Entries from February 2014 ↓

Unique Linear GG Amplifier Input Circuit

The ideal input circuit needs to be able to handle 100 watts of RF, present a 50 ohm impedance to the exciter over the desired frequencies, and prevent the driving power from being wasted in the filament transformer.

The conventional solution is to use a ferrite loaded filament choke and band switched individual low Q tuned circuits – one for each desired band.

Recently there have been several articles offering an alternate solution using a homemade coaxial line composed of copper tubing into which a length of #12 copper wire is inserted as the center conductor. The copper tubinged coax is then wound into a 1 inch diameter coil. At one end the shield and center conductor are wired to the filament transformer. At the other end the center conductor and shield are wired to the tube filament.

A 1000pf variable capacitor is wired across the copper tubing and a 2.5k ohm 2 watt resistor is wired across the capacitor to lower the Q of the circuit. The input RF is applied to the third turn from the filament pins through a .01mf ceramic capacitor.

With the tubing coiled to 15 turns around a 1 inch diameter form and the variable capacitor value at 1000pf we can tune 80 through 10 meters. The tap at 3 turns from the pins at the tube socket provide a 50 ohm impedance to the exciter over the same frequency range.

This alternate method requires the addition of a variable capacitor to the input circuit and burdens the operator with an additional adjustment. It also eliminates the need for seperate individual circuits, an input bandswitch, and a ferrite cored filament choke.

This alternate method is both effective and inexpensive.

The most troublesome part of the alternate method is the uninsulated copper tubing used to create the homemade coax shield. This brings up an obvious question, why not use real coax. Well, for two reasons. Most coax you can wind onto a 1 inch diameter form will have too much voltage drop across it at the current levels required by the filaments.

So the first step is to reduce the filament current. As an example, assume the use of four 811 tubes. If you wire all the filaments in parallel the power requirement will be 16 amps at 6.3 volts. If you wire two tubes in series and parallel their seriesed connections, the power requirement will be 12.6 volts but the current will only be 8 amps. Wire all four filaments in series and you can power the filaments with 4 amps at 25.2 volts.

With the filament current reduced we can now use RG8X to build our input coil. This further reduces the cost of this input circuit solution by eliminating the copper tubing.