivial to do with an Arduino, a couple of Hall switches (or even just one), and some external transistors to handle the coil currents. I read his post last night and would have done the setup and programming but I don't have any free Hall sensors and I refuse to take apart an already-built motor to scavenge one.

One big advantage of the Arduino over the stand-alone microcontrollers is the programming environment. Also, the ease of interfacing and controlling makes it nice to work with. Also there are tons of help, many examples of code, and lots of people showing different Arduino projects on YT.
The ease of software reconfiguring means that it's easier to experiment with than a stand-alone TTL or CMOS logic chip pulser, once you get the hang of it.

The Hall switch will send a pulse whenever the magnet passes it. You just put the Hall switch's 5 v signal output to an input pin on the Arduino, and the pulseIn() command reads the pulse timing. From this you can derive RPM very simply by knowing the number of magnets on your rotor and the pulse durations. The arduino can be programmed to send the output pulse to a coil with whatever delay or advance and pulse duration is necessary for the RPM, and the advance/retard and pulse duration timings can be sent to the Arduino live, either by two simple potentiometer controls or by keyboard mappings as MH suggests, and the whole thing can be monitored on the serial monitor using the Arduino programming environment or other terminal emulation application of your choice. Or you can use an LCD or touchscreen display very easily with the Arduino.
A little more sophisticated setup would use a ratiometric Hall sensor and then you could even select the precise magnetic field strength at which to send a pulse to the drive coil. This might actually automagically correct timing for RPM.

I'm not a salesman, but I love my Arduinos, they are cheap, easy to program and use, hard to break (I have blown one PWM channel output transistor on my Uno, out of eight...) and can do just about anything in terms of controlling little electronic devices. Or even big ones.

[Farmhand]

Well you sold me on it Tinsel, I'm gonna get one now. Any suggestions on the actual chip or start up kit I should get ? Is there USB interface, or just serial ?

Anyway I think you'll like my next video clip, I got it really thumping, it's kinda deceptive the sound doesn't match the speed kinda thing, but the tachometer jumps in the hundreds of revs. The capacitor C2 has a Sine wave on it that Matches the input pulse just like a Tesla coils wave forms, the pulse is less than 50% but the square side of the drain wave form matches to exactly the down slope of the sine on C2 kind of thing, like my Armstrong oscillator variant's wave forms.

The timing now is now almost automatic over a certain range due to the way the coils are oriented to the magnets but some auto adjustment would be very helpful. I can use my micro to switch the power to the coils off if a certain drain voltage is reached ect. as protections, which can be used to limit speed kind of.

Video will be about 15 minutes, it should be good for a laugh at my hot glue skills when prototyping.  I need to try that coil at a few degrees either side of 45* just to make sure I'm doing the right thing with the main build, it could have two motor coils and a single charging coil with two appropriate windings on it, the charging coil will be between the two motor coils so only the bottom 90 degrees of the rotor would be used for driving, (like a cradle). A 90 degree tripple, or I could make it two sets of coils on each side so two 45 degree N-S twin coil set ups that way. Dunno yet. If the testing shows me 45* is actually the best then it's a go ahead on something. I want some open space on the rotor for experimenting and to keep it upright and bottom heavy, I'll have to set up a rotor balancing device as well.

It's  really a lot of fun. I'll need to get hall's i've never used them, sounds good though.

Cheers

P.S. Here's the second video. http://www.youtube.com/watch?v=aC5Hdp6LSMU

And the circuit as it is now. I cut the circuit explanation part off the video to make two smaller uploads it's uploading now It's the first part.


..