RX1AS style 144 MHz PA with GS35b

BACK TO TECHNICAL

Project started: 20031110
Project finalized: 20040820

Project status:
Box finished 200312
Mechanical parts finished 200403
Bias/Heater circuits fitted 20040814
Power supply modified for grounded-grid 20040815
PA tested with HV & bias 20040815
Input RF circuit  20040819
Final RF test    20040820

Thanks for help and advice to: PA7JB PE1PFW PA2DW

Do not use this amplifier without a proper lowpass filter ! (eg. a coaxial stepped impedance filter available from Arie PC7M ). The combination of the halfwave resonator, and the capacitive loading does not offer much harmonic supression.

 

Construction links

The original RX1AS PA
Picture by OH5LID

PA2DW

PA3CSG

PC1T

PA4FP - sells parts

OK1BAF - a similar construction with a bigger tube

DL4MEA - not a RX1AS construction, but the manual has many good tips

EA3AXV - not a RX1AS construction, but has a good input circuit (L2 was reduced to 3 windings in my PA)

F1CXX - not a RX1AS construction

G3SEK - bias circuit and how to arrange metering for grounded-grid (page 6)

List of materials - Aluminium

This amplifier can be made on the "kitchen table" with simple tools. There are only two machined parts: The tube/stripline adapter and a teflon termination in the output coupler. The teflon termination can be made with a power drill and a strong knife. The stripline adapter is available from PA4FP.

Output coupler

The output coupler is made with parts from the local hardware store (idea by PE1PFW):
Outer tube is 1/2" copper water pipe, internal diameter 12.7mm. Inner tube is a solid brass rod, diameter 6mm. The female N-pin is removed from the plug and soldered to a hole drilled at one end of the rod. The wall feedtrough is a standard brass 1/2" water pipe fitting, which is drilled trough to allow the pipe to pass. A slot is cut in the external tread to make it press against the tube when tightened. The loading disc is made with a 83mm (3 1/4") hole cutter in a brass plate.

Input circuit

When I got around to making the input matching circuit, I stumbled over a posting on the AMPS reflector by OZ5TG recommending using a 150 Ω resistor (from cathode to GND) to simulate the dynamic resistance of the "live" tube. This value worked fine for me. I was able to tweak the input circuit (L-C-L T-network) copied from EA3AXV using an MFJ SWR analyzer. The finished PA tunes nicely to a perfect match on the input. By tweaking the coils in the "cold" circuit I avoided using a variable "floating" capacitor in series with the first coil. I think this has saved me some losses in the input circuit. By the way - I never decoupled the heater towards the cathode. It seems to work fine without it.

Heater

If you apply the nominal heater voltage to a cold tube of the Russian types (true at least for GS35b) you will far exceed the allowed heater loss, because the cold heater will have less resistance, and you will eventually blow the tube. There are several ways of dealing with this:

1. Apply the nominal heater voltage trough a resistor, which is bypassed after a delay (used by PA2DW)

2. Apply a higher heater voltage trough a permanent resistor. Choose the resistor value such that the rated heater loss can never be exceeded (about 18V / 1.8 Ω for GS35b). You will suffer some power loss in the resistor(s), but they will keep you warm on a cold day (used by PA5DD).

3. Choose a too small heater transformer, that will saturate the core when connected to the cold heater. Could be combined with a small resistor (proposed by PE1PFW).

Cooling

A typical axial ventilator (e.g. PAPST 4650N) delivers 160 m³/hr (94 CFM) which is enough for colling the GS35b. Of course the ventilator only delivers this volume at low back pressure, which is why I decided to mount two blowers. Depending on which anode loss, duty cycle and enviromental temperature you intend to operate with, the second blower can be left out. A real killer is dry air, be careful when operating in places where the humidity is low.

The air flow should be from the cathode compartment to the anode compartment, which is the standard direction (air intake over struts) of 119x119mm axial blowers in the configuration, that I have choosen. Note that axial blowers cannot be turned around, without generating a lot more noise, as the fan gets too close to the mounting wall. I learnt this the hard way, as I had two strong odd-size 127x127mm blowers, which have reverse airflow (air exhaust over struts). Not being able to turn them, I had to blow hot air into the cathode compartment. This cost me 3 blown bias circuits over 2 years of use. It is worth noting, that quite some power is dissipated in the bias circuit. Although not shown in the photos, I have also mounted a heat sink on the top of the bias metal case. Cooling air circulation around the bias circuit is vital.

As you can see from some of the constructions links, others prefer to use a tangential ventilator blowing into the cathode comnpartment, or the separation space between the grid and the anode.This will work fine aswell.





Portable use


BACK TO TECHNICAL