I get a large number of Emails concerning my amplifiers. These photos were sent to me by one particular correspondent who wishes to remain anonymous but it really is a masterpiece of workmanship which deserves its own webpage.
I am working on the project for more than two years now. A friend of mine is using and testing the amp right now. He made a lot of suggestions and often tested the amp, since he has a lot more time than me :-).
When we started the design, we decided to use it grounded grid / grounded screen grid. First we did not want the complexity of a screen grid supply, second we already had an GI7b / GS 35b amp delivering about 1.5kW Power so there would be enough power to drive this tube in this configuration.
As there were no data sheets available with a grounded screen grid, We first had to check out what grid bias would be necessary to get around 0.5 amps bias anode current. We found out that we could simply put the cathode to ground with 0 V bias the tube would perfectly run at 500 mA at a Anode Voltage of about 6.5 kV ! We choose the high current for a maximum linearity.
The rest was straight forward. The first design we simply designed from our experiences with the GS 35b. So we build a half wave anode line. At least what we thought that would be a half wave line :-).
When we started tuning, we got 2kW with almost no optimization. However the amplifier was really sharp in resonance with a 3dB bandwith of only 130kHz or so.
So we rechecked our design and found out that we hadn't built a half wave line but actually a full wave line :-)) and the Q was much too high for successful operation on full power. The output capacitance of the tube is also extremely high in grounded grid so most of the first quarter wave of the plate line is within the tube :-).
What followed were four steps of shortening the plate line and adding more capacity to the anode end of the line to get down the plate Q.
With the current design we only have a tube of about 28cm for the plate line and a 3dB bandwith of more than 1 MHz !
It is essential to get the plate Line's "Z" as low as possible to achieve a low Plate line Q. If you go the usual 70 Ohms design, you will definitely end in flashovers when you finally run the amp in full power.
We are now at about 40 Ohms !
Since the required capacitance on the anode could not be achieved with the standard flapper concept usually found in 2 meter amps we decided to add a vacuum capacitor on the anode end. We had some experiences with flashovers through 1.5mm PTFE in a GS 35b amp, so we didn't dare to try something like this in this amp.
The output coupling is on a "subchassis" within the chassis, so we could optimize the position for maximum output.
The socket we had was wired for short wave, so we decided to build our own with fingerstock material.
The blower is located on the back of the amp and blows through the plate pipe. This will give an optimum air flow for this configuration.
One word for the gain of the stage : We got a little more than 13dB you can usually get out of a grounded grid amplifier. We currently need about 80 Watts input for 2kW output. The HT power supply is at 5kV without any load and we measured 3.6kV with about 0.8 amps. So the efficiency of this amp doesn't seem to be too bad.
Currently we do not have a Bird with a 10kW Plug for 144 MHz. To measure the power we use a homebrew coupler so we think we are in a 1-2dB Range for our power measurements. We checked the output Voltage with an 200 Mhz Oscilloscope running the amp on a 1kW R&S 60 Ohms dummy (a nice commercial dummy with 200x40x40cm size.).
We will check the exact power output later after the final Power supply will be running.
The operating position in the final chassis will be upside down to meet the datasheet from eimac.
Page last updated 1 Sept 2003