[Collins] Re: Need 516F-2 replacement choke source

[jefferis at antelecom.net]

Greetings all,

My hands on experience tuning the 516F-2 resonant choke has
shown the tuning exercise, and the technique to be useful. I
have only done this to 3 units, so I certainly have not seen
the true range of parameter variation. Two of the units
started with no load (actually the minimum load determined
by the bleeder) B+ voltages in the 1050 to 1150 VDC range
(line votage for testing always set to 115 VAC).  The 3rd
was 975 VDC.  Here is my tuning data from the 3rd unit:

C1 (uf)       B+@ min load         B+ @ 32S-3 full load (230
ma Ip)
.033                   1012                              
727
.043                     960                              
726
.057                     880                              
729
.067                     870                              
730
0.1                      870                              
725

Not only does the resting voltage decrease, but the
capacitor values have little effect of the full load
voltage. You quickly get to the point of diminishing returns
and dangerously close to the resonant point, where you do
not want it to operate. With this unit, I backed C1 off to
057 uf. The other 2 units wound up at .047 uf and .068 uf
for optimum voltage sag control. I use 1600 Volt rated
metalized polypropylene caps and I believe that 0.01 uf
tuning granularity is quite adequate.  Oerating temp and
line voltage variation moves these voltages around anyway.

I believe that the resonant choke technique works much like
the swing choke, i.e., you definately want inductange
reduction as load current increases. The trick is make sure
that you do NOT resonate the choke/cap combination. You want
the choke inductance to decrease and reduce the 120 Hz
impedemance as load current increases, and you want the
choke/cap combo resonant sightly above 120 Hz. The real
difference between this and the swing choke is that the
inductance change with load current is highly non-linear
with the conventional choke, whereas the inductance change
with load current is intended to be nearly linear in the
swinging choke, and you you get a wider range of inductance
change. Typically, the swinging choke has to be physically
larger to get the same overall effect. Just my
observations... Cheers.

Bob, KF6BC

----- Original Message Follows -----
From: "Dr. Gerald N. Johnson" <geraldj at storm.weather.net>
To: collins at mailman.qth.net
Subject: RE: [Collins] Re: Need 516F-2 replacement choke
source
Date: Sun, 17 Aug 2008 16:56:56 -0600

>On Sun, 2008-08-17 at 10:53 -1000, pete wokoun, sr. wrote:
>> 
>> 
>> 
>> I built up a 516F-2 'equal' power supply a while back and
>> had concerns  also with that resonant choke circuit and
>> limited choke breakdown voltage.   I don't think that
>> original resonant choke filter works all that well; I
>found a swinging choke worked much better.  You can read
>> about my design  considerations as well as voltage
>> comparisons here: 
>>  
>> 
>>
>http://www.qsl.net/kh6grt/page4/516powersupply/516powersupply.htm
>> 
>>  
>> 
>> pete
>> 
>>  
>Without looking at your web page, I can agree. The resonant
>choke could be good at notching the primary 120 Hz ripple
>frequency at the expense of passing more of the higher
>harmonics to the filter capacitors. If it was resonant at
>120 Hz. My Shure reactance rule says 8 Henries and .05 mfd
>is resonant at 250 Hz. Not much 120 Hz filtering there,
>unless the low current inductance is really 34 Henries. But
>it can only work if the choke's inductance is constant and
>that is hard to accomplish with an iron core and varying DC
>in the choke. It may well be that all it accomplished in
>the 516F-2 was to limit peak currents into the filter
>capacitors, a requirement to keep the 5R4 from dying
>prematurely. Silicon rectifiers tend to accept much larger
>peak currents and not like the inductive kick from turning
>off the load so choke input filters are not used as much
>with them. In the low power power supplies in the 821A-1,
>each input choke was equipped with a selenium plate
>transient absorber to protect the diode strings.
>
>In my own 516F-2 replacement, I used a single high voltage
>winding with a bridge rectifier. I wanted lots of
>capacitance on the high voltage output to maximize the
>voltage and the voltage regulation. I took the low voltage
>off the transformer winding center tap with a choke input
>because I wanted to lower that voltage. It works just fine.
>Silicon rectifiers, of course.
>
>A swinging choke should nearly always give better voltage
>regulation than a fixed or tuned choke or a capacitive
>input filter, unless the capacitance is very large, then
>peak currents wipe out tube rectifiers.
>
>Fact is, given choices in single or double high voltage
>secondaries, silicon or vacuum rectifiers, oil or
>electrolytic filter capacitors, and fixed or swinging
>chokes, there are many possible circuits that will work
>perfectly fine. Its sure that some will show a little
>better or worse efficiency, power factor, voltage
>regulation (efficiency, pf, and regulation may  not
>optimize at the same time), cost, and physical size.
>
>OK, now I've read it. Just a couple comments.
>
>100 mfd 450 volt capacitors were not common, maybe not
>available at all, and  for sure were not small enough to
>fit under the original 516F-2 chassis in 1959. One needs to
>watch their ripple current rating in this application.
>Physically small electrolytics may not hack the applied
>ripple current leading to higher operating temperatures and
>shorter life.
>
>The insulated choke core will float at a voltage determined
>by the voltage division of leakage Resistances coil to core
>and core to chassis and the distributed capacitances of the
>same components for the AC component of the voltage. With 
>differing dielectric properties you might want to control
>that float voltage  by using a couple resistors or a near
>center tap on the three resistors  providing the bleed (and
>minimum load for regulation) and voltage division on the HV
>filter capacitor. E.g. about half supply. Then you might
>want to encase the choke in a housing made of 1/16" clean
>fiberglass to be sure fingers or tongues can't reach that
>floating core even if little tongues from a kitten or
>puppy.
>
>Otherwise, except that it is a bit bigger than a 516F-2 I
>think your supply is  a good design that will run a very
>long time. It is much smaller and prettier  than the one my
>dad (K0CPN, now silent key) built. I call his the 516F-2000
>but no ham at a hamfest has given it a second look. It runs
>cold though at full load. -- 
>73, Jerry, K0CQ, Technical Advisor to the CRA
>All content copyright Dr. Gerald N. Johnson, electrical
>engineer
>
>___________________________________________________________
>___ Collins mailing list
>Home: http://mailman.qth.net/mailman/listinfo/collins
>Help: http://mailman.qth.net/mmfaq.html
>Post: mailto:Collins at mailman.qth.net
>