[ARC5] An "Extra Class" Question:

[don davis]

Hi Dave:

I did high power buses for spacecraft for a lot of years, and your modulator
question is similar to a lot of the problems I had to deal with.  Some
ideas:

O  In any application where loads are switched or could be disconnected (ham
environment) I always put a resistive pre-load across the output of ~ 1 - 2%
of design max output to limit peak charging effects of an open xfmr
secondary.  For 100W RF you would have ~ 50W audio, so I would target 0.5 -
1.0 W.  Never operate a power amp without a proper minimum load unless you
know it will survive.

O  Always limit the audio range to the design limit of the system.  You can
get all sorts of funny behavior if you run full power at 5,000 Hz audio for
something designed for comms between 250 - 2500 Hz.

O  I like snubber R-C series circuits to reduce the energy in transient
spikes caused by the modulator or load.  These can be quite effective, but
are difficult to properly design (usually empirically sized) and can burn a
surprising amount of power.

O  MOVs are good -bad - good.  Good that they behave kinda like a fast
variable resistor so you don't have big di/dt that can cause splatter like a
Zener.  Bad that they can only tolerate limited energy before they short out
and take the system down.  And, good that if they short they may save your
xfmr and you can then find root cause.  Must be sized properly for Voltage
threshold and energy capability (Joules).  You can use several of these of
different thresholds with series resistors to make a softer, more survivable
system.  The devil is in the detail...

O  I'm a big fan of brute force Zener / resistor chains to handle large
power events gracefully.  In one system (Hughes Astro/NextSat) I used 8
strings of Zeners with series resistors to gracefully commutate and un-clamp
10kW events lasting several seconds.  Big, hot, expensive, but the notion of
the series resistance to minimize splatter is the key.

O  To avoid operating discontinuously with large negative spikes, others
have mentioned current-fed inductors (which is correct) but the key is to
provide an energy source that will allow it to work through (expensive) or
prevent it in earlier stages (best) or clamp it gracefully (ok compromise). 

O  Make sure all RF bypassing is ok - for obvious reasons.

73 de don ad6pb

-----Original Message-----
From: arc5-bounces at mailman.qth.net [mailto:arc5-bounces at mailman.qth.net] On
Behalf Of David Stinson
Sent: Saturday, February 09, 2013 10:03 AM
To: ARC-5 List
Subject: Re: [ARC5] An "Extra Class" Question:


Sorry, guys.  I was intending to get fresh insights on a problem, so was
intentionally vague, but that didn't work-out so well.  Please let me
explain.

I meant "Extra Class Question" in the sense of it being the sort of question
one might see on the test, not that it actually was.

Here's the situation:  the AN/ARC-5 MD-7 modulator has a G.E. "Thyristor"
across the secondary which damps modulation excursions over 300 V RMS.
I'm rebuilding an MD-7 for a friend which had a blown mod. transformer.  It
was blown because yet another doofuss decided to whack it out of the
modulator, break off the mounts for it and doubtless try to modulate his
pair of  P-P 813s with it.
I've got the modulator restored and worked except that the "Thyristor" and
its mounts are missing.  
It was dumb luck that I found a replacement modulation transformer and
finding another is unlikely, should this one blow from not having the device
installed.

At the moment, I have two series-connect 200 V MOVs across the secondary.  
This is unsatisfactory because that's 400 Volts, not 300.   
Are MOVs rated RMS or Peak anyways?
I probably knew 20 years ago..... 
And can you series them?  Like a 200 and a 100?

Whatcha think?