[AMRadio] Re: need help

Jim Candela jcandela at prodigy.net
Sat Oct 22 19:49:58 EDT 2005

Brian said,

> Of course you could use a bipolar transistor with
> it's base held 
> at the appropriate voltage of  130V DC if it had a
> good
> heatheatsink could handle the 104 watts.

Reply from Jim Candela:


    Your zener idea has merit, and this option was a
lot harder back when a lot of tube circuitry was being
designed well before I was born. It's pretty easy to
make a power zener using a small zener, resistor, and
a transistor. One such example I did is at this link:

The approach here with a -100 volt bias supply uses a
power NPN transistor as the shunt regulator. As a
bonus the collector is grounded, so there is no need
to worry about isolating the transistor from the
chassis, or heat sink.

   Now back to your 4-400 linear with a 130 volt zener
using the booster transistor approach, 104 watts is a
lot to manage thermally, and to keep a transistor away
from secondary breakdown (V-A). Power FET's are free
of secondary breakdown, so that is another

   If you want to go down this road, follow the idea I
did with my Fisher audio amp. For the 130 volt
regulator, use the circuit at the link. To get 130
volts use four 33 volt 1 watt zeners in series, and
for the transistor use two to four TIP-50 transistors
in parallel (number depending on thermal demand, and
secondary breakdown V-A demand). Depending on ambient
temperature, and heatsink, you might need a fan. Those
computer CPU heat sinks attached fans are ideal for
these kinds of projects.

   If you wanted to lower the zener burden, maybe
parallel the power zener circuit with a 25 watt 130
volt light bulb. The light bulb is like the 12AX7
filaments used in my Fisher. The idea would be to have
most of the 4-400 idle current (100 ma per 4-400) flow
through the bulb, and for the zener to shunt
additional current whenever you kick up above idle

   You could also have a floating screen supply that
is referenced to the cathode (filament xfmrxfmr.)
instead of ground. This will eliminate screen current
from flowing through the power zener and allow you to
go back to 750 volts instead of 880.

   The power zener will need to be AC bypassed. The
capacitance you choose will determine a time constant
where the amplifier will have no bias at the instant
power is supplied. You don't want too wait too long!

    I still don't think the idea is too "mainstream",
but "what the yeck it might work.


More information about the AMRadio mailing list

This page last updated 22 Oct 2017.