After successful construction of VTTC with GU-81M tubes I have now tried to to build a VTTC with a completely different kind of tubes.
I choose Czech coaxial ceramic tetrodes RE025XA made in Tesla (equivalent of 4CX250B or SRL460). They are designed primarily for power transmitters.
The advantage is small size, low cost, lower power consumption of heater compared to GU-81M and also less risk of breakage during transport than the giant "bottles" :).
The disadvantage is the need to getter them, longer heat time and higher sensitivity to wrong electric treatment.
The maximum frequency is 500MHz and the anode loss of 250W. Due to their small size, it seems more than enough :).
Not to think small, I joined 6 of these tubes in parallel :). Heater
voltage is 6V (not 6.3 V, as usual). Electron tubes taken together almost 17A heating current and total heating power of
six vacuum tubes does 100W. I used small rewound MOT (class 220 °) for heating. I rewound it to approximately 10.8 V
and the voltage on the primary I limited using ballast for 125W mercury lamp. This makes output voltage under load drop to 6V.
This type of power supply has several advantages. There is no large indush at power on, when the filament cold and have much less resistance.
Reduced voltage on the MOT primary allows for continuous operation. Heater source must be rated for continuous operation, because
ceramic tubes required gettering before use. If not used for more than a month, it is necessary to connected to the heater voltage
for a period of 72 hours. On one piece of tubes I already convinced that without gettering it
will not work - the tube at the anode voltage 250V flashed and welded g1 with g2 together :).
The RE025XA datasheet shows the maximum anode voltage 2kV. The maximum peak voltage is not noted. External surface (creepage) distance between the anode and g2 is only 8mm, which is somewhat disturbing (risk of arcover). The problem I have solved the looping tubes with duct tape :). Start of duct tape should be cut obliquely to avoid the tunnel. This increased creepage distance to more than 2 cm, which already looks much better. As a high voltage anode power supply I used 3 MOTs in parallel with a voltage doubler made of capacitors 9 from the microwave ovens (I think that would be enough to use and 2 MOTs and 6 capacitors). Secondaries are connected in series with the primaries (it works as autotransformer). Input voltage is regulated by a variac. At the output of MOTs is situated protective spark gap, which protects MOTs and other components from high voltage pulses to tens of kV in the case of rapid withdrawal of oscillations. Spark gap has a distance of 2-3 mm. The voltage supply is blocked by high frequency capacitor 40n / 7,5 kV. It is compiled as MMC ten (5x2) of capacitors 100n 1500V type TC343.
I connected the vacuum tubes directly in parallel (corresponding electrodes to each other). Theese vacuum tubes require forced cooling b air flow (see details in Datasheet). Anodes I have placed in the holes in the box, in which the fan (from the microwave) blows. Without forced cooling you can't run even the heaters only (solder on pins would melt)! The fan is therefore necessary also for gettering. First I tried to connect the vacuum tubes into triode wiring Like the VTTC with GU-81M. Thus feedback connected to g1 and g2 connected together. It turned out that this way is not the way. Failed to create a bigger spark than 3 cm. Then I tried to ground g1 and put feedback into g2. Discharges into the air were now 35 cm. After tuning better they extended to 45cm. I used primary from the first VTTC, It has several taps on the primary. This VTTC works best with the anode resonant capacitor connected to the tap with 13 turns. The resonant capacitor is 470pF 15kV 40kVAr made in Soviet Union. The primary is wound on a 16 cm diameter and is wound with insulated wire. Secondary is on 11 cm diameter, the layer of winding has a height of 32 cm and is wound by enamelled wire 1 mm. It has about 270 turns. The feedback winding is 11 turns diameter of 11 cm, placed under the secondary. The feedback winding is connected through a blocking capacitor 22n/1000V, a diode and incadescent lamp 230V 20W-40W. With a smaller bulb (10W) ruined oscillations, with larger (60W) the discharges were slightly shorter.
Operating frequency of this VTTC is about 1MHz. The largest discharge (45 cm in air) gives at the input voltage 160V~. The more increasing of the input voltage does not extend the discharges. Current at this voltage is 16A. That means input power 2560W :) with heater not included. Before MOTs it is recomended to add a circuit breaker.