Voltage regulator for G1-- CZ -- EN

Voltage regulator for G2-- CZ -- EN

"Long life" 144 MHz, 750+ Watt PA


With GU78B tube....

As our club is very active in VHF/UHF contests there has always been a need of a reliable high-power PA for 2m band. Although the current power limit in OK is just 750 W, we decided to construct a brand new PA with a big margin of power which would last "forever" and can be possibly used for EME as well. Since 2 pieces of GU78B were lying for longer time on the shelf in my ham shack I chose this tube. It's a tetrode designed for linear PAs with sufficient power dissipation. Besides I was able to get a socket for it on flea market at reasonable price. Also driving such a tube to just a little part of its potential power ensures a clean signal in comparison to smaller tubes driven to their maximum. This is very important in big VHF contests with high number of stations in a small area (a very frequent case in OK).

For fundamental construction problems at high power level (especially the anode isolation capacity) I droped the usual lambda/4 design and developed a construction with the lambda/2 resonator. Of course there's a big minus in its dimensions, not only due to total cavity length of almost 1m, but also due to the high voltage (about 105kV) at the end of the resonator. However, there are some important advantages in this construction. No anode isolation construction capacitor, small tuning and coupling capacities and the last but not least the long and massive copper pipe improves the temperature stability of the tube. The higher intended power the bigger significance of these advantages. From certain power level the classic lamda/4 design is not feasible.

Note: PA is not for sale !!!

Technical parameters:

Ua - 3100v - no load.

Ua - 2500v - at full load.

Ug2 - 315 V

Ia - 1.5 A

Ia PEP - 4A

I fil - 3.5 - 4A

U fil - 27V

U fil - EME - 28V

P out - 3000W linear (intended)

P out MAX - 4000W peak maximum to limitation

http://www5a.biglobe.ne.jp/~jh2clv/ - GU84-6m

More pics on N2DX web page.

And remind: Life is too short for QRP!!!


Practical tests and construction process.


OK1VPZ promised a big cooling turbo fan. It means that old military blowers can be dismounted and some changes will be
done at the rack and air vents for better cooling.


PA tested under load 900W in 500W out. Except of small matching problems in cathode circuit all OK.


There's no appropriate material for the rack at the moment. As soon as I have some the work can continue.


Tests of g2 protection and g2 power supply. Test OK - switch on at 2100 V of anode voltage, off at 1000V.


.Added a choke for protection of the output relay, the auxiliary transformer and the plate with regulated power supply for g2.


Temperature test. 1250 W DC-dissipation for 5 minutes. The temperature of outgoining air was about 60 degrees centigrade (not very precise measuring), but it's OK for the sufficient cooling. Estimated temperature at the tube was about 130 degrees.


Completed ready for TX

Coax relay for RX/TX switch ...

Anode current abt. 300mA of 3kV --- 900W IN

Temp of big dummy load in 3rd minute

Output PWR. This meter is up to 600W only ...

Cooling air temp 400W out after 3min. Room temp 15°C

Frame structure is a bit large ...

Completed for transport ...

Cathode connection


After a slap in tube :-) R-1Ohm 5W

Schema. Unfortunately not in full resolution....

Note .. g1 and g2 power supply are connected as usual parallel stabilizers. G2 protection observe high voltage on anode, if there is at least 1kV potential. If not, the G2 is switched to zero to protect a tube. The next protection is watching of SWR....

Test of g2 protection .. switch ON by 2100V and switch OFF by 1000V. Important circuit, when High voltage on anode is OFF!!!

From top to down:  G2 PS, HV socket, G1 PS, transformer and van.

Some of important parts pic...

On right side choke of antenna galvanic connection protection. It protect coax. relay. If the output is not connected, the G1 PS is blocked.

G1 PS and G2 protection schema. I made it in school during Czech lesson, hi... On anode voltage is divider 1 by 1000 and if the HV is more, than 1,5kV, the G2 PS switch ON. It is very important! If the G2 potential is higher, than anode, all electrons, emited by cathode are impacting to G2 and because such G2 grid is not designed for major load, burn out and the tube is destroyed!!!

Ventilation cover MK III. :-)

Mechanical assembly

End of resonator. The white material is a teflon. On left tune on right coupling disc.

Coupling adjustment..

Tune disc ...

Anode choke and 2nd. version of ventilation cover

Vacuum improvement !!!

My well proven PS - laboratory supply TESLA 2x400V. Just used in proces of GU78b vacuum improvement... Ua - 800V , Ia - 6mA , Uf - 28v , If - 4A

Detail of tuning system. Some parts were used from old military equipment duplexer RDM 6/12

Well done: Q of unloaded resonator. Qn >> 700.

The very first RF measure... where is the resonance... resonator is a bit long.... handsaw helps...
when the resonator pipe was cut by  few cm, the 65cm lenght has considered as the right one....

Socket connection.

Long way home :-)

Resonator pipe holders

HV connection and capacity from PCB plate

Not very nice, but it works!
Connection of GU78b and a resonator pipe.

How to match the copper pipe 100mm to the tube heatsink with 110mm dia? That's the question!

Tube socket bottom view with matching resistors and a LC circuit

First tube insertion into a cavity

tube in a socket

teflon legs were finaly not used

socket assembly

LC match

cavity bulkhead for socket

tinsmith job

the cavity construction is  in total 900mm long

Collection: from left GU78b , GS35b , GU74b , GI7b , RE025XA (4CX250) , GU50 , HT323 (2C39BA)

G2 stabilizer is a basic condition for PA with tetrode...

G2 PS is fundamental for such PA .. it must be designed as parallel stabilizer! If not, some electrons are returning from anode and it increase potential on G2 until the tube take a spark... and if the voltage on G2 is not stable, it change the gain and the PA produce splatters. And even it may destroy the tube at all!

On the very beginning. It has started with aluminium plate 1000x1000mm, 2mm thickness, used for cavity production..
Size:  900mm high , 224mm wide , 232mm deep. On right is a prepared PS for PA: 3100V DC 4.5kW input

Last update of this page content :12.4.2004 12.50

More: http://www.nd2x.net/OK1BAF.html


Filament and PTT connection into cavity

Input of the cooling air from the big blower ..

Socket with added attenuator resistance 125Ohm .

Last pic of testing before cooling improvement