Here are a few photos of my old and trusted PA that was built sometime around 1979. It is based on the GW3ZTH design from the early Seventies.
Two photos from the original design.
Having built a single 4CX250B amplifier for 2m I decided that the next project should use a pair of valves. The metalwork tends to be the major challenge of such a project but fortunately Roger, G4NRG (then G6ANS) was able to produce the metalwork. So we decided on a joint project where Roger produced three complete sets of metalwork and I built two amplifiers and power supplies, one for myself and the second one for Roger.
We basically followed the original design for the RF deck with the exception of the anode tuning where the two discs that formed the tuning capacitor were replaced with a large flap that sat beneath the copper lines. At some point in time the Pawsey Stub output coupling was also changed for a simple loop and capacitor which gave better results, this also allowed the load control to be brought out to the front panel.
As you can see from the photos above cooling was provided by a large snail blower that pressurised the grid compartment. The air passes through the valve bases, cooling the seals, and is then directed by the PTFE chimneys through the anode fins and out through the ends of the anode lines. This method of cooling was fine for “normal” SSB or CW working but not for digi-modes with their extended duty cycles. Firstly blowing into the relatively small grid compartment creates considerable back pressure which in turn reduces the amount of air actually flowing through the valves and secondly the hot exhaust air simply blows into the anode compartment, it circulates and then exits via the two fine mesh grills on the end panel.
I needed to improve the cooling if I wanted to run JT65b EME tests running more than a couple of hundred watts output and not exceed the recommended seal temperatures. It was suggested to me that the K2RIW method of cooling would probably improve the situation. This would require removing the existing PTFE chimneys, moving the blower so that it blows down onto the anodes, thereby pressurising the anode compartment and fabricating some form of chimney / seal between the ends of the anodes lines and the exhaust holes in the chassis wall. By “blowing” into the larger (anode) compartment there should be less back pressure and as a result an increased airflow through the anodes. In this configuration some air must still exit via the grid compart to cool the valve seals, the bulk of the air exits the sealed anode compartment via the anode fins / lines thereby cooling the valves.
I had some reservations about doing the K2RIW changes, apart from the additional metalwork modifications I was not clear on how to ensure sufficient air was passing back through the valve bases into the grid compartment to cool the seals, but I needed to do something.
The changes required the anode lines to have an air tight seal between the end of each line and its respective exhaust aperture (in the wall of the anode compartment). It occurred to me that if I was to do just this part of the K2RIW modifications it should also improve the airflow of the existing setup. To do this I used a 200mm wide sheet of 0.5mm PTFE sheet to form a tube that slid inside each anode line and protruded out through the anode compartment wall. An aluminium end-plate with two holes cut to align with the PTFE tubes was made to fit on the wall of the anode compartment. A small amount of sealant was used to make an air tight seal.
The photo above shows this being tested using a temporary cardboard end-plate. Just making these changes resulted in much better air-flow and the amplifier can now run JT65b at 600W output.
2M PA under test
The final modification (not shown in the photos) was to place a large mesh “box” over the end of the two PTFE outlets. This was sized so as not to restrict the air flow and also not allow fingers near the ends of the tubes!
I’ve not relocated the blower, the PA runs fine for now. I plan at some time to move to a SSPA.