[Central-Electronics] Solid State Power Supply for 100V

Sat Feb 26 10:16:22 EST 2005

Pete,

   Your right about this because it is pretty low level. It is easier to
hear on a hi-fi set because the frequency response is much broader, and the
dynamic range is much larger. You can see this stuff on a spectrum analyzer,
and a scope. That does not mean that WD5JKO can hear it when listening to
SSB signals on 80 meters. I never said this was a problem on the 100V. There
are some differences when going to solid state however, and here is a
partial list:

Immediate power up from turn on in a few line cycles versus whatever a tube
rectifier takes to warm up (2 - 15 seconds?). This means the B+ is there as
the tubes begin to warm up, and draw current. Many folks say that delaying
the B+ is better for the tubes. That said a 5U4 rectifier will warm up much
faster than most any indirectly heated tube, so have we been doing it wrong
since they invented the type "80" rectifier?

Higher peak voltage after turn on before the rest of the tubes warm up. The
filter capacitors better be able to handle this. This is a big concern where
the original power supply may have been designed in the days of 110 vac line
voltage. So plug your boat anchor into today's 125 vac outlets, and then
solid state the power supply... You see my point. With the 20A, 114 vac
inlet gives 6.3 volts to the filaments. I see about 6.9 volts on my
filaments. The original 450 volt DC electrolytic can is over rated with the
5U4 rectifier with 125 vac line voltage. I bet the can capacitor would
squirt capacitor juice everywhere if I put solid state diodes in there with
the original rated filter capacitors.

Higher B+ after the tubes warm up than before. This will shift the DC
operating points of various circuits, and might require circuit
modification, or adjustment to keep things operating within specification.

Momentary shorting out of transformer every half cycle by diode reverse
recovery time resulting in lower efficiency, and high dv/dt spike that might
if not suppressed punch through a diode and short it out. It might also
compromise the insulation within the transformer leading to a failure. It is
easy to suppress this spike, and using a faster recovery diode is advisable
to lower the reverse recovery time. Faster diodes however are usually less
rugged. There are all sorts of new diodes these days designed for switch
mode power supplies. One such example is the "Hex-Fred".

Choke input filters also need a R-C filter to suppress the energy when the
choke releases it's stored energy. If unsuppressed, the path is right
through the diodes in the reverse direction. This kills the 1N400x (x=1-7)
variety of diodes. A "controlled avalanche" diode is better if you want to
omit the R-C.

The power up current surge will be higher; maybe by a lot. This might
require you to switch to a slo-blo fuse of the same amperage to avoid
occasional blow outs if you happen to switch the rig on at the peak of a
given line cycle. A two step circuit with a relay is sometimes used on
higher power equipment like Linears to get around this problem.

If you switch over, and have a left over 5 volts AC filament winding,
consider using it to "buck" the incoming AC voltage by putting the 5 volt
winding in series with the primary winding. When properly phased (50-50
chance) this will lower the power company 120 volts AC to 115 AC, and might
cool things enough to be a near complete solution. The 5 volt windings are
either 2 or 3 amp rated, so this won't work on bigger rigs where the AC
input current (RMS) is much higher than the 5 volt winding current rating.
This approach is ideal for a 20A, and many boat anchor receivers.

	So I suggest that anyone considering switching to solid state consider
these issues before doing so. A little planning might pay off with a better
conversion.

Regards,
Jim Candela
WD5JKO

-----Original Message-----
From: central-electronics-bounces at mailman.qth.net
[mailto:central-electronics-bounces at mailman.qth.net]On Behalf Of peter
markavage
Sent: Thursday, February 24, 2005 7:44 PM
To: central-electronics at mailman.qth.net
Subject: Re: [Central-Electronics] Solid State Power Supply for 100V

Quote Jim - "There are those who believe that a power supply has an
audible effect on how an amplifier behaves....."
I won't dispute what you say here but I've used my 100V on both AM and
SSB with these solid-state devices for years. I've never observed any
unusual ripple on the carrier nor the audio component. Nor have received
any reports at the other end indicating my audio had some unusual
characteristics.

Pete, wa2cwa

On Thu, 24 Feb 2005 18:56:40 -0600 "Jim candela" <jcandela at prodigy.net>
writes:
>
>
>         Peter,
>
>         There are those who believe that a power supply has an
> audible effect on
> how an amplifier behaves. By amplifier I refer to audio amplifiers
> with a
> tube basis. Push pull circuitry is more forgiving than high end
> single ended
> amplifiers when it comes to the source impedance (Rs) of the power
> supply.
> Lets say from DC to 1 Khz the power supply Rs is 350 ohms, and
> decreases
> steadily above that frequency as the filter capacitance reactance
> decreases
> with rising frequency. This will continue until resonance occurs
> where Rs is
> near zero, and then starts rising again. All of this can occur
> within the
> high fidelity 20-20,000 hertz audio range. Any ripple on the power
> can also
> have noticeable impact on the sound as well. One source of ripple is
> the
> solid state rectifiers. This is where in  a dual diode (FW-CT)
> circuit the
> diodes themselves at crossover simultaneously conduct. This
> momentarily
> shorts out the transformer, and when the reverse biased the diode
> finally
> gives up, and the transformer releases a big voltage spike large
> enough to
> blow many diodes. I've seen the 1N4007 variety where the overlapping
> period
> can be several hundred micro-seconds. A suitable R-C snubber can
> gobble this
> up to no ill effect. I have an old W6SAI Radio Handbook where this
> approach
> was suggested. I think that was the 1970 edition. Another source of
> ripple
> where the audio amplifier modulates the power supply due to the high
> Rs of
> the power supply. This can be very apparent with heavy bass notes
> that can
> drain the power supply maybe 20% in a few cycles at 20 hertz.
>
>         The audio amplifier example closely parallels what we're
> doing with
> the 100V with the 6550's. The big difference is parallel operation
> at RF
> instead of push pull at audio. We still wiggle the RF at an audio
> rate
> (SSB), so this is still is applicable. For those interested, here is
> how I
> tackled this issue on a Sherwood audio amplifier (tube, circa 1962)
> using
> power supply refinements to try to lower Rs to near zero across the
> audio
> range.
>
>
> http://pages.prodigy.net/jcandela/Sherwood_S8000/HV_REG/regulator.ht
> m
>
> Regards,
> Jim
> WD5JKO
>
>
> -----Original Message-----
> From: central-electronics-bounces at mailman.qth.net
> [mailto:central-electronics-bounces at mailman.qth.net]On Behalf Of
> peter
> markavage
> Sent: Thursday, February 24, 2005 4:25 PM
> To: central-electronics at mailman.qth.net
> Subject: Re: [Central-Electronics] Solid State Power Supply for
> 100V
>
>
> I solid-stated my 100V years ago. Found some old and weak tubes
> with
> octal bases, broke the glass, cleaned out the base, epoxy a small
> perf.
> board in it, mounted some 1N-4007's or something similar (bag of 100
> for
> pennies at a flea market), and plugged it in. Been working ever
> since. No
> variac, no thermistors, no power resistors, etc. Most of my rigs
> have
> something similar.
>
> Pete, wa2cwa
_______________________________________________
Central-Electronics mailing list
Central-Electronics at mailman.qth.net
http://mailman.qth.net/mailman/listinfo/central-electronics
--
No virus found in this incoming message.
Checked by AVG Anti-Virus.
Version: 7.0.300 / Virus Database: 266.4.0 - Release Date: 2/22/2005

--
No virus found in this outgoing message.
Checked by AVG Anti-Virus.
Version: 7.0.300 / Virus Database: 266.4.0 - Release Date: 2/22/2005