The 432 MHz EIMAC 8938 Kilowatt Amplifier
In the late 1970's, I was quite involved in EME work on 144 MHz. One of my mentors was Bob Sutherland, W6PO. One night on one of our many telephone calls (The internet had not been invented yet), I happened to mention to him that I had just completed a 432 EME contact with W0YZS to give him one of his last states for WAS on that band. I then mentioned to Bob, that I was going to rig up a permanent EME station on 432. My sked with W0YZS was a jury rigged affair where I cobbled together a four yagi array over a weekend. I wanted a better setup, with a bigger antenna and a better rf source. Bob asked me what I was using for gear. My amplifier was a pair of 4CX250Bs in a K2RIW stripline. I was beating it to death on EME. I guess Bob felt sorry for me, and, over the next few weeks, all sorts of components started showing up at my doorstep. One of the boxes held a nice new Eimac 8938 1500 watt dissipation triode. Well I figured I better make an amplifier pretty quick, and show the results to W6PO so he would not think I was some kind of deadbeat. I went to the scrap metal yard and bought a bunch of aluminum plate, and started to build the chassis. I decided on using the W6PO 8938 input circuit, but opted for a stripline output tank rather than the W6PO cavity design. One of the reasons was I did not have access to much brass to build a good cavity! I used 1/8" thick aluminum plate, with 1/8" thick angle stock at the corners to build the chassis. I tapped the angle brackets for 4-40 threads and peppered the thing with holes every 3/4" or so. I did not want any RF leakage I guess! The end result was a compact case that had over 325 tapped holes! You had to give me a "A" for effort, even if the thing did not work! When I look at the amplifier today, I just shake my head. I must have been nuts!!! I did it all as best as I could, so as not to disappoint W6PO. I figured that I owed him that much.
Enclosed are a few pictures of the amplifier. It has been working faithfully since 1978 and delivers 1300-1500 watts on 432. Thanks again, Bob!
I spent considerable time planning the front panel layout. The panel is a 3/16" aluminum rack panel. Most homebrew 432 amplifiers at the time had no symmetry on the front panel. Knob placement depended on where everything fit. I wanted something that looked reasonably symmetrical. To achieve that end, I employed several gear trains inside the amplifier to route the tuning controls to the proper place. In addition, I used a series of small pulleys and dial cord to control the tuning and loading "flappers". It all required considerable planning and scrounging of hardware. I wanted this amplifier to look good as well as work well! I did not want to disappoint Mr. Sutherland! My machine shop at the time consisted of a 115 volt one speed hand drill, a hack saw, chassis punches, assorted files, and a nibbling tool!!!
The top of the amplifier shows a nifty air exhaust RF filter designed to keep rf inside the box. It is a waveguide beyond cutoff type, that provides tremendous attenuation of any rf trying to leak out of the plate compartment. The exhaust air RF filter is located just over the 8938 anode, and allows heated air to escape from the compartment. Note the air flow switch. This will turn off the amplifier should the air flow cease for any reason. The large transformer is the 5 volt filament transformer. It is rated for 20 amps, but has a current limiting resistor in the primary for two reasons. The first is to limit inrush current and then, also to drop the filament voltage to about 4.5 vac. This is needed to counteract back bombardment of the cathode that is evident at UHF frequencies.
A shot of the business end of the 8938 amplifier. The 1/16" thick brass stripline is visible along with the tuning flapper and the smaller output loading flapper shown attached to the output N connector on the back of the amplifier. The initial stripline was made from 0.062 G-10 double sided PC board, but the effects were disastrous. The two sides were not bonded together, and large RF arcs were commonplace. The stripline is a very High Q structure, and can be a rather tough adversary at tiimes. A solid brass metal line solved most of the arcing problems. Silver plating added about 100 watts of extra output over the bare brass. The stripline is a halfwave resonant length, with the HV attachment point at the low voltage midpoint area of the stripline. Note that the midpoint is next to the tube anode. An air inlet is visible on the back panel as well. A teflon chimney (not shown) couples the anode cooler to the air exhaust port on the cover plate. The small transformer visible is a 24 vac unit that operates the keying relay as well as some protection circuitry that is currently being added to the amplifier. The new circuit will protect the amplifier from loss of high voltage and also grid overload current. This is very important for contest operation where some operators may not be familiar with the amplifier operation. With the cost of tubes so high, such protection is mandatory.
A bottom view of the amplifier. Note there are two compartments inside the main chassis. The leftmost compartment contains the cathode input circuit. There is no other wiring in this area. An 8938 will take anywhere up to about 85 watts of drive. The input N connector can be adjusted so that the input coupling capacitor moves closer to the input cathode tuned line for loading adjustments. The length of that input coaxial line is approximately 1/4 wave long, and acts as a balun of sorts to contain rf currents on the outside of the transmission line inside the compartment. There are two rf chokes in the filament leads. The large connector on the rear panel is an Amphenol MS type HV connector. Note also the air intake port as well as a secondary air supply that routes a small amount of air to blow across the tube cathode and filament seals. This is important for the 8938.
The component area on the far right contains the filament transformer, bias zener diode, and assorted resistors used for keying the amplifier. If you look closely, you can see that the entire amplifier is buttoned up externally with clear plexiglass panels, to keep mice out of the amplifier. Mice are a terrible problem, and all of the amplifiers and other gear must be closed up to prevent mice from making nests in delicate areas. The plexiglass idea has worked out well. Having a beagle on the premises is a good choice too!
Closeup of the cathode tuning circuit. The gear train on the left is attached to the "GRID" tuning knob on the front panel. The gear train drives a threaded shaft that pushes a square brass plate in proximity to the input tuned line for resonance adjustments. Note that a large piece of thin walled brass sheet connects the tuning plate to ground, eliminating any sliding contacts that could become erratic or "noisy". This amplifier is a smooth operating unit!
Some of the techniques used to organize the front panel knobs involved gear trains, right angle drives, and these pulleys and dial cord assemblies to manipulate the tune and load flappers in the output stripline. This compartment is located just below the stripline compartment, and the cords run through small holes in the chassis, to adjust the flappers. I stole the nylon and brass pulleys from old defunct car radios. That right angle gear drive in the top of the picture looks like it came right out of WW II.
My main reason for detailing this amplifier, is to acknowledge the tremendous effect that W6PO had on my ham experience. Bob Sutherland gave me the push I needed to tackle big projects, and this amplifier was one of many big projects. Interestingly, Bob helped literally hundreds of hams in similar fashion. There are many amplifiers such as this one all across the country, that had their start in a similar way. Bob's quiet assistance and gentle prodding was everywhere. Bob passed away in January of 2004, and his presence in many facets of amateur radio is sorely missed. For more information you can visit http://sutherland.blogs.com/ Bob's daughter, Janice has started a very nice site there that commemorates Bob's contributions to the hobby. Please check it out, and the next time you make a moonbounce contact, thank Bob for being instrumental in making EME what it is today.