Complete information on working SM style device.

[orbs]

I do not understand why people want to make things so complicated.

One 74HC4017 and one OR gate looked simpler to me than several integrated circuits, at least for an initial testing...

One uses a shift register where all stages are clocked synchronously.  Nothing else.

OK, sound advice. Building on the rapid fire rat race controller, one could use a 74AHC164 with a 74VHC27 and 74AHC377 to generate same pulses as with the 4017 but with the nice symmetric delay properties of the 377. Unfortunately, 377 available as SMD only.
[Details: connect Q1..Q6 of the 164 to two NOR gates of the 27, and their output to the 2 data inputs of the 164. This creates a self-initializing shift register that rotates a 2-cycle pulse with a period of 9 cycles. Connect Q0/Q3/Q6 to the third 27 NOR and its output to E of the 377. Then connect Q1/Q4/Q7 to the data inputs of the 377. This yields the output signals for the first 3 coils. The 4th coil signal is connecting E of the 377 to one of its data inputs and the output again to an input, i.e. twice in sequence, to get the pulse delayed/shortened the same as the others. The 164 and the 377 use the same input clock signal.]

For precise timing, one does not use counters and state decoders, nor microcontrollers, nor DSP.

Well, if one does use an embedded controller already, then it looks like an AD9510 clock divider is pretty good. It has 4 identical outputs (twice) that can be programmed to generate the signals for the 4 coils (one divide-by-3 with 33% duty cycle and three divide-by-9 with 11% duty cycle and 0/3/6 cycle phase shifts).

For a clock generator, the AD9912 DDS looks pretty nice, although somewhat expensive (some $70). However, it has some nice properties that make its use convenient: very high frequency output resolution and built-in comparator to generate square wave output. And the interesting part is, it is used by HAM amateur radio operators, so there are already people out there who have some experience with it. While it looks like overkill to have a device designed to generate high quality sine output to then just make it into a square wave signal, it allows very fine frequency adjustments that make it simple and flexible to use (see also AN-823 and AN-837).

Apparently it's better to use an external main clock source like the Epson Toyocom EG-2102CA PECL instead of the internal PLL of above devices. Below are some pictures taken by radio operator DL7IY [dl7iy(at)darc.de] from his article about the AD9912 in FA 3/08 (German magazine Funkamateur).

[Note that there are hyperlinks in above text that become visible when you mouse over them or use the Tab key.]

[Earl]

I do not understand why people want to make things so complicated.

One 74HC4017 and one OR gate looked simpler to me than several integrated circuits, at least for an initial testing...

One uses a shift register where all stages are clocked synchronously.  Nothing else.

OK, sound advice.

Danke fuer Dein Post.  FA ist nicht schlecht, habe aber kein Abo.

I am real busy at the moment, don't really have time for a reply.

RE AD DSP, those little things are very difficult to solder.

Viel Glueck beim basteln and c u soon.  73, Earl

[poynt99]

SM Quote:

- The control wiring is vertically wound in several segments around each of the horizontal collector coils.

spherics made no mention of any horizontal coil other than a toroid (collector) coil (which has vertical windings) centered between the xyz and A coils. He said tao's depiction was correct.

spherics is saying that the collector and control (kicker) coils are reversed comparing his and SM's, and that the kicker coils are bifilar solenoids oriented inward and not part of a toroid config.



Is this thread dead? or am I just way behind everyone else?

[omnispace]

spherics is saying that the collector and control (kicker) coils are reversed comparing his and SM's, and that the kicker coils are bifilar solenoids oriented inward and not part of a toroid config.

The bifilar coil was for an experiment meant to demonstrate a concept.  I do not think it was meant to be a part of Spheric's TPU.  The positions of the 4 coils, and the interaction of the pulses provide the same effect as the bifilar experiment, but in a controlled manner where the energy can be collected.

Basically, as I see it, the top coil pairs up with each of the 3 control coils and takes the place of the bifilar windings.

[mfred68]

Hi everyone, this is my first post, i have just joined as i am also seeking the golden information of the SM TPU, as i want to build one to power my home, but after many days of reading 100s of posts about it on here, it all seems to boil down to coils getting "kicks" at the right phase and frequency, and everyone seems to be talking about making the nessesary frequency generators to drive the coils, i even read somewhere that SM explained that tubes (valves) work best, but one thing i'm still confused about (which i'm sure many of you are too) is that by watching the SM videos of the 3 different sized TPU's in opperation, none of them appear to have electronic control circuitry connected to them! they really look just like self contained toroids, so where are the tubes SM talked about? and where are the 3 frequency generators with the heat sinked mosfets? and the battery supplying them for start up?

I also just read that Steven Mark is ill and has cancer, dose that mean that if he did pass away, that the secret behind this technology will never be revield?