[AMPS] LK550-ZC Troubleshooting (long)

[Dick Green]

Remember that LK550-ZC my friend is looking to sell? It's the one that blew
it's fuses the first time he turned it on after 3 1/2 years of no use. Well,
I finally got my hands on it, and need some troubleshooting help. Here's the
scoop:

1. Lots of dust inside both the transformer and RF cabinets. Cleaned all
that out. The plate meter housing is cracked and dinged, as is the silver
part of the Load knob and part of the case in that area. Some odd-shaped
thing must've fallen against the face plate...

2. I lifted the tube deck to clean underneath it and to clean the lower
fans. Everything looked fine under there, but I did notice something funny
about the plate choke -- it doesn't look very pretty. The windings are kinda
bunched up in a couple of places. The insulation appears OK and the turns
aren't loose -- it's just that some of the turns are over other turns, even
though there's still some room left on the form. I haven't been inside all
that many commercial amps, so I don't know what to expect here -- it doesn't
look like a very professional job to me, but maybe that's normal. More on
this later -- it figures into the punch line of the story.

3. One surprise is that this amp has a soft start board from Omega
Electronics. I almost didn't notice it because it looks so much like the
rest of the boards.

4. The blown fuse problem was pretty simple -- one of the tubes had a leak.
The HV made beautiful purple swirls inside the tube before the fuses blew. I
had suspected this might be the problem -- I saw some tiny shards of glass
near the tube sockets when cleaning, although I could not find any visual
damage to the glass of the bad tube. The fuses didn't blow without the tubes
installed, and it was easy to find the bad one by reinstalling them one at a
time. A quick substitution from my SB-220 and the amp was back up and
running with three good tubes.

5. Tuned up into a dummy load on 20M and got 1450-1500W out (depending on
which of my two meters you believe), with about 100W drive from an FT-990.
HV is a tad over 3000V, dropping to 2600V at key down. The plate current
reading seemed kinda low: 700 mA. I would have figured 850 mA - 950 mA. Grid
current was about 225 mA. That seemed low to me, too. Possible meter
problems?

6. With one exception, tuneup on the rest of the bands went well --
160=1500W, 80=1650W, 40=1500W, 20=1500W, 15=1525W, 10=1325W. SWR was a
little high on 10, maybe something that can be improved by adjusting the
tuned input. On each band, I tuned up in several steps to full power. The
plate and grid dips occurred in step with max power output, so this LK550-ZC
doesn't seem to exhibit the output network problem I've heard about (anyone
know how to visually confirm whether this puppy has the right tank circuit?)
The only thing that bothers me about this is that I thought the amp wouldn't
put out this much power with only 100W drive. What gives? (Read on before
you answer.)

7. The exception was 40. Meter behavior was *very* odd there. Tuneup was
normal until I reached max drive. Then, at max power output, the plate
current meter was showing maybe a couple of hundred mils and the grid
current meter was showing something like 50 mils -- with both power meters
showing a full 1500W output! If I fiddled a bit with the tune and load
controls, I'd get a tad more output power and the plate and grid meters
would suddenly jump up to normal reading and the transformer would buzz like
it was under full load. No change in output power -- still max. I really
couldn't set the amp at that sweet spot -- the meters would drop to those
weird low reading and the transformer buzz would subside. But power output
was still max. I figured something very odd must be happening with the
meters on that band.

8. But then I tuned up on some of the other bands and the meters behaved OK.
I tried tuning into resonant antennas on several bands, and got full output.
Finally, I decided to see whether the amp could load up in the 75M phone
band on my 80M CW dipole. The output network looked as hefty as my old
SB-220 (which tunes into 3:1 fine) and since there are no PIN diodes to fry,
and the amp has an ALO circuit to detect high VSWR, I figured this would be
pretty safe -- WRONG! I tuned up at several hundred watts out from the amp
and increased drive until the output was about 800W. I thought I saw
something out of the corner of my eye -- maybe a thin waft of smoke -- but
thought it was a reflection on my glasses. I bumped the power up to about
900W -- suddenly the output dropped and I saw smoke coming out of the tube
compartment. I let off on the key and looked down through the vent holes at
the tube deck and saw that the plate choke had turned black and copper in
several places where the wire insulation had vaporized. End of plate choke.

What happened here? Why did the plate choke overheat? If I can trust the
meter, plate current wasn't any higher than 700 mA. Could it have been heat
from the tubes (the plate choke does sit smack in the middle between the
three tubes)? All four muffin fans were going full bore at the high speed
setting, so cooling was normal. Is it possible that this plate choke is a
homemade replacement?

Of course, the $64K question is, why didn't the ALO trip? Any ideas for
troubleshooting that circuit? BTW, when I flip from STBY to XMIT, the ALO
light comes on for a couple of seconds, then goes off (a relay or two click
at the same time.) I thought that might be an indication of normal
operation, but the manual doesn't mention it.

I probably have no business theorizing about this, but could the plate choke
have been damaged when the bad tube arced (which happened at least three
times)? Could this have caused the turns to bunch up? Could a few shorted
turns have allowed enough RF to get back into the power supply to cause the
low plate/grid current meter readings on all bands and the very oddball
meter readings on 40M? Extending the theory, perhaps 80M RF getting into the
power supply caused the semiconductors in the ALO circuit to malfunction.
Maybe at the same time the RF-impaired meters weren't showing the full
extent of the plate current (excessive due to high VSWR?) and the RF choke
got finished off the rest of the way. Does this sound plausible? Is there a
simpler explanation (i.e., I screwed up...)

Hopefully, it's just the plate choke -- I hope I haven't fried the grids,
too!

I guess I'm going to need to fix the plate choke first. Would it be
possible, or advisable, to rewind it myself? No inductance value is given in
the schematic, so I'd need to know that. Does anyone know if the choke is
available as an assembled unit from Omega?

Are there any other components I should check for damage? What's the best
way to proceed?

Any help would be *most* appreciated!

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