K5AM                 k5am home page      k5am amplifier page
Raytrack 50 MHz Amplifier; modifications.
Posted on:vhf@w6yx.stanford.edu, 1997 Aug 29.
Revised, 2012 Feb 1.

These are merely very rough notes on recent mods. The notes
are in random order. Some plans for work that might be done
later are also noted. Comments and suggestions for solutions
to the remaining problems will be appreciated. The amp was
obtained as back-up for the DM54 Horse Mountain contest site.

A. Power supply

1. Attached handle to top. Replaced diode/capacitor board
spacers with fixed-to-board threaded spacers.

2. Replaced each diode string (original 6 x 1N4007, 1 A, 1
kV, 30 A surge) by 10 x 1N5408, 3 A, 1 kV, 200 A surge. More
important than the current rating or the total voltage
rating may be the peak surge current rating, so step-start
may not be needed. The amp arrived with 6 x 1N4005 - only
600 V each. These shorted out during the initial test.

3. Turn on amp in SSB mode. If turned on in CW mode, the
primary relay will arc badly.

4. Plan: use CW primary relay for a step start circuit. CW
tune-up is not needed if pulse tuning is used. Pulse tuning
at 33% duty cycle works well. (Design Notes for "A Luxury
Linear" Amplifier, QEX, November, 1996, 13-20.)

5. Deleted 3 x 2.7 ohm 1/2 watt resistors in HV lead in
power supply. Limiting resistor added in RF deck; see note
B4 below.

6. Changed 2 x 18 ohm 2 watt resistors in negative lead in
power supply to one 10 ohm 10 watt.

7. Removed line cord ground wire from questionable ground at
nylon cable clamp and connected to chassis.

8. Plan: Replace sluggish circuit breakers with fast acting

9. In-rush protection, Keystone Carbon CL-10, 12 amp,
Digi-Key KC001L.

B. RF Deck

1. Eliminated cabinet jamming problem (caused by screws
which hold feet). Removed feet, attached stick-on feet 1/4
inch high to same spots. Plan: Find 1/2 inch plastic feet
with space for nut inside foot.

2. Found one cause of arcing and power supply failure. Noted
severe arcing burns between corner of HV shorting interlock
device bracket and HV connector hardware. Removed interlock;
spring was broken anyway, so it was non-functional.
(Warning: Do not remove interlock; this could reduce the
number of live 6 meter operators by one, and I need your
grid square.
) Plan: Install a good interlock. Always
disconnect all cables before opening any amp. Then use a
shorting device. My favorite method is to keep my eye on the
plate voltage meter while I switch the amp off, watch the
meter decay to zero, and immediately pull the line plug. The
220 outlet is at the front of th
e workbench in a prominent place
where it cannot be overlooked.

3. Replaced over-long HV connector hardware with proper
sized hardware. Cleaned burns from connector.

4. Added HV limit, 20 ohm, 10 watt, on rear panel between
tubes. Used 1/2 inch ceramic spacers, nylon hardware, and
solder lugs.

5. Replaced parasitic suppressors. Amp arrived with original
suppressors missing, and wrong type installed, which burned
up in 5 seconds. Used 4 x 0.375 x 0.010 inch brass strips
with 68 ohm carbon composition resistors. Amp is absolutely
stable and resistors show no signs of overheating.

6. Plan: Move screws on deck which prevent chimneys from
seating fully on deck.

7. Replaced steel hardware in plate tank circuit with brass.

8. Removed one blocking capacitor. One is enough, and the
parallel connection of two may introduce parasitics.

9. Added 10 ohm 2 watt resistor in lamp ground lead to
increase lamp life.

10. Modified grid metering circuit to protect meter from
burn-out. Circuit on request.

11. Relay is 12 VAC, incompatible with solid-state control
from transceivers. Added DC PTT circuit. Circuit on request.

12. Replaced one chimney.

13. Replaced tubes.

14. Sealed leaks in presurized grid compartment with
filament tape and hot glue.

15. Checked meter calibration. Replaced shunt.

16. Plan: Replace bias diodes; 1N5408.  QUESTION: Why should
the Raytrack have less bias and more idling current than
similar amps, such as the SB-220? The Raytrack does very
closely meet the Eimac specs for idling current. Why do
others get by with less?

17. Added in-rush protection for filaments; Keystone Carbon
CL-90, Digi-Key KC009L. This is very easy if amp is wired
for 240 V: removed strap at terminal block and attached
device in place of strap.

18. Plan: Add filament transformer primary fuse, 2A.

19. Plan: Add cathode fuse, 1 A, in series with relay

20. Added 100 ohm 1/2 watt and 0.01 disc in series across
relay coil to protect new PTT circuit.

21. Note: output 1250 watts.

C. Grid Metering Mod

D. PTT Relay Mod