866 or Damper diodes

866 or Damper diodes

hey-Hey!!!,
If y'all were going to construct a power supply for a big tube amp( B+
~650, current ~350 mA) would you use Hg vapour tubes in the PS or
damper diodes? Seems both have some advantages. Glow is nice for
sure...:)
cheers,
Douglas

Re: 866 or Damper diodes

On Oct 24, 11:10 pm, Multi-grid <pent...@netscape,com > wrote:
> hey-Hey!!!,
> If y'all were going to construct a power supply for a big tube amp( B+
> ~650, current ~350 mA) would you use Hg vapour tubes in the PS or
> damper diodes? Seems both have some advantages. Glow is nice for
> sure...:)
> cheers,
> Douglas

Mercury vapor rectifiers have been obsolete for at least 30 years and
are a needless environmental hazard so its not clear why you would
want to use them The term "damper diode" can refer to either a tube or
a silicon diode and is intended for use in a television. Silicon
diodes represent a much better choice since they are considerably more
efficient. For a nice glow, try a fluorescent light but don't break
it.

Re: 866 or Damper diodes

In article <ffp1fj0idd@news1.newsguy,com >,
Multi-grid <pentode@netscape,com > wrote:

> hey-Hey!!!,
> If y'all were going to construct a power supply for a big tube amp( B+
> ~650, current ~350 mA) would you use Hg vapour tubes in the PS or
> damper diodes? Seems both have some advantages. Glow is nice for
> sure...:)

When I was designing transmitters for RCA Broadcast (late '60's), I did
the analysis that resulted in the replacement of mercury-vapor
rectifiers by a series string of silicon rectifiers in several of their
AM and FM transmitter models. Saved several hundred dollars per
transmitter while improving reliability and efficiency.

None of the engineers there thought that the tubes had the slightest
advantage, and were glad to get rid of them; what do you think the
advantages are?

Isaac

Re: 866 or Damper diodes

"Multi-grid" <pentode@netscape,com > wrote in message
news:ffp1fj0idd@news1.newsguy,com ...
> hey-Hey!!!,

> If y'all were going to construct a power supply for a big tube amp( B+
> ~650, current ~350 mA) would you use Hg vapour tubes in the PS or
> damper diodes?

Neither, I'd use silicon diodes.

Re: 866 or Damper diodes

"Multi-grid" <pentode@netscape,com > wrote in message
news:ffp1fj0idd@news1.newsguy,com ...
> hey-Hey!!!,
> If y'all were going to construct a power supply for a big tube amp( B+
> ~650, current ~350 mA) would you use Hg vapour tubes in the PS or
> damper diodes? Seems both have some advantages. Glow is nice for
> sure...:)
> cheers,
> Douglas

Why not silicon rectifiers? Assuming you're producing DC successfully,
tubes as rectifiers can't possibly make your amplifier sound any better...

Re: 866 or Damper diodes - a definitive response

In article <3eydnRa5Bd2WGu_anZ2dnUVZ_v2pnZ2d@pghconnect,com >,
jwvm <jwvm@umich.edu> wrote:

> On Dec 25, 11:48 am, "MC" <m...@uga.edu> wrote:
> > Ah. I was assuming there would be adequate regulation and noise filtering
> > after the rectifier. If not, then the rectifier does make a difference.
> > But wouldn't a silicon diode with a capacitor across it have "soft recovery"
> > too?
>
> Good power supply design isn't rocket science. With effective use of
> filtering and bypass capacitors, it is not obvious why noise should be
> much of a problem. Even low-cost amplifiers with headphones can have
> no audible sound at low volume settings. At higher volume settings,
> white noise will be present but this is not due to power supply
> failings.
>
> One might note that power supply noise at the output stage should not
> be much of a problem anyway, at least for common collector/drain type
> circuits. The high impedance (Early effect) of output devices
> effectively isolates the power supply from the loudspeaker unless they
> saturate.

I suppose it's possible that "soft" rectification makes a
less-than-competent grounding setup more-or-less adequte. High-current
pulses from "fast' rectifiers could cause voltage drops in unintended
places, producing hum or buzz.

Isaac

Re: 866 or Damper diodes - a definitive response

On 27 Dec 2007 16:46:17 GMT, isw <isw@witzend,com > wrote:
....stuff deleted.....
>
>I suppose it's possible that "soft" rectification makes a
>less-than-competent grounding setup more-or-less adequte. High-current
>pulses from "fast' rectifiers could cause voltage drops in unintended
>places, producing hum or buzz.
>

If you are rectifying 50/60 Hz power, it is silly to specify soft
or fast rectifiers. Due to the low dv/dt (change of voltage with
respect to time), there is little charge "caught " in the rectifier
junction that will cause current spikes on the reverse voltage. If you
have a switching power supply on the other hand, with high rates of
change of voltage, yes, you will have current spikes at the reverse
voltage beginning, as the charge is pulled out of the junction.
I tried this about 10 years ago, with spectrum analyzer, and a
special circuit that is commercially used to test EMI that propagates
from a power supply. It made diddly difference what type of diode was
used. All the harmonic content coulkd be explained purely from the
normal rectified sinusoid. That is for a sinewave supply.
What DID make a difference was leakage inductances and stray
interwinding capacitances. Those were controlled by RF bypass caps to
the chassis, series bifilar inductors, and most importantly, a 0.1 mf
capacitor in series with a 22 ohm resistor, and this combo placed
directly across the power transformer primary. That circuit would
lower the Q and dampen most of the resonances. The type of capacitor
and resistor are extremely critical for safety reasons. All components
on the line side MUST be rated by the appropriate standards
committees. Otherwise you are taking chances of burning your house
down.
I did work for a large (un-named) corporation where in the past
they didn't specify the correct components..... several families lost
their lives to the wrong choice of line capacitors.
Don't f... around with power line components.
As another weird thing, diodes can affect jitter on a CD player as
the RF currents (from the crystal timebase) have their paths out of
the amp turned on and off 60/120 times a second. Tha cure for that is
not diode type, but proper elimination of RF leakage from the
oscillators.

Re: 866 or Damper diodes - a definitive response

Paul wrote:
> On 27 Dec 2007 16:46:17 GMT, isw <isw@witzend,com > wrote:
> ....stuff deleted.....
>> I suppose it's possible that "soft" rectification makes a
>> less-than-competent grounding setup more-or-less adequte. High-current
>> pulses from "fast' rectifiers could cause voltage drops in unintended
>> places, producing hum or buzz.
>>
>
> If you are rectifying 50/60 Hz power, it is silly to specify soft
> or fast rectifiers. Due to the low dv/dt (change of voltage with
> respect to time), there is little charge "caught " in the rectifier
> junction that will cause current spikes on the reverse voltage. If you
> have a switching power supply on the other hand, with high rates of
> change of voltage, yes, you will have current spikes at the reverse
> voltage beginning, as the charge is pulled out of the junction.
> I tried this about 10 years ago, with spectrum analyzer, and a
> special circuit that is commercially used to test EMI that propagates
> from a power supply. It made diddly difference what type of diode was
> used. All the harmonic content coulkd be explained purely from the
> normal rectified sinusoid. That is for a sinewave supply.

Wonder how much current was involved in that supply?
Suspect not all that much...

There have been numerous papers on this subject that show the
very thing you claim to have not found, so perhaps that was the
differential??

> What DID make a difference was leakage inductances and stray
> interwinding capacitances. Those were controlled by RF bypass caps to
> the chassis, series bifilar inductors, and most importantly, a 0.1 mf
> capacitor in series with a 22 ohm resistor, and this combo placed
> directly across the power transformer primary. That circuit would
> lower the Q and dampen most of the resonances. The type of capacitor
> and resistor are extremely critical for safety reasons. All components
> on the line side MUST be rated by the appropriate standards
> committees. Otherwise you are taking chances of burning your house
> down.

When you say resonances here, what are your referring to on a supply
that you say is sinusoidal supplied *and* has a slow dv/dt??

And are the bifilar inductor/bypass cap combos on the primary or secondary
side?? (or both?)

> I did work for a large (un-named) corporation where in the past
> they didn't specify the correct components..... several families lost
> their lives to the wrong choice of line capacitors.

No fuses?
Hot chassis?
Fires?
No three wire safety power cords?
I'm confused!

> Don't f... around with power line components.
> As another weird thing, diodes can affect jitter on a CD player as
> the RF currents (from the crystal timebase) have their paths out of
> the amp turned on and off 60/120 times a second. Tha cure for that is
> not diode type, but proper elimination of RF leakage from the
> oscillators.

Ummm... unclear here too... "their paths out of the amp"?? What amp?
How is the output of the xtal oscillator modulated by the rectified line?
And how does the elimination of RF leakage from the oscillator(s) going
to effect what? Sorry, I'm interested in this, but confused by the
explanation.

- -bear

Re: 866 or Damper diodes - a definitive response

On 1 Jan 2008 16:16:27 GMT, bear <bearlabs@netzero,net > wrote:

>Paul wrote:
>> On 27 Dec 2007 16:46:17 GMT, isw <isw@witzend,com > wrote:
>> ....stuff deleted.....
>>> I suppose it's possible that "soft" rectification makes a
>>> less-than-competent grounding setup more-or-less adequte. High-current
>>> pulses from "fast' rectifiers could cause voltage drops in unintended
>>> places, producing hum or buzz.
>>>
>>
>> If you are rectifying 50/60 Hz power, it is silly to specify soft
>> or fast rectifiers. Due to the low dv/dt (change of voltage with
>> respect to time), there is little charge "caught " in the rectifier
>> junction that will cause current spikes on the reverse voltage. If you
>> have a switching power supply on the other hand, with high rates of
>> change of voltage, yes, you will have current spikes at the reverse
>> voltage beginning, as the charge is pulled out of the junction.
>> I tried this about 10 years ago, with spectrum analyzer, and a
>> special circuit that is commercially used to test EMI that propagates
>> from a power supply. It made diddly difference what type of diode was
>> used. All the harmonic content coulkd be explained purely from the
>> normal rectified sinusoid. That is for a sinewave supply.
>
>Wonder how much current was involved in that supply?
>Suspect not all that much...

About 1 ampere (primary side). I tried different currents, no real
difference to harmonic spectrum.

>
>There have been numerous papers on this subject that show the
>very thing you claim to have not found, so perhaps that was the
>differential??
>
Yup... I know the paper you're thinking of. That was one that
appeared in one of the audio rags, it turned out that that experiment
was only repeatable (showing high harmonic content) if the spectrum
analyzer was scanned too fast. I was able to duplicate (more or less)
their incorrect results. The paper was done by someone who didn't
understand the equipment. The paper was obviously not peer
reviewed.Many of the auiophile papers are of dubious quality, even a
few from some very respected scientists.

>> What DID make a difference was leakage inductances and stray
>> interwinding capacitances. Those were controlled by RF bypass caps to
>> the chassis, series bifilar inductors, and most importantly, a 0.1 mf
>> capacitor in series with a 22 ohm resistor, and this combo placed
>> directly across the power transformer primary. That circuit would
>> lower the Q and dampen most of the resonances. The type of capacitor
>> and resistor are extremely critical for safety reasons. All components
>> on the line side MUST be rated by the appropriate standards
>> committees. Otherwise you are taking chances of burning your house
>> down.
>
>When you say resonances here, what are your referring to on a supply
>that you say is sinusoidal supplied *and* has a slow dv/dt??
>

Any 50/60 Hz power supply has very slow dv/dt compared to a switching
power supply. There is no need for fast/soft recovery diodes for a
normal wall socket power., UNLESS it runs a switching power supply off
the rough DC, as most PC power supplies do.

>And are the bifilar inductor/bypass cap combos on the primary or secondary
>side?? (or both?)

For max. isolation, they were on the primary side. Filters must be
designed for max. attenuation - that means you need to consider the
various impedances to ground and to source(s). Filters designed for
typical line filters may need extra components at high frequencies,
since stray capacitance can give low impedances and reduced
attenuation at greater than 50 MHz. The resonances due to transformer
leakage inductances and interwinding capacitances were below several
MHz. For the purposes of determining the effect of different diodes I
ran the measurements with and without filters. There were some
harmonics that were higher due to resonances, but by optimizing the
filter (damped), they dropped to reasonable values.
filtering for analog circuits is quite different that that required
for digital or computer sources. Properly working audio analog
equipment doesn't usually generate very much junk that can be injected
into the line, unlike digital. The filtering is usually
unidirectional, that is to cut down on the stuff coming into the
device from the line. In the case of large common-mode signals on the
power lines , you might need extra components on the line to limit
noise currents. That can get tricky, since you can't raise the
impedance of the ground connection for safety reasons. Isolation
transformers can work, but they have interwinding capacitances that
end up giving you back the same problem.
Linn designed little switching supplies (their "Brilliant" supplies)
that replaced their toroidal power transformers, and got around this
problem by using transormers with low parasitics.
....ooops... I digress....
>
>> I did work for a large (un-named) corporation where in the past
>> they didn't specify the correct components..... several families lost
>> their lives to the wrong choice of line capacitors.
>
>No fuses?
>Hot chassis?
>Fires?
>No three wire safety power cords?
>I'm confused!
>
Correct components I am talking about are the bypass caps, damping
resistors, and of course all the other stuff you mention like fuses,
etc. The wrong choice of caps was responsible for the fires. They
promptly changed types, and installed components that would not
readily burn. When you choose capacitors or any other component that
is placed on the line side, it must meet proper safety specs, and I'm
not talking about silly audiophile rated components.

>> Don't f... around with power line components.
>> As another weird thing, diodes can affect jitter on a CD player as
>> the RF currents (from the crystal timebase) have their paths out of
>> the amp turned on and off 60/120 times a second. Tha cure for that is
>> not diode type, but proper elimination of RF leakage from the
>> oscillators.
>
>Ummm... unclear here too... "their paths out of the amp"?? What amp?
>How is the output of the xtal oscillator modulated by the rectified line?
>And how does the elimination of RF leakage from the oscillator(s) going
>to effect what? Sorry, I'm interested in this, but confused by the
>explanation.

I did a lot of messing around with jitter (CD player)and found that in
several instances that in the case of a poorly designed oscillator,
the RF leakage can cause jitter because of RF finding a path out from
the power supply. The component of audio that was caused by this
jitter was in the order of -90 to -95 db from full output. In this
case the diodes were the path out. By installing ferrite tubes on the
supply lines I was able to control it, but proper shielding and
termination of the clock lines was more effective.
-Paul