3 x 4-400A project

Like all my amplifier projects, this one started out as a "what if?" For several years I had been using an amp with two 4-400A's in parallel, triode connected, in a grounded grid mono band 40 meter chassis. The chassis was an experiment that was never quite finished. It used cross flow cooling similar to the Heathkit method and proved to be very reliable. Although I could get 1500 watts out with the pair of 4-400A's, it took a lot of drive at the anode voltage I was running. At this level of output, I was seeing some output compression. After with this for a while, the "what if" process started up and the idea of running three tubes in parallel came to be. I had plenty of broadcast "pulls" and three spare sockets on hand. The biggest obstacles were filament supply and cooling. With a total of 45 amps current required to light all those tubes, both the transformer and the filament choke were cause for concern. I found a transformer capable of that current in my "Junqe Box" but when connected to three tubes through a typical filament choke, the voltage across the tube pins was less than the minimum required by the tube manufacturer. Eventually, a choke was wound with 10 gauge wire wrapped with teflon tape (for heat resistance) that was sufficiently low resistance to  minimize the voltage drop.

Cooling was assured by having two muffin fans blowing across the tubes and tube sockets. The sockets are mounted on a sub chassis which is mounted to the main chassis with threaded rod standoffs. Here's what the underside of the sub chassis looks like:


Notice the coax for the cathode RF drive and the ugly filament choke! Control and screen grids are connected directly to the chassis with copper strap. Many years ago I gave up on the idea of monitoring grid current in this type of grounded grid amplifier. I always ran what ever drive I needed to get the output power I wanted regardless of what the grids were doing, so why bother? You can't hurt the tubes in normal operation and as long as you stay out of compression, you can't "overdrive" them.

All of my HF amps are installed in a rack which provides selected RF and HV lines and also provides what metering is required. This makes it easier to build each individual amp. The HF amps are all built behind 19 inch rack panels. The entire chassis for this amp was fabricated from salvaged aluminum sheet. All of the components were also acquired from swap meets or salvaged equipment. Some of the parts date back to old TV sets I salvaged out in the early 60's! A few parts are from old WW II military units acquired and salvaged out through the years.

Monobanders are really simple to build. Here's a photo looking from top left:

top left

In the upper right is a small filament transformer which provides the DC center tap for the cathode return. The 0.25KVA filament transformer is just below it. The diode string provides a small amount of cathode bias. The two muffin fans pull air in from the right side of the chassis and exhaust it out the left side. I wanted to use two larger size fans but I just couldn't squeeze them in the space I had available. In any case, they seem to be able to cool things adequately. Top center and top left are the resistors and relay for the filament in-rush current limiter circuit. Left center at the coax connector is the pi-network input matching circuit. The anode RF choke is salvaged magnet wire wound on a piece of teflon rod. Notice the parasitic suppressors. I had hopped that I could build this thing in a way which would not require them. The last three amps I built didn't need them and I thought I was on a roll. Oh well! It turned out this thing made a great 190 MHz oscillator. No science was involved in the suppressor design. Since no 10 meter operation was contemplated, I didn't have to worry about too much inductance so I just made them "big enough"!

Looking at it from top right:

top right

Air variable caps are used for both output tune and load functions. No need for an expensive vacuum variable in this application. A couple 100 pF doorknob caps are used  to increase the load capacity. No screens are on the fans in this photo. I didn't have the right kind of hardware cloth handy and didn't want to hold up the project. If an when I ever get some, I might put it on. Probably never happen!

Looking from top rear:

top rearm

Along the rear apron left to right  RF input, AC mains input, HV input, and RF output. Most of my recent amp projects have an AC connector salvaged from old computer power supplies. I used to always step on the dangling line cord, now the line cord stays in the rack. The tank coil is wound from #10  bare copper wire. The first coil I made was from 1/4 inch copper tube. It was a thing of beauty but didn't fit well in the space allocated. The 10 gauge is good enough for 40 meters and is somewhat easier to work with. The rear chassis cover is made in two parts. The bottom piece stays put and has all the connectors on it. The top piece is easily removed for modifications. Notice I said "modifications" instead of maintenance? I seldom do maintenance, because by the time the thing breaks, I'm tired of it and want to modify it into something else! The input network and the output coil can be changed in a matter of minutes if need be and I can have this on 80 or 20, whatever, quite easily.

Smoke test:

under test

Here the amp chassis is placed in the rack and powered up for some testing prior to final assembly. This is at full power output running a string of fast dots. The bright orange spot on the left most tube is an artifact of the photograph, they all run the same color in reality.

Covers on:

left oblique

As you can see here, the left side of the chassis is covered with a sheet of perforated aluminum. All other surfaces are sealed so that the air flow is from right to left across the tubes and other components that require cooling. I was worried about hot spots, but it seems to work OK.

In the rack:


The finished 40 meter monobander is placed in the rack right above the 4 x 813 amp deck. Some day I'll have to finish the detail work on that one! But that's another story!

Like all my amplifiers and most everything else I build, I have no documentation! Don't ask, you won't get a schematic because I don't have one. All the RF stuff is right out of the text books and uses commonly available computer software to do the calculations. I just wire it up as I go and try to make it look like an amplifier. I do this for fun, not to prove I'm a world class RF engineer.

07jul06 lhm