Thread: Kenwood KW55
View Single Post
  #3   Report Post  
Posted to rec.audio.tubes
Omer S Omer S is offline
external usenet poster
 
Posts: 7
Default Kenwood KW55


"Patrick Turner" wrote in message
...
On Jul 10, 4:31 pm, "Omer S" wrote:
Anybody out there work on one of these?

http://www.mcmlv.org/Archive/HiFi/KW55U.pdf

I'm about to endeavor restoring one of these, and I'm wondering if anybody
did any mods to this receiver.

Omer


Hi Omer,

I once completely restored and improved a tubed Kenwood AM-FM receiver
in about 1996. A guy here bought one made for the Japanese market
which had TWO am turners because in Japan when stereo FM began they
simply put one channel in AM because they only had one mono channel at
first. There were tuning eyes for the tuning because you couldn't tune
the AM channel for FM stereo by ear with both channels running. The
second AM tuner was provided so that once the FM was tuned on
favourite you could leave it untouched then switch to listen to the AM
only station. It was all very complex and had no MPX converter for
producing FM stereo from a multiplexed carrier with supressed 38kHz
sub-carrier. I installed a kit which had a chip for MPX, and removed
the extra AM tuner used for a "stereo" FM channel. The spare sockets
were used to make a much better AM detector for the AM section which
my customer wanted to be equal to the best available with 9kHz of
audio bandwidth, 9kHz whistle filter, and low THD/IMD.

Your Kenwood is a different circuit to the one I worked on. But if you
look at Pioneer, Fisher and Kenwood, there is much similarity in the
AM and FM front ends, IF strips, and diode detectors. Yours has tubed
MPX decoding but has the usual diode matrix for L&R channel
production. Some of these things work well, and some work very badly,
and to align these sets and keep them aligned and running as quietly
as possible is a labour of love.

Previous to my work on a cutomer's AM/FM Kenwood, I had bought A Trio
receiver with single ended audio output stages with a single 6BQ5
each. I gave up trying to get the MPX decoder to work properly after a
week of ****ing around. I visited local university and college library
archives where I found old books with the theory and basic circuits
which were followed by many makers including Kenwood for the days
before tubes were retired. I copied and studied many pages from books
and applied what I began to understand when I redesigned the Trio. I
got rid of the audio output stages off the Trio and made the chassis
just for AM and FM stereo. The MPX decoder is somewhat a critical item
if you want good imaging, low noise, and low distortions. In the days
of tubed MPX to comply with the GE-Zenith format of stereo
multiplexing, the first MPX decoders propsed and recomended by RCA
used 3 tubes, and Quad made add on addaptors with just 3 germanium
transistors. Usually such primitive MPX circuits offer increased noise
and distortion when using stereo compared to just having mono
function. In all my experience since 14 years ago I have found most
generic MPX decoder chips work better than any arrangement of tubes
and tuned coils and diode matrixes. I built what is probably my final
version of an MPX decoder based on the Quad transistor circuit knowing
that with tubes instead of crummy germanium transistors there was a
chance I could get better dynamic range performance and less
distortion. But the noise is still worse than most chip sets. It sure
sounds well though. The filtering of the switching noise and remnants
of 38kHz and filtering of the 19kHz pilot tone take real expertise.
Many tubed AM-FM sets can have wonderful looking schematics which
withstand visual scrutiny but when you make the measurements which you
NEED to make with this sort of gear they just don't always measure
very well at all.

Measuring AM/FM radio performance requires that you KNOW how and why
each tube and R and C and L has been used where it is. When I set out
to learn enough to service any AM receiver dumped in my bench for a
fix, I was alone. I did know a few fellows in the local amateur radio
group. I used to go to their swap meets and sales days and found them
to be a nice lot but when I mentioned FM, one guy retorted, "****,
FM?, Hell, we are always tryin to keep the frequency stable, and not
let it vary". I didn't get anywhere with my questions about FM radio
of any type. Some amateurs were using narrow band FM, but none here
did. Also the Army used narrow band FM at around citizen radio 27MHz.
It was secure in 1955, and I have a beuatifully made transceiver from
that time, needed a jeep and two guys to get it around. But now is all
entirely a digital world for the armed forces, and not a tube to be
seen anywhere.

Tubes were at about their limit when TV came along and with FM stereo.
Once you have 20 tubes on a chassis all doing things that are high-
falootin and not easily understood by an average man in the street,
and you have a tendency for hot runninng bits and peices to
malfunction in circuits designed by been counters then you find the
Kenwoods et all will need to be serviced once a year. After tubes came
discrete circuits using individual silicon bjts which suddenly became
very reliable and cheep. The myriad of slug tuned LC coil units in
cans suddenly became much smaller. This suited FM but usually made the
sound much worse in AM. Then came the integrated circuit, and these
mostly worked better than anything with discrete bjts or tubes.

I have tried to find out exactly what is inside a typical MPX decoder
chip which performs the stereo decoding function so I could emulate
the circuit using several twin triodes, but I never found an exact
schematic of what actually does the business. Whatever it is, it works
better than a diode matrix, even though that is simpler. I recall
there were things like a pair of cross coupled differential pairs. But
exactly how they were arranged I never found out.
Tim DeParavicini put out a schematic for a "tubed MPX decoder" but
what it really contained was a 14 pin generic chip with a tubed output
stage. It was anything but a fully tubed MPX decoder and one which had
no diode matrix. Nobody I know has ever produced a fully tubed MPX
decoder without diodes, even tube diodes.

I have a few thoughts at my website....
http://www.turneraudio.com.au/am-fm-...ex-decoder.htm

Farnarkling around with AM-FM radios can occupy a man for weeks,
months or years. I spent lots of time on all that but rarely have any
need to use my knowledge which is now a bit rusty. I am now not sure
what a "quieting factor" is. But I do know one had to use a lot of
very good limiting to make FM reception quiet. The "limiter" stages
using 6AU6 are very important, and two such stages are best, although
usually there is often only one. The limiter stage is an SE pentode
gain stage which operates normally like any linear gain stage which
has maybe 30dB over drive. It works in a state of being wildly over
driven and any changes in IF frequency must not cause any output
amplitude changes and where any amplitude changes to the IF exist,
they are to be eliminated as much as possible along with any noise. FM
is thus a quiet medium compared to AM. But you can make a heck of an
effort to get an FM receiver to work and still find there is an
annoying rustling noise from a receiver, and that the SNR is only a
lousy 45dB.

Its usually easy to get good mono with low noise from a tubed FM
tuner, but more difficult if you add the MPX unit.

Patrick Turner.



Hi Patrick,

Fantastic and excellent work!

Although I didn't get that far yet with the KW55, I will keep your web site
as a referance for later.

Right now, I like to replace the usual suspects, lytics, paper caps, and
carbon comp resistors. The diodes seem to be OK (FR1K), but I'm thinking
of seeing how Schottsky diodes work. I don't know if you came across these
before or used them, but what is your opinion of using FRED's instead of the
standard 1N5408 diodes?

http://ixdev.ixys.com/DataSheet/DSEI12-12A.pdf

http://www.allelectronics.com/make-a...T-E3/45/1.html

Any mods you'd suggest for the power amp section, resistors between the
EL84's?

thanks

Omer