[Elecraft] ProAudio Engineering KX33 Power Supply

[Howard Hoyt]

Hi Ed,

Response to step load changes is a mostly unpublished parameter of power 
supplies in general.  Both linear and switching supplies have time 
constants in the voltage control circuitry, and both can be poorly 
designed.

Designing a linear regulator can be potentially more simple, as most of 
the components affecting the steady-state and dynamic characteristics of 
the supply are designed into the control IC and a few external 
components.  This takes most of the performance characteristics out of 
the hands of the manufacturer of the supply. Indeed most linear 
regulated supplies like the Astron are mere executions of application 
note designs offered by the manufacturers of the regulator ICs.

The performance of switchers can depend more heavily on design 
execution, as the characteristics of the inductive and filter components 
can affect the regulation.  Our testing has shown the Kx33 running 14.2V 
to have less than 10mV of overshoot when returning to receive mode from 
a 2.7A KX3 transmit load.  At least one inexpensive switching supply 
spiked to over 19V in this same test.  This is caused by the response 
time of the switcher's duty cycle servo to output node voltage changes 
and is not to be confused with its output impedance causing voltage drop 
with load.  We are working on an addendum to the test reports published 
on our website which will show load step response of competing supplies 
along with currently available ripple and static voltage regulation specs.

Cheers & 73,

Howie - WA4PSC



On 1/22/2016 9:51 PM, elecraft-request at mailman.qth.net wrote:

>>I have come to suspect that the MFJ supply created
>>transients which somehow escaped being regulated out.