Electric Motor Secrets

[Harvey]

Hi Robert,

I have to think on this a bit - things don't come to my mind like they did when I was younger

Let's see - #3 is inversely positioned when #1 is active. However, D1 does not let that side drop much below ground so essentially we light the lamp in reverse and inversely charge the cap that is across #3 with any transformer action there.

When #2 is firing, #1 is isolated because of D2. Even though there is some transformer action there, the upper side is clamped to your B+ line through A1 and the lower side is floating even though it may drive low. I doubt that any spatial (RF or the like) transference of energy exists there.

When #3 fires we have the condition mentioned in my previous post. So any transformer action on #2 is simply additive to the conduction experienced later in the cycle. So its a bit like #1 above during the first part of the cycle and then later conducting if the low side is not too low and the capacitor bank is drained enough to allow conduction to occur.

So I don't think the large rotor is having any real adverse effect - if I'm looking at it right

It's very interesting to consider the magnet flux and timing involved

Are all the MOSFET gate signals similar or can you change the length and timing there?

[Robert49]

hi Harvey

I will modify the circuit soon. I have some high speed solid state relays I can try.

There is probably something else to take into account.
Because of the size of the rotor, at all times there is a closed magnetic circuit between two of the coils which is when #1 is active there is a transformer effect with #3 and when #2 is active the transformer effect is with #1 and when #3 is active the effect is with #2...
That energy must be going through the diodes or cancelling the current from the power supply?????????????
What do you think?

Robert

[Harvey]

Thank you Robert

I know it is difficult to get the traces to light up sufficiently and still maintain a crisp defined signal especially when it is varying consistently. You have a nice shot there to show the voltage differentials. Unfortunately, I am unable to determine the precise timing and the vertical transitions are difficult to ascertain. One thing is certain however in both shots, you have 6 total firings every 5 divisions (of course we are only seeing two of the three MOSFETS) so if we had your time-base settings we could determine the RPM etc.

What I was looking for was a lengthening of the Q1(upper left) signal compared to the Q3 (lower center) as this could account for some of the increased wattage. This is because the actual on-times could differ for the two devices but the power over time could be the same even though the voltage and current is different when differentiated.

After reviewing the process involved, I believe part of the increase could be via the diodes and stray flow through Coils 1 & 2 during the Coil 3 firing. In order for this to happen, the capacitor bank drops sufficiently to allow the two diodes to conduct - so it would occur at the latter part of the Coil 3 cycle. In this situation Q1 and Q2 are both off, but the two coils will conduct in parallel through the diodes (D1 & D2 *) once the voltage is sufficiently dropped in the capacitor bank. This increases the A1 thru A3 reading but because it is a fractional part of the waveform the meters may not be differentiating the signal as well as we may like. Whatever the case, this is an undesirable result because all three coils are active simultaneously.

There are ways to keep this from happening but it would require some changes in the circuit - some means to prevent Q3 from turning on if D1 or D2 are conductive. (* I've assumed D3 is at the lamp)

I love the simplicity of the design

[Robert49]

Hi Harvey.

Thank you for the good comment.

I seem to double the power of my motor. I intend to try driving a car alternator with it and see how it goes.
I'm retiring at the end of this year (65) so next year is full time on my projects.

Robert

[Harvey]

Hi Robert,

As always I applaud your craftsmanship

The readings you were able to take look quite promising - but I am interested in the actual timing involved. Do you have a scope?

From what I can see in a cursory look, you have done a good job in capturing the energy and putting it to work in the third coil motor.

Keep up the good work!

[Robert49]

Hi gyula.

The cap bank value is 60 micro-farads at 2250 volts.

Since the strength of a magnetic field is determined by amperes/turns, I will try a bigger capacity on the cap bank which should reduce the voltage on A-3 but raise the amps at the same time. If I'm right the motor should be stronger.

The cap on A-3 is 220uf 450v and is there just to even the spikes.

Thanks again

Robert49

[gyula]

Hi Robert,

Well, I did not study the motor pictures, I watched only your recent schematic where coil 1 and 2 is shown and read your description on the third coil so I answered accordingly, sorry.

Now you seem to have some extra output for sure. I wonder what is the capacitor value for your cap bank? Thinking on a cheap DC/DC converter for a possible looping but other than a PC power supply is not cheap and a computer supply is still not good without modifications as per its standard output voltages. (Nevertheless a PC supply is able to run from a 190V DC input instead of an AC mains.)
Or you may drive a motor as a generator (or a normal generator) by this motor's shaft and add up the two outputs.

Wish you good luck, very interesting setup!

Gyula

[Robert49]

Hi Gyula

I can see where you're getting at with the timing but if you look at the pictures on post 1536, I would have to build a new motor to do that.

The stator has three sets of four coils and each coil in a set is 90 degrees from the other arranged north-south-north-south.
The timing wheel has four slots and activates the opto sensors for each set in a 1-2-3 manner which gives 12 pulses per rotation. The motor was built to run this way.

Now when I load the motor as to reduce the speed to half, the amps on A-1 go up to 1.1a and on A-2 to 350ma. On A-3 the amps go up to 800ma and the voltage drops to 189 volts. No apparent change on the light bulb.

So input watts on A-1 =143 watts
A-3 = 151.2 watts
The motor feels strong.

Robert

[gyula]

Hi Robert,

Yes, I did not consider the lack of the lamp load may cause damage to the lower MOSFET, sorry.
Would like to ask about the timing of the 3 switches. What I would do as timing is that I would switch on coil 1 and coil 2 in a push-pull fashion, with lower than 50% duty cycle for both and regarding coil 3 I would switch it ON when neither of the coil 1 and 2 are ON (this latter requires a less than 50% duty for the two motor coils).

Sequence 1:

coil 1 is ON with 30% duty, coil 2 and coil 3 are OFF

Sequence 2:

coil 1 is OFF, coil 3 is ON with 30% duty, coil 2 is still OFF

Sequence 3:

coil 1 is OFF, coil 2 is ON with 30% duty, coil 3 is OFF

Sequence 4:

Coil 1 is OFF, coil 2 is OFF, coil 3 is ON with 30% duty

Now Sequence 1 can be started again and so on. How are your switching sequences done, I wonder.

Regarding the load on the shaft and the voltage goes down on A3 is okay but the amps rise is interesting if you mean the lamp current rise? How much current rise is involved approximately, does it mean that the 15W captured energy drops only a few watts? And if you meant the A3 current rise when the shaft is loaded, then it may indicate that coil 3 starts loading the input 130V voltage source? maybe because the load retards the rotor and timing slips a little for the lower MOSFET?
I know it may be difficult to attain the switching sequence I propose, it may need a microcontroller (Arduino or similar).

Thanks, Gyula

[Robert49]

Hi Robert,

Your coil 3 may form a resonant LC tank circuit with cap bank via the lower MOSFET's drain-source (nanoFarad) capacitance and this may explain why the current meter A3 (652mA) shows higher value than A2 (250mA): the circulating resonant current inside a resonant LC circuit is (loaded Q) times higher than the current flowing into the LC circuit. Of course this does not explain everything, I do not mean it.
Have you checked the input current (A1=850mA) increases when you unconnect and then connect the lamp to the third set? Also, how the input current changes when you mechanically try to load the motor shaft?
Interesting circuit for sure, thanks for showing it.

Gyula

Hi Gyula.
Thanks for answering my call.
There is about 15 watts on the light bulb when A1 is at 130v input.
I can't disconnect the light bulb. There has to be a load there otherwise the voltage rises and burns my 950volt transistor.

When I load the motor shaft, the amps on A1 rise but on A3 the voltage goes down and the amps rise.

I will make more tests and report.

Robert

[gyula]

Hi Robert,

Your coil 3 may form a resonant LC tank circuit with cap bank via the lower MOSFET's drain-source (nanoFarad) capacitance and this may explain why the current meter A3 (652mA) shows higher value than A2 (250mA): the circulating resonant current inside a resonant LC circuit is (loaded Q) times higher than the current flowing into the LC circuit. Of course this does not explain everything, I do not mean it.
Have you checked the input current (A1=850mA) increases when you unconnect and then connect the lamp to the third set? Also, how the input current changes when you mechanically try to load the motor shaft?
Interesting circuit for sure, thanks for showing it.

Gyula

[Robert49]

It was worth the wait, congrats! So it works!

Hi.
Looks like it does.
I tested again and again and each time came up with positive results.
At some point I had 7 meters connected to make sure.
The thing is I don't know why it works, I did it out of inspiration rather than knowledge. I think the geometry of the motor and the timing is probably the key.

Anybody has an idea please let me know.

Robert

[Blargus]

It was worth the wait, congrats! So it works!

[Robert49]

Hi guys.

Here's a drawing of my setup.
I'm driving the motor with two sets of coils, capturing the radiant spikes and driving the third set of coils with that energy.
I tested it with different voltages but the higher the input, the higher is the gain.
Example:
Point A-1 : 130 volts at 850 milliamps = 110.5watts
Point A-2 : ??? volts at 250 milliamps
Point A-3 : 199 volts at 652 milliamps = 129.75 watts
See post #1552 for diagram.

Robert

[Robert49]

Hi guys.

I haven't posted for a while, took a break!

I have reconfigured the circuit to better use the radiant pulses coming out of the coils.
I am now running on two sets of coils , collecting the spikes in a cap bank and running the third set on that energy. The spikes from this set are going to another cap bank and lighting a bulb. At first glance, higher speed, higher torque with less power in.

I will post the results soon.

Robert