Kapanadze / Dally

[ion]

I stumbled onto this schematic (posted on OUdotcom) and thought it be interesting for discussion.

It is someone named Dally's attempt at Kapanadze replication

It shows the use of the coax as a shorted line that is pulsed with an SCR pulse circuit similar to what Peterae and I have been playing with. I believe Grumpy was also going to test something out with the coax.

I am impressed with the high quality of the schematic and level of detail, especially in the series pass +200 volt regulator circuit, where  frequency compensation (R6 C4) around the TL431 regulator is used.

Main points are:

1) The use of a shorted end coax line  that overlays other windings that is pulsed by SCR T7 operated in avalanche mode. Reflected pulse in the line is clamped with diode KA204.
2) The use of logic chip (K1534na3) with gate delay to drive to the SCR (this does not appear to be an oscillator, rather it seems to be a debounce circuit for S1)
3) The use of ATX power supply 12 V for startup, then keep alive circuit Consisting of L4 and Br3 take over.
4) Coil L1 is driven from the secondary of the push pull oscillator with frequency adjust by R 1k.
5) Coil L2 is resonant with capacitor 50 OM (translate?)
6) An earth ground is used on L1 and L4.
7) Push Pull regulator built around TL494 provides 200 Volt regulated power supply for SCR pulser and square wave drive for L1
8 A 1:2 pulse transformer Tr1 provides impedance match to coax line.

SM seemed to use shorted end transmission line (lampcord) in his open TPU, and may have used this technique in other devices, as lampcord would make a transmission line.

The Spherics simple dual coil test was also a shorted transmission line.

Printed circuit artwork seems to be available for the device.

More info here:

http://www.liveinternet.ru/users/edward_lee/ and on OUdotcom.

Comments welcome.

[PhysicsProf]

  I respect the research work of Itsu, and he has videoed a Dally replication:

http://www.youtube.com/watch?v=MQUciCy9p3A

and 2 days ago:
http://www.youtube.com/watch?v=utkIltw6WrU

Going to these vids will also provide links to similar Dally replication efforts.  I agree with ION that this work is intriguing.

[Hoppy]

I stumbled onto this schematic (posted on OUdotcom) and thought it be interesting for discussion.

It is someone named Dally's attempt at Kapanadze replication

It shows the use of the coax as a shorted line that is pulsed with an SCR pulse circuit similar to what Peterae and I have been playing with. I believe Grumpy was also going to test something out with the coax.

I am impressed with the high quality of the schematic and level of detail, especially in the series pass +200 volt regulator circuit, where  frequency compensation (R6 C4) around the TL431 regulator is used.

Main points are:

1) The use of a shorted end coax line  that overlays other windings that is pulsed by SCR T7 operated in avalanche mode. Reflected pulse in the line is clamped with diode KA204.
2) The use of logic chip (K1534na3) with gate delay to drive to the SCR (this does not appear to be an oscillator, rather it seems to be a debounce circuit for S1)
3) The use of ATX power supply 12 V for startup, then keep alive circuit Consisting of L4 and Br3 take over.
4) Coil L1 is driven from the secondary of the push pull oscillator with frequency adjust by R 1k.
5) Coil L2 is resonant with capacitor 50 OM (translate?)
6) An earth ground is used on L1 and L4.
7) Push Pull regulator built around TL494 provides 200 Volt regulated power supply for SCR pulser and square wave drive for L1
8 A 1:2 pulse transformer Tr1 provides impedance match to coax line.

SM seemed to use shorted end transmission line (lampcord) in his open TPU, and may have used this technique in other devices, as lampcord would make a transmission line.

The Spherics simple dual coil test was also a shorted transmission line.

Printed circuit artwork seems to be available for the device.

More info here:

http://www.liveinternet.ru/users/edward_lee/ and on OUdotcom.

Comments welcome.

There has been two complete schematics posted and the one posted by ION looks to be a development on the other and more sophisticated but its not the circuit applicable to the Daly self-running video on OU.com which can be seen in the following link: -
http://realstrannik.ru/media/kunena/attachments/630/Dally2012.JPG

The components in need of alternatives are: -
Diode KA204 -  step recovery?
IC K1534NA3 - Probably a 4011B Quad 2-I/P Nand
KT315 - Any general purpose NPN?
T7 - 400V or 600v non sensitive gate power thyristor?
DR+R?

The L1 - L4 coil used in the Daly self-running video has been stripped-down and photos are posted in reply 14019. It may or may not suit this schematic but the build looks the same.

My understanding is that the three unlabelled caps across the input to the ATX PSU amount to around 10uF.

TR1 is a small ferrite toroid approx. 7mm dia x 4mm x 2mm.
TP is not iron cored as the symbol indicates. It has been translated as being a ferrite toroid taken from an ATX PSU, approx 4cm diameter.
The TL494 is a PWM oscillator chip which can be found in an ATX PSU.

I'm currently winding a L1 - L4 coil as I think this offering deserves replication.

The PCB applies to the other schematic - see link above.

Hoppy

[ion]

Thanks Hoppy I was about to post that second schematic and make a few corrections to my earlier post, trying to absorb a lot in the last day or so.

Note that the first and second schematic shows a clamp diode or zener across the coax and resonator cap in series with the pulse transformer Tr1.

I believe these are important elements to be considered in testing the coax.

The diodes will clamp one half the ringing reflected wave and create DC current polarization in the line.

A shorted line reflects an inverted pulse!

The capacitor will facilitate ringing of Tr1 which will be reflected to the SCR allowing it to reset.

Here is a chart showing the effect on the open end of a shorted line as a function of wavelength.

Perhaps we can reverse engineer to find the 1/4 and 1/2 wave resonance of 5.9 meters of RG58 50 Ohm or whatever was used (75 Ohm TV grade?).

[Hoppy]

Thanks Hoppy I was about to post that second schematic and make a few corrections to my earlier post, trying to absorb a lot in the last day or so.

Note that the first and second schematic shows a clamp diode or zener across the coax and resonator cap in series with the pulse transformer Tr1.

I believe these are important elements to be considered in testing the coax.

The diodes will clamp one half the ringing reflected wave and create DC current polarization in the line.

The capacitor will facilitate ringing of Tr1 which will be reflected to the SCR allowing it to reset.

Here is a chart showing the effect on the open end of a shorted line as a function of wavelength.

Perhaps we can reverse engineer to find the 1/4 and 1/2 wave resonance of 5.9 meters of RG58 50 Ohm or whatever was used (75 Ohm TV grade?).

Thanks ION for posting the chart - very useful.

The co-ax is 50 ohm and is apparently from a LAN installation. the white sheath is unusual (usually black) but I think the cable used is probably old.

The co-ax on the stripped-down Dally coil is less than 5.9m.

Hoppy

[exnihiloest]

Neither the TL494 nor the transistor КТ926 allows high frequencies. The highest frequency components are likely around some hundreds Khz. This is too low to get resonant modes in the coax, because at these frequencies the coax length is much shorter than a quaterwave. Imho one can discard line effects and can consider the coax coil as an ordinary coil of a double-wire (the inside one and the shield).
As the current goes in opposite directions in the shield and in the central conductor, the magnetic field from the coax is canceled. The coupling between the coax and the other coils can be only capacitive, especially near the not connected end of the coax. Although the shield is connected to the ground, at the opposite end the potential is not null due to the inductive impedance of the "coax shield coil" that is in series with the ground.

[ion]

Neither the TL494 nor the transistor КТ926 allows high frequencies. The highest frequency components are likely around some hundreds Khz. This is too low to get resonant modes in the coax, because at these frequencies the coax length is much shorter than a quaterwave. Imho one can discard line effects and can consider the coax coil as an ordinary coil of a double-wire (the inside one and the shield).
As the current goes in opposite directions in the shield and in the central conductor, the magnetic field from the coax is canceled. The coupling between the coax and the other coils can be only capacitive, especially near the not connected end of the coax. Although the shield is connected to the ground, at the opposite end the potential is not null due to the inductive impedance of the "coax shield coil" that is in series with the ground.

Both devices you mention may be operated in avalanche mode i.e. extremely sharp turn on pulse from a high voltage supply. While the repetition rate may be in a few hundred kilohertz max due to recharge rate, the frequency components from the fast risetime pulse due to avalanche breakdown may easily be in the tens of megahertz if not even higher. Do not confuse the maximum frequency from the data sheet for this large die device (KT926) or SCR T7 with the turn on time when operated avalanche mode. The recharge rate of the pulse discharge capacitors sets the maximum repetition rate possible. It is possible KT926 is operated as a simple switch since the applied voltage is 150 and the max c-e is 200 V.

As I mentioned before, the SCR is operated in pure avalanche mode with a debounce  switch circuit used to turn the circuit on and gate the SCR to avalanche mode.(unless whoever drew the schematic forgot the feedback line for the gate circuit, thus turning it into a one shot instead of oscillator). It then operates independently at a frequency determined by the recharge rate of the capacitor C11 in conjunction with inductor DR+R.

In Dally1, note that "R", the SCR gate resistor must be experimentally determined and will differ for each SCR used. It is probably a high value, the idea to bring the gate close to turn on threshold of avalanche mode, not normal SCR mode.

In the case of KT926, if operated in avalanche mode, it is gated by 74HCT100 oscillator. The inductor and capacitor set the recharge rate, which is faster than required by 74HCT100 gating.

The idea is to produce (comb effect of) very fast risetime pulses with relatively low repetition rate into the coax.

These items are merely a replacement for the small spark gap used by Kapanadze. As such they do not actually carry full power to the load by induction, but may just be a catalyst towards NMR effects, if we are to believe that theory.

Standing waves of electrostatic potential may show themselves at intervals along the skin of the coax when the coax is impulse excited at its resonant frequency.

[exnihiloest]

@Ion

I was only speaking of this setup:
http://realstrannik.ru/media/kunena/attachments/630/Dally2012.JPG

I don't confuse the working frequency and the frequency components of the pulse.
The diode КД203 and the transistor КТ926 are power components. I didn't found their datasheet but I saw their case. They are not designed for HF.
So their inside parallel capacity should be rather high and prevent high frequencies (2-3 Mhz max?). Moreover the inductance of TR1 probably increases the impedance. 

What maximum frequency do you think you have in the coax? For me, no more than 2 or 3 Mhz at a significant level; probably less.

[ion]

@Ion

I was only speaking of this setup:
http://realstrannik.ru/media/kunena/attachments/630/Dally2012.JPG

I don't confuse the working frequency and the frequency components of the pulse.
The diode КД203 and the transistor КТ926 are power components. I didn't found their datasheet but I saw their case. They are not designed for HF.
So their inside parallel capacity should be rather high and prevent high frequencies (2-3 Mhz max?). Moreover the inductance of TR1 probably increases the impedance.  

What maximum frequency do you think you have in the coax? For me, no more than 2 or 3 Mhz at a significant level; probably less.

As I understand it, avalanche mode is a different mechanism that does not rely on current injection into the base to effect turn on, rather, the entire die is shock excited, and an entrained  electron flow is produced. True, the die capacitance will limit repetitive pulsing at a high frequency, so there is a recovery time involved.
I agree that such a device would only be usable as a low frequency oscillator when connected normally using the base as a feedback or drive element of the oscillator.

It is true that HF transistors (and SCR's) use a different construction technique that allows more efficient propagation of the gate signal across the die.

It is probably the fast risetime that produces harmonics which excite the line at it's resonant frequency not the actual repetition rate.

Tr1's inductance will be a function of the load it sees (plus any leakage inductance). We need to determine this, but it should not be a huge factor in pulse coupling if properly engineered.

I have not tested any 5.9 meter lengths of RG58A/U coax yet, but I am sure coax as used in the TV and LAN industry can support 100 MHz  or more without extreme loss.

[exnihiloest]

...
Tr1's inductance will be a function of the load it sees (plus any leakage inductance).
...

In a 100W HF transmitter powered under 12v, the push-pull power stage is followed by a transformer whose the primary has only one turn of a big cylindrical conductor inside a double core, and the core is constituted of several sections of different ferrites in order to cover a wide band (1.6-30 Mhz).
I'm very pessimistic about the possibility of Tr1 to pass high frequencies at a signicant power if it is built as an ordinary transformer that we can find in switching power supplies.

[ion]

In a 100W HF transmitter powered under 12v, the push-pull power stage is followed by a transformer whose the primary has only one turn of a big cylindrical conductor inside a double core, and the core is constituted of several sections of different ferrites in order to cover a wide band (1.6-30 Mhz).
I'm very pessimistic about the possibility of Tr1 to pass high frequencies at a signicant power if it is built as an ordinary transformer that we can find in switching power supplies.

I agree that it cannot pass large amounts of power. My understanding is that these low power pulses are merely a catalyst of some sort.

Time will tell.

[Hoppy]

I agree that it cannot pass large amounts of power. My understanding is that these low power pulses are merely a catalyst of some sort.

Time will tell.

I'm progressing as close as possible to Dally's coil build and will use a small 15VA 12-0-12 toroidal transformer and bipolar's for the inverter.

Hoppy

[ion]

I'm progressing as close as possible to Dally's coil build and will use a small 15VA 12-0-12 toroidal transformer and bipolar's for the inverter.
Hoppy

It seems to me that this coil could be tested on the bench with a 150 to 200 volt DC power supply for the avalanche circuit, and an oscillator driving small amplifier that can generate clean 4.6 kHz square waves.

Input current would have to be carefully monitored with a few shunt resistors so that power input could be computed, but this is not impossible.

Of course it could not be easily looped with this method.

[d3x0r]

Did anyone here start a replication of this?  maybe is some other thread?

[TinMan]

I found this video on YT about a new kapanadze device that has been built in Turkey.
The P/in and P/out look way to good to be true?
http://www.youtube.com/watch?v=KtadR0skY00

[ion]

I found this video on YT about a new kapanadze device that has been built in Turkey.
The P/in and P/out look way to good to be true?
http://www.youtube.com/watch?v=KtadR0skY00

That video has been around for a while.

Where do you see power in / power out measurements? I only see a clamp amp meter being used. Without knowing the voltage on each side as well as current, we cannot know the power.

Am I missing something?

[TinMan]

That video has been around for a while.

Where do you see power in / power out measurements? I only see a clamp amp meter being used. Without knowing the voltage on each side as well as current, we cannot know the power.

Am I missing something?

No nothing being missed ION.
I was going on the input wireing,which looked to be single phase and i believe 230/240 volts over there.The output to the electric motor looked to be 3 phase,as it had 4 wire's going to the motor(and very thick aswell)
Then they were also driving all the light's,and heating elements aswell.All of this couldnt have been drawing from a single circuit breaker from a 240 volt single phase supply.

But as you say,no where near enough information given to make any accurate measurement's-just my guessing here from what i could see in the video.
Like i said-way to good to be true.

[Grumpy]

hmm

gaps are after the coils - at the top of them

could it be that simple?

[TutorialFE]

Wesley News

http://www.youtube.com/watch?v=_NVBiY13osE&list=UU1ZrGlkxBvG9WOwIs2eApLA