It’s not often that genuine “high-end” audio equipment crosses my
workbench, so I’m sharing my experience of the special occasion. I’m
not going to wax poetic about the “tube sound” but I will talk about
the design and my efforts toward restoration.
You’ll find photos at
http://picasaweb.google.com/circuitsmith/ConradJohnson
The subject at hand is a pre-amplifier that appears to have been made
in the early ‘80s (1982 or later). It uses vacuum tubes in the signal
path and high voltage rectifier, a hand full of transistors to
regulate the B+ voltages and IC regulators to supply the tube
filaments with DC voltage.
The reason it crossed my path is a small child was playing with it;
repeatedly turning it on and off. The resulting hot-switching
transients opened up several resistors in the HV power supply and blew
one of the two filament regulators. I also found the output mute
circuit inoperative and a shaky mod job of replacing capacitors.
Thankfully, the owner had a schematic.
Clearly the designer gives high priority to supplying clean DC power
to the amplifiers. Each cascode triode stage has its own R-C filter (8
total) after the voltage regulator. These filters take up a sizeable
chunk of the main PC board. The 12X4 rectifier delivers a nominal
output of 510VDC. Seems quaint in this age of MP3 players that run on
one AA cell. Unfortunately the rectifier filament is not on a separate
transformer winding, so the heater-cathode voltage is also 510V, far
in excess of the 100V average 400V peak max rating listed in my RCA
tube manual for the 12X4. This is one oversight in an otherwise
conservative design. Line level, phono and filament grounds are
loosely coupled with big Rs and small Cs.
The smoothing capacitors were originally two 200uF 350V caps in series
but whoever “re-capped” the unit replaced them with a “universal”
60-90uF 500V cap in parallel with one of several 20uF non-electrolytic
and put a 2K resistor in series with the rectifier cathode to reduce
the voltage (to 430V nominal) and keep from over stressing the cap.
Unfortunately this nearly eliminates the regulators ability to
tolerate low line voltage. I restored the original configuration with
two Nichicon high temp (105C) 220uF 350V caps and bleeder resistors. I
like the idea of some extra resistance to limit cathode current of the
rectifier since it feeds such a large smoothing capacitor (10-20uF
would be run-of-the-mill with a 12X4) so I replaced the 2K with a 649
ohm 3W resistor. Now the HV regulators deliver clean DC down to a line
voltage of 104VAC.
The first HV regulator uses a stack of zener diodes, fed a constant
current, plus an emitter follower to deliver 410VDC. Four R-C filters
provide extra smoothing to each high level amplifying stage. The Cs
were originally 60uF electrolytics but have been replaced with non-
electrolytic 20uF units. All the Rs were blown from the switching
transients. A couple of the zeners were shorted and reduced the output
voltage to ~350V. The second regulator starts with the 410V and uses
more zeners and an emitter follower to deliver 380V to the low level
phono amplifiers, through 4 more R-C filters.
The mute circuit uses a relay and a unijunction transistor to short
the output lines to ground for approx. one minute after the unit is
turned on. There is also a mute button on the front that controls this
relay. A wise choice, considering it was able to deliver a 200V peak-
to-peak sine wave without clipping, driven from the phono input. This
circuit had been bypassed however. A timing capacitor became leaky and
prevented the timer from timing out. Evidently, someone mistakenly
changed the relay and miswired the replacement. I restored the circuit
to its original configuration.
Another relay and timer to prevent hot switching would make the unit
more robust and toddler resistant.
Tim Brown
circuitsmith * at * verizon * dot * net