W6PO design
Robert I. Sutherland, W6PO “High performance 144-Mhz Power amplifier”
Reprint of Eimac Amateur Service Newsletter - W6SAI... AS-47 Aug 22, 1971
W5UN's webpage
1.5 KW output @ 144 MHz.
Made in Democratic Republic of Congo
The construction follow the design of Robert W6PO. (See above the description on W5UN link).
At the first time under push ON, and some tuning, it operated with no problem at all at about 1,2 Kw output, but in the worry to improve the output, the performances, the suppleness of the fine tuning, it was necessary to make some adjustments and some modifications.
Example:
1) The length of the resonant lines:
The lengths of the resonant lines are theoretical on the paper.
It is necessary to modify their lengths on their supports to the near mm.
2) The anode contact:
It has been made the first time according to the W6PO design, with ribbon brass.
After having used a block of brass, (110x110x5 mm) the thermal dissipation increased and the alignment
of the plate capacitors Input and Output was perfect.
3) The top of the anode lines and the capacitors of 100pF were too much hot. The original grid lid
cover on the top of the cavity was closed, only above the tube, so that the
air flow is directed on the right and left them of the cavity and cool the capacitors & lines while leaving.
4) I didn't have HV and filament feed through capacitors; I used a part of PTFE Teflon rod.
I used ceramic capacitors at this place. I have also modify the location of the 1000pf capacitor near the
HV connection in the cavity, (small influence on the line)
5) I also added a setting of the filament voltage. (Was 6v !!!!). Eimac® mentions that for the life of the tube, the voltage must be from 4,75 to 5,25 V max, It is often disregarded, but it’s important..).
See original manuscript of Richard L. Measures, AG6K for what became "The Nearly Perfect Amplifier" (January, 1994 QST)
6) The adjustment of the Input variable capacitor (front face) has to be made with an X6 reduction.
7) The adjustment of the Output capacitor (rear face) was not precise. To pull or to push the coaxial line with the hand is not sufficient. It was necessary to put a thin adjustment with 2 screws and 2 springs that made
good job on the N square connector. A screw for blocking is foreseen, once the optimum adjustment is OK.
8) a step start of 3s for the filament transformer was added to avoid the" inrush current ".
I could not add a silver to all parts of brass and copper of the cavity. But to the point of aesthetic view and oxidization, For sure, it will better. (Probably also the output?!)
All it takes a lot of time, but the result is worthwhile.
The controls protection of Grid, Plate, and HV are made according to a GM3SEK design (Control triode board).
I hesitated to use an other of control triode board from WD7S PRODUCTIONS that is also good. Interesting point is the DS-1BARGRAPH DISPLAY BOARD, excellent for who don’t have any meter.
We never know what it would be able to result on enormous damages, if a protection is not used. It is a Cheap insurance for the tube.
The switching with coaxial relays was replaced by a QSK vacuum relays box (Jennings RF65-12S).
The new scales of the instruments has been achieved with the "trial version of "Meter" by Tonne Software
I printed them with a laser printer color. The result is like "exit of factory". (The first tests with Photoshop were not so beautiful).
The present design of the front face is not very fabulous, I don't have here in 9Q land, a machine operating with X-Y and so large enough to print the 19 inches tracing plastic format in spite that is achieved already under AutoCad ™.
The HV supply is not described in this article, but… Careful with the securities, (step start, etc…..)
High voltage + HF radiations = 
Thinking well two times before putting the fingers there and be sure that all capacities are empty.
The high voltage was reviewed (before was 3000v with No load and 2500v with 1,2kw) by the replacement of the power transformer, + QRO.
Now the HV voltage with NO load is: 3500v, and 3400V with 1850 w output. These tests have been made to check the conditions and the amplifier output with acceptable norms by Eimac®.
Therefore, I should not pass over 1500w tune in normal traffic. I don’t make some higher tests. I’m certain to pass over the 2Kw in CW tune mode, but it is not necessary, and enough for me.
For info: here is the datasheet of the 3CX1500A7/8877 from Eimac®.
* * I forgot…. About the tube conditioning, (being new), the filament voltage, with blower was applied without HV during 48h. This amplifier was manufactured in 9Q/ Kinshasa on a barbecue (only workbench that I have, hi!!), during some of free week end. It was very long because, the lack of spare time, parts and adequate tools.
Now remain to build & install:
- The yagi antennas
References: Articles from DUBUS DJ9BV: Yagi Antennas for 144MHz (1990)
DJ9BV Reiner Bertelsmeier, Technical report, DUBUS magazine 1/90
DJ9BV Reiner Bertelsmeier DUBUS Magazine 3/88, "gain and performances of 144 Mhz antennas".
DL5GBF Peter Gruchot, "2m EME antenna with open wire feed system".
- Rotors site and azimuth, (azimuth will be a YAESU G800 and the site will be a recuperation of satellite dish motor).
- The mast preamplifier (under construction with GaAsfet MGF1302)
- The switching. (Insulated plastic box with coaxial relay TOSHTU models: CZX3500 and vacuum relay Jennings RF65-12S)
- The sequencer: preamplifier, amplifier, TX, (under construction)
- The coaxial cable will be Andrew's LDF4-50. (Approx. 50m).
- A filter pass band is under construction. We never know!!!! Made with the Software ELSIE Filter design.
Other articles will follow about this…..
73 Georges 9Q1EK
The moon between the fingers. For now....!!!! next..... Fire with the antennas.. hi!
Life is too short for QRP 
Emails from SM5DQC three days before silent key.
