[Eiron]

Peter Wieck wrote:

Andrew Jute McCoy exuded more lies:


I didn't say it wasn't *also* a feedback circuit to some
fractional extent.


Either it is ZNF or it is not. Now that you have admitted that it is
not, then you have also admitted to all the rest of your lies... Not
the least of which is that you actually have an amp, rather than a
concatenation of loose parts and a bad schematic of a marginal plan of
a dysfunctional amp.

Show the amp. If Raymond can do it without fanfare, blather, pretense
or reams of crap, perhaps you can do it even _with_ all of that.

We have seen photos of what Andre claims to be the amp, though he lied
about the date. Obviously next time he will change the date on his camera
before taking any 'old' photographs.
The photograph was tastefully posed, hiding the bottom of the 300Bs
so we couldn't see if the valves were plugged in or just resting there;
the chassis is supposed to have different valve sockets depending on
what he is playing with at the time.

Just noticed another nice feature of the 'finalized' circuit:
fixed 2.4v bias, max input voltage 2vRMS!

--
Eiron.

[Peter Wieck]

Eiron wrote:

We have seen photos of what Andre claims to be the amp, though he lied
about the date. Obviously next time he will change the date on his camera
before taking any 'old' photographs.

A couple of things: The tubes in the submitted photo were Chinese
Silvertones. What was artfully posed behind were the _boxes_ that
purportedly containted the WE variety. Then, later, came a B&W photo of
a WE tube in a glass case with one of its vanity-published books behind
(also showing a 30 year-old photograph). But no actual connection
between any actual WE tube and any actual "real McCoy" amp. Even the
317s were concealed.

The amp in the picture was at best a breadboard attempt, not even the
beginnings of a finished product that one would leave exposed to polite
society.

So, we have a "finalized" amp that has had its "circuit updated". One
shudders to think how many more iterations this accretion of loose
parts will experience before either its creator shucks its mortal coil
or his landlord gets tired of changing fuses, or its tiny little
bed-sit burns down. But one thing is almost a certainty. It will never
be complete to any reliable operational degree.

Consider the elephant: Much trumpeting and activity, not much product
to show for a couple of years thereafter. In McCoy's case, the
tumpeting happens every day but lacks the activity and will produce no
product... Oh, right... other than the excrement it dumps here.

Peter Wieck
Wyncote, PA

[John Byrns]

In article ,
flipper wrote:

On Sun, 21 Jan 2007 01:32:17 GMT, Chris Hornbeck
wrote:

On Fri, 19 Jan 2007 01:10:38 -0600, flipper wrote:

I understand your explanation but I don't understand why that doesn't
introduce the same voltage differential (larger actually because of
the AC peaks) but varying between opposite extremes at 60Hz rather
than 'DC'.

I don't mean thermal emissions but the cathode voltage distribution vs
grid you were speaking of.

Excellent question, perfectly expressed.

With AC heating all physical points of the filament/cathode
have an instantaneous average of zero volts grid-cathode due
to the heating voltage. This occurs by definition from
the "hum nulling", and is true however the nulling is done.

To say it backwards, if the heating voltage could appear
as grid-cathode voltage it would appear as hum.

I understand your logic there but I'm not convinced it's air tight
because it seems to me there are two effects at play, one 'direct',
the hum emissions, and the other indirect, the voltage distribution
affecting the grid's amplification of an active signal.

It might be that the voltage distribution effect would cancel if the
grid-cathode signal transfer characteristic were perfectly linear but,
then, if it were perfectly linear then a DC differential would be of
no consequence as well.

To elaborate my thinking, take the hum with AC at it's 'peak', and no
signal. If it's nulled then you have the AC component on one end of
the filament equal in magnitude to the AC component on the other, but
in reverse phase so they cancel when summed at the plate. Simple
enough because the grid is static. But, I'm not convinced that the
effect on the signal also sums to 0 because amplification is not
linear and, at any instant in time other than zero crossing, the
signal places things at a different spot on the linearity curve than
where the hum nulling took place. And since it's in a different place
on the linearity curve the effect will not balance when summed at the
plate.

(Non linearity also suggests there's always residual hum because there
is always 'some' difference in amplification between the positive and
negative peaks. Strikes me that low voltage filaments would help to
minimize that.)

Seems to me that you have a perpetually rotating 120Hz (assuming it
matters not which 'side' of the filament is which way up or down)
distortion rather than a static one.

I can imagine that perhaps the human ear perceives it differently but
it's there.

Or did I make a mistake in there somewhere?

I didn't read every word, so I don't know if you made a mistake in there
or not, but it is important to note that even with no signal on the
grid, the hum balancing only nulls the 60 Hz component in the output.
The nonlinearity you describe means that harmonics of the supply
frequency will be generated and will not all be nulled, 120 Hz hum in
particular will appear at the plate, even with a perfectly filtered B
supply, as a result of using a 60 Hz supply to light the filament.


Regards,

John Byrns

--
Surf my web pages at, http://fmamradios.com/

[Eeyore]

John Byrns wrote:

flipper wrote:
Chris Hornbeck wrote:
flipper wrote:

I understand your explanation but I don't understand why that doesn't
introduce the same voltage differential (larger actually because of
the AC peaks) but varying between opposite extremes at 60Hz rather
than 'DC'.

I don't mean thermal emissions but the cathode voltage distribution vs
grid you were speaking of.

Excellent question, perfectly expressed.

With AC heating all physical points of the filament/cathode
have an instantaneous average of zero volts grid-cathode due
to the heating voltage. This occurs by definition from
the "hum nulling", and is true however the nulling is done.

To say it backwards, if the heating voltage could appear
as grid-cathode voltage it would appear as hum.

I understand your logic there but I'm not convinced it's air tight
because it seems to me there are two effects at play, one 'direct',
the hum emissions, and the other indirect, the voltage distribution
affecting the grid's amplification of an active signal.

It might be that the voltage distribution effect would cancel if the
grid-cathode signal transfer characteristic were perfectly linear but,
then, if it were perfectly linear then a DC differential would be of
no consequence as well.

To elaborate my thinking, take the hum with AC at it's 'peak', and no
signal. If it's nulled then you have the AC component on one end of
the filament equal in magnitude to the AC component on the other, but
in reverse phase so they cancel when summed at the plate. Simple
enough because the grid is static. But, I'm not convinced that the
effect on the signal also sums to 0 because amplification is not
linear and, at any instant in time other than zero crossing, the
signal places things at a different spot on the linearity curve than
where the hum nulling took place. And since it's in a different place
on the linearity curve the effect will not balance when summed at the
plate.

(Non linearity also suggests there's always residual hum because there
is always 'some' difference in amplification between the positive and
negative peaks. Strikes me that low voltage filaments would help to
minimize that.)

Seems to me that you have a perpetually rotating 120Hz (assuming it
matters not which 'side' of the filament is which way up or down)
distortion rather than a static one.

I can imagine that perhaps the human ear perceives it differently but
it's there.

Or did I make a mistake in there somewhere?

I didn't read every word, so I don't know if you made a mistake in there
or not, but it is important to note that even with no signal on the
grid, the hum balancing only nulls the 60 Hz component in the output.
The nonlinearity you describe means that harmonics of the supply
frequency will be generated and will not all be nulled,

100% correct.

Graham

[Chris Hornbeck]

On Sun, 21 Jan 2007 15:29:00 GMT, John Byrns
wrote:

In article ,
flipper wrote:

(Non linearity also suggests there's always residual hum because there
is always 'some' difference in amplification between the positive and
negative peaks. Strikes me that low voltage filaments would help to
minimize that.)

Seems to me that you have a perpetually rotating 120Hz (assuming it
matters not which 'side' of the filament is which way up or down)
distortion rather than a static one.

even with no signal on the
grid, the hum balancing only nulls the 60 Hz component in the output.
The nonlinearity you describe means that harmonics of the supply
frequency will be generated and will not all be nulled, 120 Hz hum in
particular will appear at the plate, even with a perfectly filtered B
supply, as a result of using a 60 Hz supply to light the filament.

Right, and the magnitude of the effect can be easily measured
as the amount of 120 Hz in the output. It's small but non-zero,
even in a very linear valve.

There was once interest here in using an ultrasonic AC heating
signal. And we still haven't discussed square wave heating voltages
yet.

Great thread, much thanks to all,

Chris Hornbeck
"History consists of truths which in the end turn into lies,
while myth consists of lies which finally turn into truths."
- Jean Cocteau

[Iain Churches]

"Arny Krueger" wrote in message
...
"flipper" wrote in message

A few exceptions don't prove a general rule.

Oh puleese, RDH, textbooks, and training materials aren't
"exceptions."

I see no specific reference to any proper textbook or training material.

RDH and Tremaine are legacy publications, and represent only a tiny
fraction of audio publications that are extant.

RHD has been a reference textbook since it's first edition.
Tremaine was a standard record industry and broadcast teaching
manual during the 1960's. It was written in the format of question and
answer specifically for that purpose. Howard Tremaine had a Ph.D


Furthermore, I can find no references to "Lewis Yorke" in any standard
list of publications. I found a post alleging "Lewis Yorke" wrote "High
Fidelity Valve Amplifiers", but guess what - I can find no references to
"High Fidelity Valve Amplifiers" in any standard list of publications.
Must be some kind of off-the-wall pamphlet.

Published 1964. Focal Press. 254 pages.
Some "off the wall pamphlet" :-)))
Standard UK educational text book.
Lewis Yorke was a London university professor.

Shall I quote you some more, Arny?
Please refer to the standard textbooks by
Patchett. AN. M.Sc. Ph.D.C.Eng. FIEE. FIERE.MIEE
Bailey AR. M-Sc (Eng) B.Sc (Eng) Ph.D
BELL EC. B.Sc. AM Brit IRE.

All the above wrote textbooks used in establishments
teaching audio engineering during the thermionic era.
All of these learned gentlemen go to great lengths to
differentiate clearly between heaters, and filaments.

So, one is inclined Arny, to take the word of such
academics rather than yours. I understand you
repair second-hand computers for a living.


Cordially
Iain

[Jon Yaeger]

in article , flipper at
wrote on 1/21/07 7:50 PM:

On Sun, 21 Jan 2007 18:54:11 GMT, Chris Hornbeck
wrote:

On Sun, 21 Jan 2007 15:29:00 GMT, John Byrns
wrote:

In article ,
flipper wrote:

(Non linearity also suggests there's always residual hum because there
is always 'some' difference in amplification between the positive and
negative peaks. Strikes me that low voltage filaments would help to
minimize that.)

Seems to me that you have a perpetually rotating 120Hz (assuming it
matters not which 'side' of the filament is which way up or down)
distortion rather than a static one.

even with no signal on the
grid, the hum balancing only nulls the 60 Hz component in the output.
The nonlinearity you describe means that harmonics of the supply
frequency will be generated and will not all be nulled, 120 Hz hum in
particular will appear at the plate, even with a perfectly filtered B
supply, as a result of using a 60 Hz supply to light the filament.

Right, and the magnitude of the effect can be easily measured
as the amount of 120 Hz in the output. It's small but non-zero,
even in a very linear valve.

Doesn't that mean, though, that the original theory of AC eliminting
'the voltage differential distortion effect that DC has' is not
correct? It's 'different', in that it's rotating, but not 'less'.


There was once interest here in using an ultrasonic AC heating
signal. And we still haven't discussed square wave heating voltages
yet.

This all kind of jives with RCA R-10 which, in describing direct and
indirect heating, list two advantages for indirect heating, that it
reduces AC hum and separates the cathode from the heater for, as they
put it, "freedom from electrical disturbances which might be
introduced through the filament supply lines."

They also mention, in the same breath, that they developed some tubes
with those features "particularly for automobile or dc line radio
sets."

My interpretation is the reasoning was to reduce heater noise from
things like the HV vibrator and generation/alternator injecting noise
on the DC and that makes me wonder if this is where the 'DC heating is
bad' idea came from.


Great thread, much thanks to all,

Chris Hornbeck
"History consists of truths which in the end turn into lies,
while myth consists of lies which finally turn into truths."
- Jean Cocteau


Chris,

I wonder if the DC heating caveat arises the observation that the voltage is
higher on one end of the filament than the distal end?

Jon

[Chris Hornbeck]

On Sun, 21 Jan 2007 18:50:56 -0600, flipper wrote:

Right, and the magnitude of the effect can be easily measured
as the amount of 120 Hz in the output. It's small but non-zero,
even in a very linear valve.

Doesn't that mean, though, that the original theory of AC eliminting
'the voltage differential distortion effect that DC has' is not
correct? It's 'different', in that it's rotating, but not 'less'.

I've always thought that it was both different and less, but I
now see that it was sloppy thinking. Your line of argument raises
two questions:

1) Does AC heating modulate transconductance, in some rough
proportion to observable twice-heating-frequency noise? An
analogy to DC heating's increase in the Child exponent
makes this a tough conclusion to avoid. Thoughts?

2) How could this be calculated? We can assume (I think!)
that the effect would be one of simply modulating transconductance
at the AC heating rate, and assuming a symmetrical physical
geometry. IOW, ignoring physical geometry and working from
curves.


I have equipment to measure down almost -100dB below signal,
and even a test-bed amplifier (that I can no longer pick up!)
but designing a suitable test format is really beyond me.

Both classic DC and AC heating effects are small but measurable,
but is the second-order effect of AC heating? Thoughts?

Much thanks, very exciting!, a new idea, very grateful,
but not yet dead,

Chris Hornbeck
"History consists of truths which in the end turn into lies,
while myth consists of lies which finally turn into truths."
- Jean Cocteau

[robert casey]

I wonder if the DC heating caveat arises the observation that the voltage is
higher on one end of the filament than the distal end?


Some filament tubes were designed with that in mind. Miniature tubes
used in battery portable radios (like 1U4) and some subminiature
filament tubes specify which end of the filament is to get the more
positive end of the filament supply. Which would mean that the control
grid is designed to compensate for the gradient of cathode voltage from
one end of the filament to the other.

[Eeyore]

robert casey wrote:

I wonder if the DC heating caveat arises the observation that the voltage is
higher on one end of the filament than the distal end?

Some filament tubes were designed with that in mind. Miniature tubes
used in battery portable radios (like 1U4) and some subminiature
filament tubes specify which end of the filament is to get the more
positive end of the filament supply. Which would mean that the control
grid is designed to compensate for the gradient of cathode voltage from
one end of the filament to the other.

Makes perfect sense.

Graham

[Chris Hornbeck]

On Sun, 21 Jan 2007 23:13:52 -0600, flipper wrote:

lotsa good stuff snipped. Gotta sleep; showdown at the
dayjob early AM corral; probably lots more time to
investigate when unemployed tomorrow

Here's another thought. Maybe picking a 'lousy' tube to test would be
better since the large non linearity should amplify the effect.

My test bed takes "50-watter" sizes, 211, 845, like that.
There were zero bias cutoff valves in that format, but I
don't have any.

I guess I'm not as interested in a measured outcome as in
a format for thinking about the issue. Lazy, too, but
that's not really the whole thing. Anyway, night,

Much thanks, as always,

Chris Hornbeck
"History consists of truths which in the end turn into lies,
while myth consists of lies which finally turn into truths."
- Jean Cocteau