Electric Motor Secrets

[Peter Lindemann]

Test This

Peter, the resistor values of the optotrigger bridge are already 500 Ohms. The datasheet of the 2N3055 NPN transistors can be seen here:
2N3055 pdf, 2N3055 description, 2N3055 datasheets, 2N3055 view ::: ALLDATASHEET :::
It says that these transistors are rated to 15A max collector current and max total heat dissipation of 115w. The diodes are rated 3A 1000v ultrafast. Here is the datasheet for these diodes:
UF5408 pdf, UF5408 description, UF5408 datasheets, UF5408 view ::: ALLDATASHEET :::
I just tried to run the motor with 12V on the primary side. I also put a 12v battery on the output. If I run the motor with only one transistor, ir can barley turn itself and is drawing about 0.4A, but the current is jumping with each impulse. If I run the motor with two transistors at 12V, the motor runs at about 40RPM (very slow) and is drawing a little bit more current. The transistors are slowly getting hot. I doubt that replacing the third transistor will give much improvement if all three transistors will run on 12v.
Why cant we use the previous circuit where the output was not isolated from the rest of the circuit? When using the previous circuit I did not notice any heat in the resistors, also the speed was good and with the new coil lenght, the speed should get better too.
Thanks,
Jetijs

Jetijs,

If you think that the new isolated output section is the problem, then hook the system up like before. Leave the three output windings completely isolated from each other and the rest of the circuit. Collect the collapsing field off the three transistor collectors again with your diodes and connect the batteries as before.

Let me know how this works.

You should still be heat sinking your transistors.

Also, you did not read what I said about the 500 ohm resistor. You have a 1K resistor on the LED side of the opto unit. To run the motor on 12 volts you will probably need to reduce this to a 500 ohm to get the same level of light across the opto device. Ohm's Law is still operating, the last time I checked.

Please slow down and check everything. I am 8,000 miles away from you and can only HELP you. In the end, this is your motor and you are building it and testing it.

Peter

[Jetijs]

Peter, the resistor values of the optotrigger bridge are already 500 Ohms. The datasheet of the 2N3055 NPN transistors can be seen here:
2N3055 pdf, 2N3055 description, 2N3055 datasheets, 2N3055 view ::: ALLDATASHEET :::
It says that these transistors are rated to 15A max collector current and max total heat dissipation of 115w. The diodes are rated 3A 1000v ultrafast. Here is the datasheet for these diodes:
UF5408 pdf, UF5408 description, UF5408 datasheets, UF5408 view ::: ALLDATASHEET :::
I just tried to run the motor with 12V on the primary side. I also put a 12v battery on the output. If I run the motor with only one transistor, ir can barley turn itself and is drawing about 0.4A, but the current is jumping with each impulse. If I run the motor with two transistors at 12V, the motor runs at about 40RPM (very slow) and is drawing a little bit more current. The transistors are slowly getting hot. I doubt that replacing the third transistor will give much improvement if all three transistors will run on 12v.
Why cant we use the previous circuit where the output was not isolated from the rest of the circuit? When using the previous circuit I did not notice any heat in the resistors, also the speed was good and with the new coil lenght, the speed should get better too.

Edit: Just replaced the burned out transistor. With all three coil strands/transistors I can get about one revolution per second. The amp draw is 0.65A at 12V.


Edit 2: Today I received a spool of 1.18mm diameter copper wire I ordered. Maybe I can wind three strands and 60 turns with this thicker wire and use the previous circuit?
Thanks,
Jetijs

[Peter Lindemann]

Transistor specs

Peter, I hooked everything up. First I tried to run the motor on only one transistor and power coil strand. The motor ran very slow, about 60-70 RPM and the current draw was 0.8A. Then I tried to run the motor with two transistors and power coil strands. This time the motor reached 2000RPM and more, drawing 1.6A. I figured that the first transistor was dead because it could not be that with one coil I have only 70 RPM but adding the second coil made such an improvement. So I tried to run the motor again with only one transistor, but this time I used the second transistor. Also this time I got only 70-80 RPM and the current draw was 0.8A. I left it running for a minute or so and then the transistor blew. I touched ir and almost burned my fingers, it was HOT. So I disconnected it and hooked up the two remaining transistors and two power coils. I started the motor and it went on fast and the RPMs increased slowly. Also I checked the transistors for temperature. They started to get hot very fast, so I stopped the motor, because I did not want to burn them out. But still I could see that the charging battery on the output went up from 24.9V to 25.1V in that short time the motor was running at high speed. I did not notice any neon flash at all in the whole operating time, even when the transistor blew. What could be the problem?

Jetijs,

Don't forget, the peak current pulses on the transistors are now approaching 7 amps. Are your transistors RATED for this current and mounted on proper heat sinks? If not, that is part of the problem. If the neon lights are not flashing and the transistors are hot, then it was a THERMAL failure of the transistor junction from too much current and/or not enough cooling. We are starting to push the transistors out of their SOA (safe operating area) so other precautions must be made. The transistors can run all day at 50C, which is too hot to touch, but they must be mounted on heat sinks.

One thing you can do is run the system on 12 volts until you get these other problems worked out. The inductive rise-time is no longer a problem, but now the current is on too long for the transistors and is heating them up.

Also, what diodes are you using in the output section?

Please replace the blown transistor and run the unit with all three coils on 12 volts. To do this, you may have to lower the resistor in the LED section of the opto device to 500 ohms. Hopefully we can leave all other resistors as is.

Peter

[Aaron]

rotor/torque relationship

Elias,

It looks like a certain amount of work by the coils attracting the rotor will be expended over a longer distance (time) on a larger rotor. So basically, the rotational distance is reduced by a smaller diameter rotor. So for equal relationship to the coils, it is compressing the work into a smaller period of time translating to more torque. To my comprehension at least.

What about the relationship of torque which is aided by a flywheel effect from a larger and/or heavier rotor (of course larger doesn't necessarily mean heavier depending on material but lets assume identical materials like aluminum)? Would that be an equal, more, or less trade off?

Peter, if this isn't enough on topic, we can make a new thread for rotor size / torque relationships. That could be an entire topic of itself and would be applicable to all motor types I would imagine.

[Peter Lindemann]

Please Relax!!!

Hello

I am trying to point out something very important for increasing motor torque in attraction motors, but nobody seems to be interested in it so I decided to perform some primitive calculations to verify my claim and I found interesting results, very interesting. I performed these calculations at a position where the rotor is halfway slided across the stator, and to obtain more complete results, the value of the mean effective force must be found by integrating it all over the period of the "on state" of the coil. But at least this may be a crude estimation for the mean value of F_effective along its attraction cycle.

I calculated the effective force exerted on the Iron rotor to produce torque in two rotor configurations: The first rotor resembled Jetijs' rotor, which is about 5 units long (each unit is the width of the stator) and the second one only 2 units long.

Consider these:
- alpha is the amount of displacement of the rotor from its alignment position with the stator in radians.
- F_pull = beta * F_slide
- F_effective = F_pull * sin(alpha) + F_slide * cos(alpha)
- F_slide = F
Here is what I found out by simple math:

In the first rotor size (5 units of length):
sin(alpha) = 0.2
F_effective = (0.97 + 0.2*beta)F
beta = 1 then F_effective = 1.17F
beta = 2 then F_effective = 1.37F
beta = 3 then F_effective = 1.57F

In the second rotor size (2 units of length):
sin(alpha) = 0.5
F_effective = (0.85 + 0.5*beta)F
beta = 1 then F_effective = 1.35F
beta = 2 then F_effective = 1.85F
beta = 3 then F_effective = 2.35F

As it is demonstrated by these calculations, by increasing beta (by decreasing the airgap) the effective force increases much more in the second geometry.

beta = 1 then F_effective_2/F_effective_1 = 1.15
beta = 2 then F_effective_2/F_effective_1 = 1.37
beta = 3 then F_effective_2/F_effective_1 = 1.50

I think that the value of beta can be about 3 with a small airgap and thus changing the rotor length from 5 units to 2 units can result in 1.5 times increase in the effective force.

The most interesting observation was that even when beta = 1 which means that F_pull = F_slide, a slight increase in the effective force was resulted (1.15 times).

I hope that this makes someone interested !
Or at least corrects me if I am in error.
I have attached the calculations.


Elias

Elias,

Thank you for running these experiments and reporting your results. Your data supports what has already been said dozens of times on this thread. We know that decreasing the air-gap increases the magnetic attraction. We also know that ANY changes to the geometry that increase the tangential vector of the magnetic forces on the rotor increases the torque as well. Making the rotor piece smaller in diameter is one way to accomplish this, and there are others.

You are always welcome to ADD what you know to this thread if it is ON TOPIC. Your air-gap and geometry test data is welcome. Thanks for running the tests and bringing to our attention again.

Peter

[Jetijs]

Peter, I hooked everything up. First I tried to run the motor on only one transistor and power coil strand. The motor ran very slow, about 60-70 RPM and the current draw was 0.8A. Then I tried to run the motor with two transistors and power coil strands. This time the motor reached 2000RPM and more, drawing 1.6A. I figured that the first transistor was dead because it could not be that with one coil I have only 70 RPM but adding the second coil made such an improvement. So I tried to run the motor again with only one transistor, but this time I used the second transistor. Also this time I got only 70-80 RPM and the current draw was 0.8A. I left it running for a minute or so and then the transistor blew. I touched ir and almost burned my fingers, it was HOT. So I disconnected it and hooked up the two remaining transistors and two power coils. I started the motor and it went on fast and the RPMs increased slowly. Also I checked the transistors for temperature. They started to get hot very fast, so I stopped the motor, because I did not want to burn them out. But still I could see that the charging battery on the output went up from 24.9V to 25.1V in that short time the motor was running at high speed. I did not notice any neon flash at all in the whole operating time, even when the transistor blew. What could be the problem?

[elias]

No one seems to be interested in what I am trying to point out, so ...

Hello

I am trying to point out something very important for increasing motor torque in attraction motors, but nobody seems to be interested in it so I decided to perform some primitive calculations to verify my claim and I found interesting results, very interesting. I performed these calculations at a position where the rotor is halfway slided across the stator, and to obtain more complete results, the value of the mean effective force must be found by integrating it all over the period of the "on state" of the coil. But at least this may be a crude estimation for the mean value of F_effective along its attraction cycle.

I calculated the effective force exerted on the Iron rotor to produce torque in two rotor configurations: The first rotor resembled Jetijs' rotor, which is about 5 units long (each unit is the width of the stator) and the second one only 2 units long.

Consider these:
- alpha is the amount of displacement of the rotor from its alignment position with the stator in radians.
- F_pull = beta * F_slide
- F_effective = F_pull * sin(alpha) + F_slide * cos(alpha)
- F_slide = F
Here is what I found out by simple math:

In the first rotor size (5 units of length):
sin(alpha) = 0.2
F_effective = (0.97 + 0.2*beta)F
beta = 1 then F_effective = 1.17F
beta = 2 then F_effective = 1.37F
beta = 3 then F_effective = 1.57F

In the second rotor size (2 units of length):
sin(alpha) = 0.5
F_effective = (0.85 + 0.5*beta)F
beta = 1 then F_effective = 1.35F
beta = 2 then F_effective = 1.85F
beta = 3 then F_effective = 2.35F

As it is demonstrated by these calculations, by increasing beta (by decreasing the airgap) the effective force increases much more in the second geometry.

beta = 1 then F_effective_2/F_effective_1 = 1.15
beta = 2 then F_effective_2/F_effective_1 = 1.37
beta = 3 then F_effective_2/F_effective_1 = 1.50

I think that the value of beta can be about 3 with a small airgap and thus changing the rotor length from 5 units to 2 units can result in 1.5 times increase in the effective force.

The most interesting observation was that even when beta = 1 which means that F_pull = F_slide, a slight increase in the effective force was resulted (1.15 times).

I hope that this makes someone interested !
Or at least corrects me if I am in error.
I have attached the calculations.


Elias

[Peter Lindemann]

Correct!

I was acting too fast and did not see your edited post. I edited the diagram, here it is:

Is it right now?

Jetijs,

That looks right now. Wire it up and see what it does.

Great work!!!

Peter

[Jetijs]

I was acting too fast and did not see your edited post. I edited the diagram, here it is:

Is it right now?

[Peter Lindemann]

Almost..

Yes, I lowered the resistor value to 500 Ohm in the phototrigger voltage divider.
There are no neon bulbs right now acros the transistor emmiter and collector. I will add them now. As for the direction of the electricity, I put two small dots on the core in the diagramm, the dots are showing the top of the power and output coils. So If all coils are wound in the same direction should it now look like this?

Jetijs,

Did you see the last edit of my previous post?

Peter

[Jetijs]

Yes, I lowered the resistor value to 500 Ohm in the phototrigger voltage divider.
There are no neon bulbs right now acros the transistor emmiter and collector. I will add them now. As for the direction of the electricity, I put two small dots on the core in the diagramm, the dots are showing the top of the power and output coils. So If all coils are wound in the same direction should it now look like this?

[Peter Lindemann]

Minor Modifications

Peter, did you mean the circuit should look like this?


I understand about the wire size, we basically want to cut the lenght by 75% (That would be aprox 10 metres of wire) and use 6 strands of twisted wire.
Thanks,
Jetijs

Jetijs,

Not quite right. Please note that the output coils will have the OPPOSITE polarity of what you have drawn. When the field collapses, the magnetic flux lines cut the wires in the opposite direction as when the electricity is applied. So, if the top of the power coils is POSITIVE, then the top of the output coils is the NEGATIVE and the bottom is the POSITIVE (opposite polarity as the power coils). Also, the NEGATIVE of the output coils goes to the NEGATIVE of the second battery, not the negative of the first battery. Also, the second battery is now ISOLATED and not connected to the positive of the first battery.

Also, did you ever lower the resistor values in the voltage divider run by the photo-transistor? Are they still 2K?

And finally, are you protecting your 2N3055's with neon lights across the emitter-collector junction?

Everything else looks good!

Peter

[Jetijs]

Edit: I wound all the wire from the startor core and folded it twice so I got exactly 1/4 of the lenght. I wound 6 strands on a spool and put them aside and I twisted the remaining six strands with a hand drill. Then I wound this thick 6 strand wire on the startor core and the wire lenght was just long enough for 59 turns (almost 60 as you predicted) Here's a picture:


Now I only need you to verify I have drawed the circuit diagram correcltly

[Jetijs]

Peter, did you mean the circuit should look like this?


I understand about the wire size, we basically want to cut the lenght by 75% (That would be aprox 10 metres of wire) and use 6 strands of twisted wire.
Thanks,
Jetijs

[elias]

I started a new thread,

Peter,

I started a new thread as you asked (http://www.energeticforum.com/renewa...html#post12929) and deleted my previous posts, but I put the portion of it that in fact is closely related for producing excess mechanical torque in attraction motors here:

http://www.energeticforum.com/12844-post358.html

Any comments on this?