[Collins] 75S3-C Power switch problem.

[Dr.Gerald Johnson]

Power switch problems are more common with the 32S and KWM-2 
where the line current is higher. When I built my own supply for my 
32S-1, I didn't bring the line switch circuit to the transmitter but 
mounted a toggle switch on the power supply. Its lasted quite well 
that way.

When switching transformers feeding filter capacitors with solid 
state rectifiers, the rotary wafer switch is a poor choice, because 
the filter capacitor (and sometimes the transformer core) takes 
maximum current on the first cycle the contacts touch, even when 
they have a minuscule contact area. That tends to cause 
contact erosion, occasionally massive contact erosion. The wafer 
switch should be operated positively, never slowly. Like the knife 
switch on the very old power panel board, slap it open and 
closed. Another cause of high current at turn on is residual 
magnetization of the transformer core. That's often noticed when 
the transformer goes "tung" when powered. The core has remained 
magnetized at the last polarity before turn off and when 
magnetized has lower inductance than if not magnetized. Worse if 
the first half cycle of power is the opposite direction the primary 
current peak is especially large. And according to Murphy's law 
the first half cycle will be the opposite direction of the last one 
from the last power down.

After checking the silicon rectifiers and filter capacitors for shorts, a 
temporary fix for the receiver is to wire a jumper across the 
contacts and use an external switch like a power strip or 
computer accessory control console with its row of switches.

A new wafer is the proper historical fix, though if I had that 
problem come up I'd be very tempted to create a cam on the 
switch shaft to operate a microswitch snap action switch module 
that would stand millions of operations and handle far more 
current safely than the wafer switch.

One cause of flash could be an accumulation of dirt on the 
wafer's insulator making a path to ground.

Isolating the current/voltage from the wafer would need some sort 
of power supply and a relay. The inductive kick from the relay coil 
can be as much of a problem as the transformer and filter 
capacitor inrush current.

It might be effective to create a step-start circuit to limit the 
transformer and filter capacitor inrush current initially. The 
transformer current can be a little high if on 50 Hz and 125 volts. A 
step start circuit might consist of a negative coefficient thermistor 
such as was used in some TV sets 30 or 40 years ago. Probably the 
thermistor was a disk a bit over an inch in diameter and about 
3/16" thick. It would start out with a relatively high resistance and 
as it heats the resistance would go down to under 10 ohms. That 
makes for a lot softer start. Otherwise the transmitter step start 
often uses a series resistor in the line that is shorted by relay 
contacts as some rectified voltage rises. Sometimes its a special 
relay power supply or it could be a receiver DC voltage, could be 
that receiver DC voltage with a series resistor and a large 
electrolytic across the relay coil to cause more time delay, though 
the inherent 10 millisecond delay of the typical relay lets it close 
after the first cycle of AC anyway. A 0.1 to 0.25 second delay is 
plenty.

A relay step start circuit could be external to the receiver as could 
the thermistor soft start.

73, Jerry, K0CQ, Technical Advisor to the CRA

--
Entire content copyright Dr. Gerald N. Johnson, electrical engineer.
Reproduction by permission only.