This amp was a desperate attempt to get something
playing for the Århus Triode Audition (2000) in Denmark this august. I am currently
waiting on parts for my 71A power amps and thus used the chassis for these amps
to get something up and running. This project is the result of that attempt.
Despite the very simple filament supply of the AV5,
there was absolutely no hum at all in the system we were playing. The KC3 is
also known for microphonics but not even a tad of microphonics
in these amps. I guess something finally worked okay from the start for a
The KC3 is biased around 3V, this means it only
takes about 2V RMS input before drawing grid current and thus clipping. It's
gain of about 20 times gives me enough to drive the AV5's to reasonable power
levels (2Wrms). The AV5 is biased at 65mA with about 49V on the cathode
(Ua=285V). It dissipated about 18.5W and using the 5K load it produces up to 2W
The 50nF/47K filter gives a low roll-off at around
90Hz. This means the amp is not meant to play the low-end of the spectrum. If
you want to play full-range change the resistor to 250K and the cap to 100nF.
The reason I use a high-pass filter is that I use my amps for the Oris system
which only runs from 150Hz and up. The less power I use up in the low end (which
is not audible throught he Oris anyway) the more control the speaker
will have in the region it is reproducing.
Listening to some music with lights dimmed at Bjørn Aaholm's place after
getiing back from the ATF, I noticed an orange glow
coming from the tubes. The plates are shiny nickel so I thought it was a
reflection from the filament off of the glass. A few minutes later when my eyes got used to the dark
I noticed the entire anode was glowing dark orange. Whoah! I quickly hooked up
my meter to check out what was going on. No, everything tested as it should...
18W on the plate, it should handle 20W. A quick check confirmed the other
channel was also glowing a bit, same thing. I contacted AVVT and they said this
was okay, no problem, "they like to run hot". Okay then, let them
play, they sound very nice!
With my Puppy power projects ended I rebuilt the chassis to it's final version
using the AV5. As the setup is almost identical to the first Århus AV5 version
I listed it on this page. I made some big changes in the power supply, just like
with the Puppy Power 3. For the AV5 I reconfigures the use of the four 10uF
capacitors to get as much voltage as I could (more power). What I got was
275Vdc, this is what I have to work with. The AV5 is biased at 30mA (-27V) and
dissipates just a tad over 10W, very nice! The KC3 got a different bias at
150V/2.5mA and a cathode bypass to get a bit more gain in the low end. Not much
else to say, take a look at the schematics. Ah yes, filament supplies are also
current regulated, I replaced the LM317 for the 71A to a LM350 (3Amps) to handle
the 1.3A of the AV5 so that it won't overheat.
I had to put the puppy to sleep. It was just too old and worn out. One
of the AV5 tubes seemed to be playing a bit softer than the other. After some
checking I noticed it's emission had dropped from 50mA to about 15mA in one
channel, and only after a month or two of playing. Tubes went back to the factory, no
problem getting replacements they said. As I am planning a new chassis and setup, the AV5 will be
replaced by a different tube. Back to the good old stuff I guess... UX245.
I'll try to put up new pics as well. The ugly black paint has made way for the nice copper
that was underneath. Looks totally different.
Frequency range -1dB with 1Vrms output: 32Hz -
Frequency range -3dB with 1Vrms output: 23Hz - 43kHz
note: The OPT's were designed and
made with a bottom frequency of 60Hz in mind, the somewhat
"quick" roll-off is a result of this (and thus intended to be
Frequency range -1dB with 1 Watt rms output: 61Hz -
Frequency range -3dB with 1 Watt rms output: 37Hz - 30kHz
High range is a bit less then I
would have expected.
Maximum output before clipping: 1.49 Watts rms.
One of the AV5 amps next to Hartmut Quaschik's turntable.
Here's the AV5 amp with a 31 as driver instead of
Here shown with clip-leads for charging the battery.
In the beginning...
B+ supplies for both KC3 driver and AV5 output tube.
filament supplies for the KC3 and AV5.