Itsunadze привет

No news from Dally like he was disappeared in earth, but somebody has wrote in Dally blog and informed all that he have success. According to his info i made a new schematic. He was lightened bulb on full brightenes. Conect coax like on schematic and try. His coil on pvp pipe have diamter 5 cm, lenght 30 cm. 1st 1000 turns #0.3mm, 2nd 100 turns #1mm, 3rd coax on full lenght 30cm. He conected light bulb on 2nd winding 1mm. Try first on your 4th winding when on 2nd. Effect is much bigger when he has putted ferrite rods inside coil. He was melted some conductors from excessive power. So coax imput was shorted and conected on trans, coax output was shorted and conected on ground braid. I don't know did anode of DSRD was on the mass or not. Try first without and than with mass.

Reg. Сергей В.

Any comment will be welcomed.

The first circuit is my preferable driving method of DSRDs.
I achieved 1200V @ 2ns FWHM to 50Ω load with this driver and the кд204 diode.

Hello verpies, ПРИВЕТ Сергей В
 
@verpies
 
Can I ask you to share some tech info regarding the first driver from #947 post?
1st - what was the gauge of the wire used by you for the coils
2nd - the number of turns
3rd - diameter of the former
4th - have you used any ferrite cores and if so what type?
5th - what was the supply on L1?
 
6th - any other information that you will find useful it's welcomed.
 
Also I would like to ask you if you can do a small test for me: if you replace the kd204 diode with let's say 1n5408 (if you have any) what are the best results?
 
For me with 24 volts supply on L1, L1=L2 10 turns, with 0,3 mm wire (european standard), with 4mm former and 12nF for C2 the best result using an 1n5408 diode was 9ns wide pulse and 300V peak.
 
I'm curious if you can do better. This way I will know (with some degree of incertitude) if I can do better and need more testing or I need a russian diode (kd20x...) and things will become better. If I can obtain 5ns with 600-800 volts peak and I can go up with PRF without significant loses I have achieved my goal.
 
@Сергей В
 
Thank you for the info supplied to us. I will study your new info and try to apply some on my circuit and see if I can improve the results.
 
Thanks to all.
 
@itsu keep up the experimenting and maybe the lights will come up .
 
I have a suggestion. If you want to pulse L1 with 150V at any desired frequency why don't you try to create a 3rd oscillator with the exact specification from Dally plans (regarding PRF and PWM) and use it in series connection with L1 and the 150 Volts supply. This way you will not be concern anymore about the working frequency of your supply and the voltage that comes out from it (see the attached schematic)
 
Mihai

Mihai,

L3 has a greater influence over the entire coil if the nanopulser it's connected over the shield of coaxial cable.

I too found that, even so that i can not measure the nano-pulse with my HV probe as my scope goes crazy.
But it think that is normal as the pulse is now on the shielding and not "shielded" anymore.
It did not show any different voltage/signal on the L1, L2 or L4 coils (same base sign wave with spikes ontop).

Also my 400v capacitor connected in parallel with L2 has been melted during this tests

Wow, my L2 resonanting cap gets warmer with increasing frequency so that it feels "warm" when using a 100nF cap at 13Khz. It does not differ when pulsing the inner or outer coax leads.

Nothing from what I have tested can go under 9-10ns.

Be aware the my HV probe says in the specifications its minimum fall (and probably rise) time = 7nS.
So i don't expect to see any pulses go below this.

PS My driver for the nanopulser it's looking like the 3rd attachment. Any comment will be welcomed.

My pulser circuit also looks like that as i also toke it from that forum.
Its a KD226D with a 9nF/3000V ceramic cap in series with the CATHODE (like that drawing, so NOT as the Dally circuit), but without the resistors.  Instead i only have hooked up the (shorted) 50 Ohm coax cable across the dsrd (like the Dally circuit).

Presently my toroid is the 10x6x4 one, so former is 4mm.
Wire is 0.28mm teflon isolated
L1 = 4 turns,  L2 = 12 turns
Best result was with a KD226D diode at 120V at 4.2Khz repetition frequency: 1500V / 12nS.

The 1N5408's never got near that.

@itsu keep up the experimenting and maybe the lights will come up .
 
I have a suggestion..........

I have no problems with pulsing the L1 coil. It runs ok by the inverter and i peak it on the resonance frequency determined by the LC combination of L2/C2 (presently 1uF = 4.2Khz)
But thanks anyway for the suggestion :-)

Regards Itsu

@verpies.
 
What do you think about the configuration from attachement?
 
Small description (as I see it): 10ns@1kv pulse travels to coaxial (50 ohms) cable. At the intersection there are 2 small coaxial cables with lengh L=0.2 'til 1 meter (the value should be determined by calculations or by testing) and then forwards to the 50 ohms load. How the pulse will look like at that resistor if we presume that the time taken for the pulse to travel up and down through the small coaxial cables it's under the 10ns period of time (they virtually travel the 2L distance under 10ns aka they re-join initial pulse). 2 cases: one with the small coaxial cables shorted and one with open end.
 
Thank's in advance.
 
Mihai
 
PS The ideea it isn't mine and I'm curios the way you see it, and if it can be applied somehow to our quest.
 
Mihai

Right, "Keep It Short Stupid" principle.

On the back burner does not mean that i have put it aside completely.

F.I. i did rewind my L3 coil partially with another good quality new coax cable (RG-316 which has a copper cladded steel stranded inner conductor).
This coax is much thinner and 7 meters of it only covered 40% of the total coil, so it was asymmetric.
But the same pulse leakage was seen with this (again shorted) coax.
I will need to order about 20M to cover the complete coil.

something about ...


oh yes; Dally's is assymetric also.  In fact a length of his L4 is doubled back on itself also assymetrically....


He noted that his was foil, and if I remember right, his was more like a single conductor along a foil shield;  all of the shielded I have is also covered in braid.


Although, your homemade coax should do the trick there.  I wonder how exactly the electron flow looks in a good cylinder ( like as it's self inducting, it's going to cause more helical paths...)  and although the propagation is actually in a skin effect (so the outside skin is best right?)

Toroflux
For sale... ( I should make a better video that's more to the point of what I mean...)

This can be put on a string/rope, and will rotate in 2 dimensions as it moves; it sort of rolls along.... in a certain interpretation it's a image of the magnetic flux around a moving particle...

The modular approach of my setup enables me to quickly swap different modules and make modifications to the circuit.

I don't think this will bring any advantages but have no idea if this is simply done or takes elaborate electronics to accomplish

Regards Itsu
 

What do you think about the configuration from attachement?

When the pulse reaches the cental point it will branch 3-ways: left, right, up.
The resultant impedance of these three 50Ω coaxial cables will be like one cable having 16.6Ω impedance (50Ω/3).

This impedance mismatch will result in partial reflection of opposite polarity back down into the feeding cable.
The rest of the pulse's energy will be distributed evenly among the 3 branches.  The upper branch is terminated with the 50Ω resistive load thus no impedance mismatch exists there and the pulse will not be reflected from this far end. However the right & left branches will pulses from their far ends..

In the first case, the left & right branches are shorted at their far ends, which will result in reflection of the pulse. This reflection will have an opposite polarity to the forward pulse.

In the second case, the left & right branches are opened at their far ends, which will result in reflection of the pulse. This reflection will have the same polarity as the forward pulse.

The reflected pulses will eventually meet in the center, they will add up if left & right branches had the same length, and their combined energy will be sent back down into the source and up into the 50Ω resistive load. The energy of the reflected pulses will be shared equally by the upper and lower branch.

Because the two left & right branches have combined impedance equal to 25Ω in parallel (50Ω/2) and the reflected pulse is distributed equally into two up&down branches, also having 25Ω combined impedance in parallel (50Ω/2), then for the reflected pulses, there is no impedance mismatch at the central point and no further reflections will be generated into the left & right branches.  In other words all of the reflected energy from the left & right branches will be transferred equally to the upper & lower branch (the source).