Some of you may remember a bunch of questions I asked about an Amp
International LK550-ZC I was repairing for a friend. This report may be of
interest to anyone who needs to repair one of these amps. I should point out
that I got a tremendous amount of help from Carl, KM1H, and Billy at Omega
Electronics.
To recap, my friend and his wife had been estranged for 3 1/2 years. While
he was out of the house, his ham radio equipment sat idle. The couple
reconciled, and when my friend moved back all his ham gear worked except the
amp. When he turned it on, the fuses blew. I took the amp and tested the
three 3-500Z tubes one at a time, quickly finding that one of them had a
serious gas leak (beautiful purple light show.) After replacing that tube,
the amp tuned up OK except on 40M, where I was unable to load the amp
properly and output was erratic. I also found that, on all bands, the plate
and grid current readings were significantly lower than they should have
been at full output power. Finally, there was a nasty crack and some deep
gouges in the plastic cover of the plate meter, a deep gouge in the metal
insert in the Load knob, and some deep scratches and gouges on the right
front and side of the amp.
Soon after the first tune-up pass, the plate choke burned up. When
inspecting the amp earlier, I had not realized the significance of some
bunched up turns on the plate choke. Evidently, the arcs from the bad tube
had damaged the plate choke enough for it to become resonant on one or more
of the ham bands, which resulted in the insulation melting and smoking.
Members of this reflector recommended that I install a glitch resistor to
prevent further damage from such tube accidents. I ended up installing a
15-ohm 25W Ohmite wirewound tubular enamel-coated resistor in the HV line
just after it comes off the HV board. There was a 2-3 inch wide corridor
between the HV board and the case that was an ideal location for the
resistor. I used a pair of wire clip mounts inserted into the resistor body
and I bent slightly to prevent rotation. I mounted the clips to a pair of
ceramic standoff insulators I installed in the floor of the amp. I soldered
the HV line to the resistor lugs, which point straight up. The combination
of the ceramic mounts and the wire clips avoids the problem of not being
able to securely screw both the resistor lugs and solder lugs to the
insulators without breaking the ceramic, and eliminates any problem with the
wire clips being grounded to the chassis.
After installing the glitch resistor, I turned the amp on and the cheap
cement resistor in the step start circuit blew. I replaced it with a higher
quality enamel-coated resistor, and when I next turned the amp on, I saw a
bright purple flash from the tube compartment. I was able to turn the amp
off before any damage occurred. On Carl's recommendation, I used my old
SB-221 to retest the tubes (after installing a glitch resistor in the 221.)
I don't have a variac, so the 221's low-HV CW setting was preferable for
testing the tubes. I quickly found that a second tube from the LK550 was
leaking, providing yet another pretty purple light show. Unfortunately, the
resulting tube arcs blew a transistor and zener diode on the electronic bias
switch (Rich's design) that I had built into the 221 for QSK. Just a couple
of bucks worth of parts, though.
Under the circumstances, the seals on the third tube couldn't be trusted.
The second bad tube had worked for a while before blowing up, so I figured
it might only be a matter of time before the third tube went. I felt that
the gas leaks had to be more than coincidence and concluded that either the
tubes were from a bad lot or had sustained some sort of physical damage
(read on to find out more about that.) Either way, all three tubes had to be
replaced.
I replaced the plate choke with a new one obtained from Omega Electronics.
It had at least 1.5 inches more winding on it than the old one, evidently
because the original choke was designed to work on 11M.
I traced the inability to load on 40M to a broken lead on the switched fixed
capacitor in the output circuit. The lead was actually broken off at the
solder joint. When I saw that, I became suspicious about what had happened
to this amp. The broken capacitor lead, two gassy tubes, cracked meter
cover, gouged load knob and numerous scratches/gouges led me to conclude
that the amp had sustained a sharp physical blow. The owner finally admitted
that his wife probably threw something at the amp after an argument three
years earlier. To me, it looks more like it fell or got pushed off the
table. Anyway, that convinced me that a sharp blow to the amp compromised
the tube seals, broken the 40M capacitor lead (perhaps cut to begin with)
and did the cosmetic damage.
The erratic output on 40M was puzzling. Everything would look OK until the
output got above about 750W. Then the plate and grid current would suddenly
drop to nearly zero, even though an external wattmeter read 750W! I was
really puzzled by this until I happened to slip a Bird wattmeter into the
line. It showed nearly zero output power. Since the other external wattmeter
was a peak reading unit, it was immediately clear that the amp must be
pulsing. Then I noticed a faint pulse in the "on air" light in the FT-990
transceiver. The problem was due to RF leaking into the rig via the T/R
line, which runs through a homebrew Yaesu band decoder box. A couple of
bypass caps in the homebrew box solved that problem.
The low meter readings were due to the meter calibration resistors. I
discovered that someone had replaced the original grid current calibration
resistor on the ALO board. The installed resistor was twice the correct
value and the solder joints showed evidence of a non-factory replacement. My
friend knew nothing about this, so it was probably the previous owner, a
contester well-known for fiddling with his amps. Perhaps he got tired of ALO
trips on 10M and thought changing the grid current calibration resistor
would prevent that (it doesn't have anything to do with the ALO circuit.)
Replacing the grid calibration resistor did the trick. The plate current
calibration resistor had to be dropped about 18 ohms, probably due to aging
in the shunt resistors (I also had to tweak the grid calibration resistor
from the stock value, but only by 3 ohms.) I calibrated the meters by
feeding a known fixed current through the shunt resistors. Now the meter
readings make sense.
I smoked the parasitic resistors myself with too much key-down tuning time
on 10M while adjusting the ALO trip. Replaced the resistors and hooked up an
electronic keyer so I can tune up safely with dits at 75 WPM and a 1:3 duty
cycle. That works very well. Got that idea from Rich's web site.
I used the amp in the CQWPX CW contest, and it worked fine. Need to install
a resistor to quiet the fan down when it switches to high speed (sounds like
a freight train), and might do well to install some ferrite beads on the AC
lines.
I bought the amp from my friend for $475. The repair parts cost about $100.
I got three full-output medical pulls for $230. They work better in the
SB-221, so I ended up using the two 221 tubes and an NOS I had in the LK550,
two of the pulls in the 221, and one pull as a spare. So, I'll have about
$800 in the LK550 plus a lot of work (which was very educational and sort of
fun.)
73, Dick, WC1M
--
FAQ on WWW: http://www.contesting.com/ampfaq.html
Submissions: amps@contesting.com
Administrative requests: amps-REQUEST@contesting.com
Problems: owner-amps@contesting.com
Search: http://www.contesting.com/km9p/search.htm
|