[Boatanchors] Resonant Choke Filters (was Please review &

[Rob Atkinson]

On Mon, Jan 5, 2015 at 7:37 AM, Rick Poole WA1RKT <wa1rkt at comcast.net> wrote:
> Thanks, Gary.  But, I'm looking for something I can look up on the web.  I'm just spoiled by all this here modern day technology, I guess.
>

That's your problem.  You are looking for a quick answer.  If you are
going to be serious about vintage equipment, you need to build a
vintage reference library of quality authoritative information.  The
internet is full of garbage requiring a mental crap filter, but that
can only be developed after reading texts.  Any moron can throw junk
up on a website.  Go to hamfests, look for books by George Sterling,
Fred Terman, old west coast Radio Handbooks edited by Orr and earlier,
old RCA tube manuals, GE tube manuals, QSTs and other radio
periodicals from before 1960, ARRL handbooks pre-WW2, Radiotron
designers handbook (the Red Bible) and if you see a copy of the IT&T
Reference Data for Radio Engineers, grab it.   There is usually at
least one person at a hamfest with boxes of books and 95% of hams
ignore it because they don't know what end of a soldering iron to pick
up.

Don't overlook Electric Radio, probably the best English language
hobby radio periodical published today.  Get all the back issues of
that you can.  Speaking of ER, I see in issue number 94 from February
1997, an article on the design of the Collins 30K by the guy who
designed it, Warren Bruene.  Let us look at page 13 on the power
supply from which I will quote:

The high voltage DC power supply was rated for 2500 VDC and 300 ma.
It employed a conventional single-phase full-wave circuit using a pair
of 866A mercury vapor rectifier tubes.  A two section DC filter was
used using a 12 Hy filter choke and a 2 mfd filter capacitor in each.
In addition, the input choke was tuned to 120 Hz with a 0.10 mfd
capacitor.  At that time it was common practice to use a swinging
choke in the first section whose inductance varied from 25 Hy, with
just bleeder current load, down to 5 Hy at full load current.  Collins
engineers had been using the tuned choke idea for several years
because it reduced hum better and had better transient response to a
varying load such as the keyed final.

With an untuned input choke, there is a minimum "critical" inductance
required to prevent the power supply from starting to act like a
capacitor input filter.  When this happens, it causes the no-load
voltage to rise resulting in poorer regulation.

The maximum load resistance is just that of the bleeder resistance
(100,000 ohms) when the final is keyed OFF.  The value of critical
inductance in the single phase full-wave rectifier circuit is Lc =
R/1130 = 100,000/1130 = 88 Hy.

Instead of an 88 Hy choke we tuned the 12 Hy input choke to 120 Hz to
achieve the same result and with a much shorter transient
time-constant.  The choke is rated for 12 Hy at full load current of
300 ma.  The inductance of this choke is approximately 18 Hy when
passing just the bleeder current of 25 ma.  Therefore it takes a
capacitor of only 0.1 mfd to tune it to 120 Hz using 60 Hz primary
power.  The best value can be found experimentally by measuring the DC
output voltage using different values of tuning capacitance.  The best
value produces the lowest DC output voltage.

The transient voltage across the choke may be up to double the DC
output voltage.  That is one reason why this capacitor has a DC
voltage rating of 5000 volts.  The second reason is that there is a
high value of 120 Hz current circulating in the resonant circuit.
This creates a significant amount of heating.  It has been our
experience that a paper dielectric filter capacitor with a DC voltage
rating of twice the DC power supply output voltage is sufficient.
Plastic film capacitors available now heat less.

See what you miss only looking online?  Subscribe to Electric Radio today.

73

Rob
K5UJ