Hi RATs,

I want to build a SE amp with some parts I dig up. I have two BTU-3 3k
primary rated at 15 watts, one power tranny with the following outputs:
700ct @ 120ma, 6.3v @ 5a and 5v @ 3a. and one UTC choke value unknown at
the moment. Tubes on hand are KT88s, EL34s. I was thinking about connecting
them as triodes. Will this work or should I use different tubes.

Regards,
Ken

KEN wrote:

Hi RATs,

I want to build a SE amp with some parts I dig up. I have two BTU-3 3k
primary rated at 15 watts, one power tranny with the following outputs:
700ct @ 120ma, 6.3v @ 5a and 5v @ 3a. and one UTC choke value unknown at
the moment. Tubes on hand are KT88s, EL34s. I was thinking about connecting
them as triodes. Will this work or should I use different tubes.

Regards,
Ken

700v CT will give you +472v with SS diodes after a CLC filter for the anode
supply.
With triode connection of KT88, and with cathode bias of around 50v, the Eak
voltage
would be about 425v, about right with about 65 mA of idle current.
This is 27 watts per KT88.

If you have 2 x KT88 in triode, there will be a total of 54 watts of
dissipation,
and if the maximum triode efficiency is 27%, which should be the case at 425 v
Ea,
then then you'll get 14.5 watts of output

The OPT you have is 3k primary, and we assume the sec is 8 ohms.
So with two tubes, the load seen by each KT88 is 6k, which they will
work into OK and at low thd.
I am not sure what load gives the maximum power, but 4k to 7k is about right
for KT88 in triode.

If you use a 5AR4/GZ34 tube rectifier, then it will give only about 410v,
and with cathode bias, of around 40v, the Ea-k will be only 370v, and the
max efficiency will be lower, even if the idle current of the tubes is raised
to make it around 27 watts for each tube.
You would only get 10 watts fom a pair in triode.
The load of 6k per KT88 is a little high with such low Ea and high Ia.
But fixed bias would be OK to get the headroom with the higher Ea.

I'd go for the SS rectifier option, and use a resistor of 30 ohms in series
with C1
of the CLC supply, say 2 x 470 uF/250v rated, in series for 235 uF for C1, then
a choke
of say no less than 2H, ( could be more ) and then C2 = 4 x 470 uF, in series
and parallel,
to give 470 uF at the OPT.
The ripple voltage should measure a few mV or less.
The choke of 2H can have a 10 ohm plus 1.25 uF in series strapped across
which makes the L a damped resonant filter with a resonant F = 100 Hz,
or if you live where the mains is 60 Hz, C = 0.87 uF.
The ripple value will be about 15 dB lower, since the main component of the
ripple is 100 Hz.
Chokes of 5H should need no such resonant RC attatched.

The use of a tube rectifier means that C1 of the CLC cannot exceed about 47 uF,

or excessive charge currents will wreck the tube.
It also means ripple at the output of the CLC will be higher, simply because
there
isn't as much C in the filter, so L has to be higher, and since you'd be
pulling
perhaps up to 150 mA from the PS, then if its dcr = 100 ohms, then that's a DV
drop of 15 volts,
and it dissipates 2.2 watts.
C2 can be as higher than C1, although if the amp is turned off, then on again 5
seconds later
then the 470 uF C2 does raise the workload of the tube rectifier for a second
or two.
With a tube rectifier, the cold turn on is gradual with GZ34, along with the
output tubes,
and nothing gets severly stressed.
The GZ 34 is capable of higher current supply than you propose, so as long as
C1 is only
47 uF, the peak currents needed at turn on to charge C2 via the choke shouldn't
wreck the
tube because the choke prevents high charge currents pulsing into C2.

If EL34 are used with the same voltages, ease the anode currents down by 10%
to allow for EL34's lower anode dissipation limits.
I like about 21 watts max with EL34/6CA7.
You should get 12 watts max with 2 x EL34 with Ea = 480v, triode,
or 10 watts with Ea = 410v, which is better suited to EL34 in triode.

Patrick Turner.