Low-Noise Instrumentation Power Supply
For as long as I have been playing with
high-end receiving equipment, and in
particular the ancillary, or so-called
accessory equipment, I have been
made painfully aware of the need for
very quiet power sources - read, ultra
clean DC power ... power as "battery
pure" as possible.
The 19-inch rack panel of the power
supply chassis with left and right views
There's nothing special or unusual
about the layout, which I've kept as
simple and basic as possible.
However, behind the panel, there's
more than meets the eye.
I've decided to put a line monitor on the
panel, since this information is very
important to the proper functioning of the
Note that the Ground Fault condition can
be simulated in order test the reporting
In the building of the PLC-1 (at the bottom of this page), I added a full battery of power line filters. I made these filters so that they
could be switched in and out of the circuit in order to be able to evaluate their need, effectiveness and utility. The panel at the left is
from that design. In order to keep the PCU/PCR front panel uncluttered, I'll probably forego the switches and just hard-wire the filters
Differential and Common mode filtering to eliminate both conducted and radiated power line noise. The Pi network removes induced
noise and the "H" (double Tee) filter removes circuit, or conducted noise.
The "C-Shunt" is just a high efficiency lowpass filter using ceramic and mica capacitors. The Gas Discharge tube is a neon tube
wired for 140 volt conduction. I have seen this tube flash during summer storms. It's interesting that transients of that duration can
come into your home on the power lines !
I also added MOV and TVD devices. The TVD or TVS supressors are faster than MOV (varistors), but the two behave in different
ways, particularly in terms of speed. Just as with LP filtering using capacitors, it's a good idea to cover the transient (timing) capability
over an area of interest in decade fashion. For example, when making a DC line RF-proof, I generally parallel the bypass capacitors
in the following values; 0.1, 0.01, 0.001, and 100nF.
Noise comes in any imaginable form, frequency, duration, and wave shape. It is important, therefore, to respect the spectral content
of the noise envelope. One important rule is to evaluate the rise time of a pulse, since it defines the total spectral bandwidth of the
pulse. I use various generators to measure the effectiveness of these filters, like Van de Graaff, Tesla, variable DC brush motors,
etc. but it would be impossible to simulate all possibilities. Therefore, it's simpler and faster to approach the subject with broadband,
The basic regulator circuit, not
showing the noise filters.
This device will feed the Preselector, Multicoupler, Demodulators, Audio Mixer and
Power Amplifiers and anything else in the RF or AF chain of accessory gear.
Some snaps of the finished PSR-1
unit, sitting in a rack along with the
PLC-1, Power Line Conditioner.
Note that the regulated voltage is adjustable from about 11 Volts to 14 Volts (knob between
the meters), and the Ampmeter indicates almost 1/2 Amp.
The Power Line Conditioner reads 120 VAC at about 1 Amp.