wrote in message
...
With the gear hooked to the supply, even with it turned-off, when I
turn on the supply, the voltage transient cleans or erases the
programming in the radios. It's not the radio, as it does it with 2-3
different ones.

Does anyone have any idea how to solve this problem, or if a faulty
component might be causing it?

Theory of LM723 operation (courtesy David Metz):
[An Astron schematic, when reading, is helpful]

The LM723 is a twenty some year old IC designed to be a universal voltage
regulator.
Being cheap and well understood it found in most of our bench supplies and
is the core of the popular Astron series.

Any bench supply consists of a transformer with a secondary of about 18
volts AC,
a rectifier block and a filter capacitor. With no load, the DC voltage
across the terminals
of the filter are going to be 18 to 30 volts.

Under load the voltage will sag due to the impedance of the secondary of
the power transformer. If it sags too far, the supply will loose regulation
and it will pass 120 Hertz ripple (hum) to whatever you are powering with
it.

Since the voltage across the filter capacitor can vary wildly as the load
changes
we need a method of regulating the voltage. The work of regulation is done
by
the pass transistors (Q2 to Q5). They in turn are controlled by the 723
regulator.

Lets look at the 723 in a little more detail now. Note that it has its own
filter power
supply made up from diodes D1-D4 and filter capacitor C2.

The 723 has its own internal highly regulated voltage reference supply (pin
6).
Internally the 723 compares this reference voltage to the output of the
power supply.

Voltage adjustment pot RV1 sets the "ratio" of the reference voltage to the
output of the supply.
This sets the output voltage from the supply.

The output of the 723 is pin 10. This voltage drives the base of Q1. Q1
acts as a simple amplifier to increase the current available to drive the
current hungry bases of the pass transistors.

As the load increases on our power supply, the voltage from the wiper of
RV1 drops as well.
When this occurs the 723 increases the voltage from pin 10 driving the pass
transistors harder through Q1.
The more base current through the pass transistors, the higher the output
voltage.

Thus we have regulation, the 723 continuously changing its output voltage to
meet the minute changes in load on the
supply. Remember, the 723 does the thinking, the pass transistors do the
work!

THE PASSING LANE
Let's look a bit at what the pass transistors do. The key thing here is
current carrying capacity and heat dissipation.
In our example, lets say your power supply uses the common 2N3055 NPN power
transistor.

This transistor can pass up to 15 amps of collector current. Note that it is
not wise to
run any device at its maximum rating. Thus for a 25 amp power supply you
would want to have more than
two 2N3055's in parallel. For the last supply I built, I used four.

If you used higher current rated transistors, you could get by with less of
them.
If you simply placed the multiple pass transistors in parallel,
slight differences in gain between them would cause unbalanced current flow.

The highest gain transistors would hog most of the current flow and most
likely fail.
To balance the current small value resistors called "Emitter Ballast
Resistors" are
added in series with the emitter lead of each transistor.
These can be between .01 to .1 ohms and should be of at least a five watt
dissipation.