Agentgates´s TPU setup with strange wavehill hump

[ketone]

 Hi all,         
               
                Haven't been able to start my build yet{materials}....but was wondering if anyone has based their build angle of the primary windings on the golden triangle? It has angles of  36°-72°-72°.

  afaik...agentgates is getting great results with 45 degrees

[agentgates]

@teslaalset

Thanks for the useful advices. I put more caps in parallel just to be sure. BTW I am moving on to decrease the current intake and step a bit higher with voltage as it more behaves as a voltage device rather than a current one. So I placed current limiter resistors in series with the coils.

@Lindsay

Yes we are optimising my old design to make the fact of OU less questionable, more compact and more visual  without using more TPUs. In my original design I used 2 stages of TPUs to boost up the output with LC stages and full bridge rectifiers to keep the field spinning. We are about to repeat it with a single TPU without ugly spikes, inefficient back EMFs and heating problems.

Some major modifications:

1. Gaps between secondary windings to decrease back EMF, power consumption and remove many components intended to smooth the output. With this configuration for one shot all you have to see is a big exponentially rising hump with the same falling edge and 2 or three hard to see long and almost completely flat waves as back EMF oscillation. See the 2 scope shots I sent for comparison, those were taken by the 9 turn secondary coil of mine.

2. More primary coils to distribute heat on the coils and (it is important) to fill the gaps between the pulses as it seems the winding gaps give the pulse delay not the driving frequency. We layed 2 other primary between the already wound one and shoot them in 120 degrees phase out. (I'm currently simplifying the circuitry to make it cheaper to build and more available to all testers)

The correct output signal with one primary at 50% is like this: -__-__-__-__-__.

If you increase or decrease the pulse width you'll get smaller peak, instead of shorter or longer in duration, also the gaps remain stable at any frequencies. This is why the 2 more primary.

Input signals:
---___---___---___---___
_---___---___---___---__
__---___---___---___---_

Output signal:
--------------------------

So you should get a stable HV high current DC with some insignificant waves on it's top without LC.

3. We use thinner primary wires than the original has, as it seems the device likes to spin electrons in thicker conductors in its environment. By using thinner wires we decrease the weight on the IGBTs and we can step below to low power transistors or darlington array or similar.

4. We use thicker on the secondary to get the power collected there rather than on the primary.

[wings]

Hello IceStorm,

Thank you for the advice I am aware of their inaccuracies (I'm a software engineer  ).

UPDATE

Just an idea, if somebody has a stepper motor driver handy can save a lots of time with shift registers and transistors as in functionality they give similar result.

Arduino stepper controller code

http://code.google.com/p/rsteppercontroller/downloads/list
http://code.google.com/p/arduino-m5451-current-driver/w/list
http://code.google.com/p/stepperserialcontrol/downloads/list

http://reprap.org/bin/view/Main/StepperMotor
http://www.arduino.cc/en/Tutorial/StepperBipolar

[agentgates]

@wings
Thanks for the link. I tried the arduino stepper motor library last night but it doesn't really comply my demands. It has inaccurate output and needs feedback to work correctly.

Video in a few minutes with some explanation as I saw many of you had problems with understanding the concept.

(sorry for the delay there is no sound, I'm going to record it again )

[ramset]

Gotta get some popcorn

I used to love the movies


Chet