Good points Hal. I am a firm believer in closing the door on transients
by having a series impedance, line to line HF capacitance, common mode
impedance, line to neutral HF capacitance and finally line to line and
line to ground and neutral to ground spark gaps and MOVs. A $5000
Amplifier is worth $100 worth of solid filtering on the input of the
PSU. Invest in a good Corcom or Filter Concepts LLN filter and some
spark gaps.
Tomm
harold b mandel wrote:
>Working with telephone company power supplies for
>about 44 years has shown me that Will Matney's, (et al),
>comments are dead on.
>
>For me, power supplies were in two usual flavors:
>High Voltage/Low Current and Low Voltage/High Current.
>
>For the most part, the most frequent failures have been
>with semiconductor junctions in switching type supplies.
>
>After that, semiconductors in Class-A (ferromagnetic)
>rectifiers exhibited the shortened MTBF.
>
>What was found was that Power Line Disturbances
>on the order of 1 X 10 e-12 seconds, and sometimes
>with potentials of thousands of volts were trashing
>the junctions.
>
>Heavy equipment on the same line as the rectifiers, such as
>elevator motors, HVAC equipment, assorted compressors,
>all using open-frame contactors were inducing spikes in
>the a.c. mains supply and these spikes just went right
>through the filtration steps and wore down the diodes.
>
>Years ago, when semiconductors were less impervious
>to PLD's, (over-engineered rectifier banks, etc.),
>the frequency of power supply failures was much lower.
>
>As equipment evolved to faster microprocessor times
>and switching supply circuits additional protection
>needed to be added to the power plants in the form of
>MOV's, choke-reactors and gas-discharge devices.
>
>In one circumstance the general manager just didn't
>believe that PLD's were the cause of his frequent
>failures and we were authorized to rent a Power Line Analyzer
>with a strip-chart recorder. Every morning at 0200,
>when the building heating saystem cut back in there was
>a huge spike of around 3,000 volts for only 12 microseconds.
>
>It took a ferromagnetic conditioner with a buck and boost
>winding and capacitor to filter that spike, but it worked.
>
>Hal Mandel
>W4HBM
>_______________________________________________
>Amps mailing list
>Amps@contesting.com
>http://lists.contesting.com/mailman/listinfo/amps
>
>
>
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|