|
GS-15B 23cm Amplifiers by KD5FZX Click on callsign for e-mail |
|
Single GS-15B 300W High Performance Amplifier. |
||
|
Operational Instructions & Data for 300W PA |
||
|
Complete 1 x GS-15B "tabletop" PA by W6PQL. |
||
|
Preparing the GS-15b to be water cooled. |
||
|
Description & pictures of a water cooling system. |
 Click on pictures to enlarge |
|
This project is based on the excellent GS-15B PA built by
N8OU. He sent his prototype for
examination and testing; with this proof in hand that the GS-15B would perform
well on 23cm, a second PA was designed using the N8OU prototype as a reference. The
biggest change in this second PA is that it employs a 1/4 wave length input cavity instead
of the original 3/4 wavelength design. This reduces the physical dimensions and increases
PA bandwidth without any compromises. It results in an excellent match; no detectable input
SWR with an input circuit return loss exceeding 23dB!! Anode tuning is smooth and easy.
Comments about tubes:
Variations in tube performance were observed. With 10 Tubes tested, three tubes worked stable
up to 400W+, three worked stable up to 350W+, three worked stable up to 300W+, and one worked
stable up to 275W+. The lower power tube showed a rapid de-tuning jump when drive was
increased, indicating grid or screen distortion. In a more general test, 20 tubes were tested
for operation at 300+ Watts output, 18 passed, the distorted grid tube already noted and an
additional leaky tube were the two failures. See N8OU’s article about tube behavior and power supply recommendations. This
is a significantly better "yield" than found with the GS-23B (one of 5 to 7 is good for 23cm),
and the GS-15B can be found for $15 to $30, delivered (NOS, old stock), which is significantly
less expensive than the $150+ for each GS-23B. NOTE: There is a pulse version of GS-15B with
the designation GI-44B. This tube is NOT suitable for use in this amp. The screen to plate
distance is longer and this tube is designed for higher plate voltage. Also the screen to
plate capacitance is lower so the cavity resonance frequency will be about 30MHz too high.
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
NOTE ON GS-15B FILAMENT PERFORMANCE:
The tube filament was tested to determine the voltage at which the heater reaches full
emission. It was a pleasant surprise to find that 5.2 volts is the lower limit without any
reduction in emission. It is therefore recommended that no more than 5.5 VAC be applied to
GS-15B filaments. This will extend the tube life and reduce the effects of secondary emission
(cathode back heating).
Measurements:
Left: Looking down on the desk with the PA and test set-up. Right:
Drive (left) is 50W scale, output (right) is 500W scale. This power measurement was done
with 10W drive in class B and shows 350W output. The screen voltage was raised to 350V
for more gain and less grid-current. Tests to 500W output were successful.
Materials:
The cavity ring is cut from 4X3” aluminum tubing and is machined to final size. Aluminum is
OK for this PA since the anode impedance is not too low. The screen - plate capacitance on
GS23B is 11.5pF. This equates to 10.68
ohms at 1296MHz. The coaxial section of the cavity is 11 ohms to match this, and use of
aluminum results in significant material losses. This makes silver plating inportant for
the GS-23B 23cm PA. Not so with the GS-15B; The GS15B screen to plate capacitance is 1.8pF.
This equates to 68 ohms at 1296MHz. The screen ring diameter is 42 and the anode is 12mm.
If we use the value in between of 25mm, the impedance is 67.2 ohms. This is not very
critical in this application; the cavity is not getting hot so the losses are very low.
Silver-plating would have very little impact here. All other metal parts are brass and
copper.
|
|
Click on pictures to enlarge |
 Single GS-15B PA |
Download a 1:1 scale (original prints out at 2/3 scale). Print using your browser's print function. |
 All machined parts for 1 x GS15B |
|
Single GS-15B PA cavity construction details. |
go to the top