Homebrew 4-400 Pair Utilizing Microwave Transformer

Several years ago I had an opportunity to see a master of RF amplifier construction, Randy (VE7RZ), make quite a few grounded grid amplifiers utilizing a standard microwave transformer. Not only did he use the high voltage winding but also re-wired it to provide filament windings as well.That was something I really had some interest in, especially if it was all self contained. All of his amplifiers utilized the 813 pentode in pairs or  a single one.
So, in January of 2010, Randy wound me a  microwave transformer that had a 2100 vac secondary and 2- 5v/14amp filament windings for a pair of 4-400's, which I've got plenty of. He had never done one of these but felt it might work out. It was as beautiful a transformer as you'll ever see . . . but all that voltage and current from a little thing like that?

Short and sweet microwave transformer conversion technique
The transformer was modified just the way it is shown on multiple websites on the internet. Simply snip the attached lead on the secondary from the frame giving the two wire secondary for your full wave bridge rectifier. Knock out the existing Magnatron filament winding, which is held in place with a couple of metal shims, and replace it with your own filament winding at approximately one turn per volt using No. 12 insulated house wiring.
Finished Amplifier
Sometime earlier I had acquired a nice piece of test equipment that had a case that I thought would make a perfect amplifier. I set about stripping it all out, putting a piece of aluminum over the old chassis top and soon had a shiny new chassis to start to work with. I drilled two holes for the 4-400 sockets and attached the transformer. I ordered a nice new Ameritron plate choke along with a pair of "shorty" plate heat sinks and proceeded to start to mount everything next to the 4-400's when low and behold, it was too tall for the case.

You can see in Pic #2 & #3 how I solved the headroom problem. I found a piece of 8" aluminum bar stock and mounted it on the back of the chassis. I rounded the end of that and mounted the plate choke perpendicular to the tubes. It seems to work just fine but it took me a while to come up with it.

                                                           Below is the underside of the amplifier.


Top Side of Chassis

Circuit Design: This is the standard plain vanilla grounded grid circuit. The meter  is in series with the filament center tap to ground. The tank circuit is a modified AirDux tank coil. I found the taps for each band by detaching both plate leads from the tubes and fastening one 2800 ohm resistor from the top of the plate choke to ground. The 2800 ohms was chosen because the plate impendance of a 4-400 is 5600 ohms so two in parallel is 2800 ohms. I then took my MFJ analyzer and attached that to the rf output connector while bridging the antenna relay with a jumper to complete the circuit. For each tap just start with a quick and dirty calculation for each band and zero it in by trial  and error looking for a 1:1 match on the analyzer. I will put a schematic online as soon as I make one using a good schematic drawing software program.

The amp tunes and runs about like you'd expect. About 700W-800W output which is on the low side but as you read further you can see why..

The standing voltage is 3000 vdc but under keying it dips all the way down to about 2150 vdc. Even under idle, at 120 ma, it dips down to about 2400 vdc so output would benefit with better voltage regulation. The filter is only 24 uf so undoubtedly that has something to do with it. I plan on putting on an external filter cap which will raise the overall capacity and hopefully the drop won't be so prevelant.

HEAT! This puppy puts out a lot of heat. Not only are the 4-400's putting out an awful lot of heat but also the microwave transformer is getting VERY hot. The muffin fan below the 4-400's seals pretty well when the case is put on and lined up with the hole in the case bottom. I plan to put a small computer fan to blow air across the transformer in an effort to cool it down. After all, it is providing 2000vdc output from the secondary but also 2- 5volt/14amp filament windings. I cut about a dozen louvered holes in the top of the case which did a lot to let some of the heat out.

I fulfilled my goal of making a totally enclosed RF Amplifier, 160M-10M, all enclosed in one cabinet.

Because the heat became so much of an issue, I decided to take out one of the 4-400's and only run with one. What a difference ! The amp now runs just warm and the output is almost as much as it was with two tubes. The poor microwave transformer was just over saturated with running the filaments of two 4-400's at 14amps each as well as the high voltage.
My BAD !