The KT805A is a higher power version.
ALL the same, it's a crap circuit if your paying that kind of money and it won't last very long as the circuit is very basic with no protection.
but this circuit is just really 'crap' draw any real power from it and you will get smoke because that HV will
be reflected into the base of your transistor and another one bites the dust as in Queen song!.

Where is your zener protection???

Zener ?  that won't fix it for our use, you need the frequency fixed between strict limits
1.5mhz for a 40-meter band and 2mhz for a 37.5m band it need feeding into a buffer and then a driver
with fast on and fast off time MOSFETs have a lot of capacitance that needs disputing real fast in this circuit
and then it will need tuning like a 160meter band transmitter tank to minimize loss and on the output
I would use a fast diode and a VDR, not a zener.
But then you're building it not me.

For Kacher, the protection from overheating is more necessary. A large radiator is required for the transistor. Overvoltage protection is only necessary when powered from relatively high voltages.

Thank you Sergy. Personally i use the same transistor for the last 4 years without any overvoltage failures across transistor's base.

One of the most important components of katcher is this nf Range capacitor at the input right after the chock. Ruslan uses a 100nF but you can go a lot lower. The chock is nessesary there and needs to be at least 50 times higher in inductance than the thick primary and more. But do not over do it there. The reason for this araingement is as follows:
When transistor conducts, this nf cap is discharging through the primary before current starts flowing again from the battery. This delay of input battery's current happens due to this chock. So this cap in combination with the primary gives a nice voltage oscillation and just before the cap goes empty, transistor closes. This action superimposes a beautiful high voltage spike right in the middle of the Cap/prim voltage oscillation. Like in the old times 
This has nothing to do with OU of cource, but makes your Kachers work very close to the operation that they were designed for. 
     

   Jeg:   Why do you want to build the simple Kacher circuit, when you have a much better controlable Kacher circuit already?   Although the controlable Kacher circuit is more complex, with many more components, each one is hogging up the output.
Yet it can be controlled as far as duty cycle, frequency, dead time etz...  so in that sense, it's the right way to go for the grenade set up. So, does your controllable Kacher circuit still work? And what are you intending is the purpose of building the simple Kacher?
   
   Apecore:  Good to see you back, and that you've got nice big artistic and decorative wall hanging.    I'm still wondering what we are all missing...
   Jeg: The two caps after the choke on the Ruslan simple Kacher are there to filter any AC component that may try to make it's way to the transistor. However, they in themselves do not produce a spike, but a sine wave, instead. If that is not what you are talking about then what capacitor are you mentioning that will produce the spike. 
   Sorry for the blurry image of the diagram.
   Picture of Ruslan's simple Kacher circuit. Same as what I built up and use.

Oh, you didn't build that heap of junk Shark was pedaling  I made it and a PCB board too,
It doesn't work all the Mos Fets he says to use have far too much 'C' it's junk! but doing a circuit to solve that switching problem is a good idea but not that one.

   AG:  Not sure who you are posting to, but,  Geo has shown the Akula kacher circuit built up, and it seams to be working just fine.
   No one is "pedaling" anything.

   There is nothing "random" about the tuning of the simple Kacher circuit. It can be tuned to the proper frequency, and it's input is also controllable if using a DC to DC converter at the input, up to 50v, or so.  Ruslan, Stalker, and Adrian have shown self running, using this simple Kacher circuit.   Nor I have not burnt the 2SC5200 transistor, as yet. 

Jeg, NickZ,

Thanks to you all....and of course all other active and passive members.....  good to see this thread is still alive..
Yes.. i guess most of us do miss something... as they probably did not yet hang it on yhe wall.

As i said longtime before,..  i m only here for the experiment and second hoping to see if some "wierd" results are scalable and explanable.
Of course not in manufacturing any kind of Boost buck converter with a grenade coil.

My personal approach is very down to earth as it should be simple.

Disruptive pulsing with HV field, proabably disrupting the magnetic field created by the inductor.
During this disruptive stage there should flow energy inside the grenade/ inductor coilpair.

On the otherhand some schematics show the 28T yoke winding having a different function then powering the kacher circuit.. 

Justawatt and Magpwr posted it..     « Reply #20544 on: February 16, 2019, 06:49:54 PM »..  "We drink now and do nothing more schematic"

This 28T does generate squarewave's ( also disruptive)...... and in some schematics connected in series with a cap to the grenade...
By changing the series 28T cap..  ( as mentioned by AG .. switch capacitor).......i can create "Radio moscow"noise out of the yoke..  with out the kacher running.

With no doubt...  28T and inductor loop magnetic fields both affect the grenade.......  some moment here the kacher burst comes in...
This kacher "burst"  as T1000 several times pointed out is very critical in tuning to this "Charging" performance stage.

In my opinion.... we still aint able to make a "unipolair"puls @ antenna with the kacher setup ..  it still has a sinewave......
Did someone accomplished this "unipolair puls"at antenna?......

Greetings