[Iain Churches[_2_]]

One often sees references on this group to RDH, but very
few people seem to know about the excellent book entitled
Audio Cyclopedia, by Howard Tremaine, who was a designer
at McCurdy Industries, a Canadian company manufacturing
studio and broadcast equipment.

The book, in question and answer format, is comprised of 25 sections,
with more than 1700 pages. The second edition, fourth printing, which
is the edition most often seen, was published in 1975, is mainly tube/valve
orientated but does also have some SS circuits.

Chapters include:

Basic Principles of Sound
Acoustics, Studio techniques.
Microphones,
Attenuators,
Vacuum Tubes, Transistors and Diodes,
Audio Amplifiers,
Disc Recording,
Cutting Heads
Magnetic Recording
Loudspeakers.
Power Supplies,
Test Equipment.
Audio-frequency measurement
Installation
Charts and tables.

I am told by a former colleague from the UK who worked also at
Canadian Broadcasting that the book was one of their training manuals.

Highly recommended!

--
Iain

[Patrick Turner]

Iain Churches wrote:

One often sees references on this group to RDH, but very
few people seem to know about the excellent book entitled
Audio Cyclopedia, by Howard Tremaine, who was a designer
at McCurdy Industries, a Canadian company manufacturing
studio and broadcast equipment.

The book, in question and answer format, is comprised of 25 sections,
with more than 1700 pages. The second edition, fourth printing, which
is the edition most often seen, was published in 1975, is mainly tube/valve
orientated but does also have some SS circuits.

Chapters include:

Basic Principles of Sound
Acoustics, Studio techniques.
Microphones,
Attenuators,
Vacuum Tubes, Transistors and Diodes,
Audio Amplifiers,
Disc Recording,
Cutting Heads
Magnetic Recording
Loudspeakers.
Power Supplies,
Test Equipment.
Audio-frequency measurement
Installation
Charts and tables.

I am told by a former colleague from the UK who worked also at
Canadian Broadcasting that the book was one of their training manuals.

Highly recommended!

--
Iain

I found a copy in a local technical college library in 1995 and I paid
10c a page to copy
1,000 pages on the photocopier over two days.

The Audio Encyclopedia is a good book, to be sure, and slightly more
modern than RDH4,
but it all mainly obsolete now like most of what is in RDH4 because we
have
developed better ways of achieving good sound in loungerooms
compared to what was done in 1960.

But for anyone whose mind is stuck in the past, and who has unlimited
time
to waste during funded retirement for fixing old junk, the AE is a good
guide.

For myself, its a reference book, and I seldom need to read it now
because I just apply basic known principles to overcome problems and get
on with it.
I have very limited time, and have to earn my money, so I cannot waste
time on junk.

I condensed what I need know and what 95% of other ppl need to know at
my website
for good loungeroom sound.

But for those into cutting LP records or movie track sound et all, and I
don't know anyone at all interested,
and willing to spend the time alone making progress the AE is helpful,
but
one still has to learn what isn't in the books by doing an
apprenticeship.

There are dozens of intersting schematics of commercial circuit designs
in the AE
but I won't ever be using many...

Patrick Turner.

[BretLudwig]

The AC comes in three editions, of which the second is far and away the
most common. Unlike the RDH 4, whose price has come down a lot since it
was put up for free, legal download, it is still copyright in the US and
elsewhere so it can't be freely distributed. Furthermore, Tremaine's
family are apparently being prickish and refusing to allow reprint.
Apparently the publishers of "The New Audio Cyclopedia", wished to put a
.pdf of both earlier editions in as a cd-rom with the new one and the
family were impossible to deal with.

CUT COPYRIGHT TERMS TO REASONABLE LENGTH!!!!!!!!

The AC is not primarily about domestic hi-fi and is, like a lot of the old
Audel's and Sams books, written in Q and A format. Nonetheless it is quite
informative as long as one does not expect to be fed and burped and
changed by a book, let alone one dating from 1969.


It is important to remember before slagging old tube equipment that it
was not built to mil spec and was not intended to hold up for 50+ years.
They did not realistically expect anyone in 2008 to give a flying
Philadelphia cream cheese about amplifiers made in 1958. That's true of
American, British, or German equipment, or any other. Whether Quad Leak or
Radford, marantz, McIntosh or Fisher, or Klangfilm or Siemens, it has in
fact held up better than anyone thought it would.

I have several military radios, a Collins S-Line ham receiver and a lot
of old test equipment that was built better, but it cost the price of a
house.

--
Message posted using http://www.talkaboutaudio.com/group/rec.audio.tubes/
More information at http://www.talkaboutaudio.com/faq.html

[Engineer[_2_]]

On Jul 29, 12:05*am, "BretLudwig" wrote:

(snip)

*It is important to remember before slagging old tube equipment that it
was not built to mil spec and was not intended to hold up for 50+ years.
They did not realistically expect anyone in 2008 to give a flying
Philadelphia cream cheese about amplifiers made in 1958. That's true of
American, British, or German equipment, or any other. Whether Quad Leak or
Radford, marantz, McIntosh or Fisher, or Klangfilm or Siemens, it has in
fact held up better than anyone thought it would.


Agreed, but it can also be "remanufactured" to even better than
original specs (modern caps and resistors) as long as the iron is
good. As long as we can get the tubes it should live "forever"...
well, you know what I mean.
Cheers,
Roger

[Iain Churches[_2_]]

"Engineer" wrote in message
...
On Jul 29, 12:05 am, "BretLudwig" wrote:

(snip)

It is important to remember before slagging old tube equipment that it
was not built to mil spec and was not intended to hold up for 50+ years.
They did not realistically expect anyone in 2008 to give a flying
Philadelphia cream cheese about amplifiers made in 1958. That's true of
American, British, or German equipment, or any other. Whether Quad Leak
or
Radford, marantz, McIntosh or Fisher, or Klangfilm or Siemens, it has in
fact held up better than anyone thought it would.


Agreed, but it can also be "remanufactured" to even better than
original specs (modern caps and resistors) as long as the iron is
good.

Yes- But that's the heart of the problem. Try to find a coil winder
who can make you a transformer *precisely* to the Williamson or
Radford specification. I am, not sure about the situation in the US
but in the EU out of some thirty or so commercial transformer
manufacturers, there seem to be only two that can work to such a high
standard.

Iain

[Patrick Turner]

Engineer wrote:

On Jul 29, 12:05 am, "BretLudwig" wrote:

(snip)

It is important to remember before slagging old tube equipment that it
was not built to mil spec and was not intended to hold up for 50+ years.
They did not realistically expect anyone in 2008 to give a flying
Philadelphia cream cheese about amplifiers made in 1958. That's true of
American, British, or German equipment, or any other. Whether Quad Leak or
Radford, marantz, McIntosh or Fisher, or Klangfilm or Siemens, it has in
fact held up better than anyone thought it would.


Agreed, but it can also be "remanufactured" to even better than
original specs (modern caps and resistors) as long as the iron is
good. As long as we can get the tubes it should live "forever"...
well, you know what I mean.
Cheers,
Roger

But even if the iron is crook, it to can be replaced with better mades.

Its a case of replacing the handle and the blade of dad's old axe,
but the nice box for the axe still has a use.

Patrick Turner.

[Patrick Turner]

Iain Churches wrote:

"Engineer" wrote in message
...
On Jul 29, 12:05 am, "BretLudwig" wrote:

(snip)

It is important to remember before slagging old tube equipment that it
was not built to mil spec and was not intended to hold up for 50+ years.
They did not realistically expect anyone in 2008 to give a flying
Philadelphia cream cheese about amplifiers made in 1958. That's true of
American, British, or German equipment, or any other. Whether Quad Leak
or
Radford, marantz, McIntosh or Fisher, or Klangfilm or Siemens, it has in
fact held up better than anyone thought it would.


Agreed, but it can also be "remanufactured" to even better than
original specs (modern caps and resistors) as long as the iron is
good.

Yes- But that's the heart of the problem. Try to find a coil winder
who can make you a transformer *precisely* to the Williamson or
Radford specification.

I have demonstrated to myself that doing "*precisely*" what Raddy or
Willy done all those
years ago is a ******thon.

One does not need vertically divided bobbins for starters.

In most cases, one can improve the design of any given amp made over 20
years ago
by removing all the circuitry and parts and re-doing the whole damn amp
in the manner I have described in many ways at my website.

Most ancient designs of OPT were focused mainly on minimizing copper and
iron
because GOSS and copper had a real price many times what the real price
is now.

Skilled winding tradespeople were both cheap and plentiful in 1958,
so guess what, how did they manage some performance quality without much
iron or copper?

They made special laminations with a larger window area to centre core
area ratio
so more turns could be placed on.

The modern trend is for much lower winding resistances which will
withstand a saturated KT88 indefinately.

Old OPT just heat up and die with shorted turns.

Wiily's OPT had 4,400 P turns of terribly thin fragile wire.
Core was a 44mm stack of 32 tongue material, and window was 75mm x 25mm.

Nowdays one would use 2,200 turns with 60 stack of 51mm tongue to get
saturation
at the same F as in the original Willy OPT at the same voltage applied.
But because the P wire would be so much thicker, and the S wire size,
the OPT
would produce 64W easily at less than 5% winding losses, instead of
Willy's
16W and 10% winding losses.

My website gives an enormous amount of info on how to wind
excellent OPT that will perform better than the Willy models ever did
and which will
be more rugged.

I've never seen the actual winding plans for the Radford amps.
Old secret stuff. Bah, let them keep thie bloody secrets, I don't care,
what I will wind will be better anyway.


I am, not sure about the situation in the US
but in the EU out of some thirty or so commercial transformer
manufacturers, there seem to be only two that can work to such a high
standard.


Lundahl and Sowter?

I have just nearly completed reforming a VAC 7070 amp with a quad of
300B per channel.

The brand of OPT used is unknown, but they do seem to have a big
simularity
to something made by Hammond, which is the brand used by many
hobbyists in the US and elsewhere.

Hammond ain't the cream though.

It don't matter though, because the VAC OPT do have **enough** bandwidth
to allow
faultless and blameless wonderful sound and technical performance.

So we have a situation where the Hammond is the main choice because its
affordable and good enough.

Something wound like an old Williamson would be 3 times the price, and a
waste of money imho.

If I was a commercial winder, I'd charge twice what Hammond does,
because I know I'd offer a slightly better performance at HF due
to much better interleaving, and better ranges of impedance matches
*without* wasting turns on the secondary.
See my website for details.

Feel free to wind anything you fancy at my website such as OPT No1.

Clever dicks will do a lot better than I have.

Patrick Turner.







Iain

[BretLudwig]

Pat T

"I've never seen the actual winding plans for the Radford amps.
Old secret stuff. Bah, let them keep thie bloody secrets, I don't care,
what I will wind will be better anyway.


I am, not sure about the situation in the US
but in the EU out of some thirty or so commercial transformer
manufacturers, there seem to be only two that can work to such a high
standard.


Lundahl and Sowter?

I have just nearly completed reforming a VAC 7070 amp with a quad of
300B per channel.

The brand of OPT used is unknown, but they do seem to have a big
simularity
to something made by Hammond, which is the brand used by many
hobbyists in the US and elsewhere.

Hammond ain't the cream though.

It don't matter though, because the VAC OPT do have **enough** bandwidth
to allow
faultless and blameless wonderful sound and technical performance.

So we have a situation where the Hammond is the main choice because its
affordable and good enough.

Something wound like an old Williamson would be 3 times the price, and a
waste of money imho.

If I was a commercial winder, I'd charge twice what Hammond does,
because I know I'd offer a slightly better performance at HF due
to much better interleaving, and better ranges of impedance matches
*without* wasting turns on the secondary.
See my website for details.

Feel free to wind anything you fancy at my website such as OPT No1.

Clever dicks will do a lot better than I have.

Patrick Turner."

Visit any transformer plant winding stuff under 25 lbs. or so piece
weight, you'll see a largely dickless workforce. Transformer winding is
women's work, quite literally, on a commercial basis.

--
Message posted using http://www.talkaboutaudio.com/group/rec.audio.tubes/
More information at http://www.talkaboutaudio.com/faq.html

[Patrick Turner]

BretLudwig wrote:

Pat T

"I've never seen the actual winding plans for the Radford amps.
Old secret stuff. Bah, let them keep thie bloody secrets, I don't care,
what I will wind will be better anyway.

I am, not sure about the situation in the US
but in the EU out of some thirty or so commercial transformer
manufacturers, there seem to be only two that can work to such a high
standard.

Lundahl and Sowter?

I have just nearly completed reforming a VAC 7070 amp with a quad of
300B per channel.

The brand of OPT used is unknown, but they do seem to have a big
simularity
to something made by Hammond, which is the brand used by many
hobbyists in the US and elsewhere.

Hammond ain't the cream though.

It don't matter though, because the VAC OPT do have **enough** bandwidth
to allow
faultless and blameless wonderful sound and technical performance.

So we have a situation where the Hammond is the main choice because its
affordable and good enough.

Something wound like an old Williamson would be 3 times the price, and a
waste of money imho.

If I was a commercial winder, I'd charge twice what Hammond does,
because I know I'd offer a slightly better performance at HF due
to much better interleaving, and better ranges of impedance matches
*without* wasting turns on the secondary.
See my website for details.

Feel free to wind anything you fancy at my website such as OPT No1.

Clever dicks will do a lot better than I have.

Patrick Turner."

Visit any transformer plant winding stuff under 25 lbs. or so piece
weight, you'll see a largely dickless workforce. Transformer winding is
women's work, quite literally, on a commercial basis.


Aha, you know the secret behind secret men's business. Its women.

And indeed they wound many trannies, and made most of the tubes, and
other radio coils
and they did anything that suited their ability for quickly doing
dexterous work repeatedly.

Not many women worked in the design offices afaik.
So that's where the clever dicks were; beavering away to produce good
designs.

In the next office along the corridoor, some ******* called Bean Kownter
laboured away to
dumb down the design to make it use less women's work to wind it.

Its a **** of a world you know :-)

Patrick Turner.

--
Message posted using http://www.talkaboutaudio.com/group/rec.audio.tubes/
More information at http://www.talkaboutaudio.com/faq.html

[Iain Churches[_2_]]

"Patrick Turner" wrote in message
...


I've never seen the actual winding plans for the Radford amps.
Old secret stuff. Bah, let them keep thie bloody secrets, I don't care,
what I will wind will be better anyway.

All good coilwinders are very careful about giving away too
much information. I suppose that's understandable really. It's
their living and they have had to learn what they know the
hard way.

I once talked to Dr. Gavin Sowter about cloning a Radford
OPT. He said, "Hmm, quite a challenge. It is extremely
complex.

In the spec for the STA100 which Radford wrote
for the BBC, he stated, "the amplifier can run
indefinitely at full power into an open or short-circuit
load. I have never seen such a claim from any other
maker!


I am, not sure about the situation in the US
but in the EU out of some thirty or so commercial transformer
manufacturers, there seem to be only two that can work to such a high
standard.


Lundahl and Sowter?

Indeed.


I have just nearly completed reforming a VAC 7070 amp with a quad of
300B per channel.

The brand of OPT used is unknown, but they do seem to have a big
simularity
to something made by Hammond, which is the brand used by many
hobbyists in the US and elsewhere.

Hammond ain't the cream though.

They seem to have several different standards of transformers for differing
requirements. A chap I once spoke to at an audio fair, who made
very handsome custom-built Williamsons told me that he had never
been able to make one work properly with a Hammond. I have used
Hammond iron (mainly mains xformers), and found them to be OK,
but their regulation is not as good as the more expensive makers.

Feel free to wind anything you fancy at my website such as OPT No1.

An acquaintance of mine has been looking for a coilwinding machine for
several years. He come accross all kinds of junk, but never a hig-quality
machine. AVO in the UK used to make one, I think it was marketed
under the name McCade or something similar. It could wind four
bobbins simultaneously. I have never seen one, especially for sale.

I have a feeling that the Radford machines were built in-house. I
rememember on a visit to the Ashton Vale factory seeing three or
four machines.

Clever dicks will do a lot better than I have.

I doubt that many people have enen an inclination to try
Those of use who fettle tube amps have little enough spare time
as it is, without dabbling in the black arts as well:-)

Regards to all
Iain

[Patrick Turner]

Iain Churches wrote:

"Patrick Turner" wrote in message
...


I've never seen the actual winding plans for the Radford amps.
Old secret stuff. Bah, let them keep thie bloody secrets, I don't care,
what I will wind will be better anyway.

All good coilwinders are very careful about giving away too
much information. I suppose that's understandable really. It's
their living and they have had to learn what they know the
hard way.

I once talked to Dr. Gavin Sowter about cloning a Radford
OPT. He said, "Hmm, quite a challenge. It is extremely
complex.

Bull****.

I have heard this bull**** often.

Its all bull**** so the guy can avoid treading on toes, or taking any
time to talk to anyone.

Sowter won't talk to us right here, and what's he worried about?

Is he a snob?

Is he worried that 4 hobbyists will "steal" his ideas during the next 12
mths?

How the **** are those hobbyists ever going to wind something more
cheaply than
what happens in the Sowter factory?


Radford OPT are just OPT, right?

Follow the golden rules in RDH4, and you get an outcome
at least as good as Radford, or what anyone else can achieve, or better.

My 400W rated OPT in my 300 watt PP amps have 100 stack x 51 core
section, GOSS wasteless.
P = 1,050 turns in 5 sections, two layers each.
Sec has 6 six single layer sections each divided into 24 t + 48 t each
to give a wide range of lossless load matches.
So all up there are 17 windings in my "complex" tranny which isn't very
complex at all.

Because the insulation between P and S = 0.75mm, the Cshunt is low, and
because the interleaving pattern is
6S x 5P, I get 250W bandwidth from 19Hz at saturation to 270kHz, -3dB,
for what is a tranny for 1,200 ohms : 5.6 ohms
with 1,050 : 72 turn ratio.



In the spec for the STA100 which Radford wrote
for the BBC, he stated, "the amplifier can run
indefinitely at full power into an open or short-circuit
load. I have never seen such a claim from any other
maker!

I have never seen any tube amp able to sustain high levels into a short
circuit.

Of course, the above statement you quote is a lie, and grossly
misleading.

Full PO of 100 watts cannot ever be sustained into an open circuit, or a
short circuit!!!!!

If the load was say 1 ohm, you might get 50W at clipping, not 100W.

But all the amps I have seen which were meant to have 8 ohms connected
will die in the arse
pretty soon if 1 ohm or a short is connected and the input level is left
at the same level as
used to make 100W into 8 ohms.

If you work out the anode dissipation when the amp begins to clip into 1
ohm,
you'll be horrified.



I am, not sure about the situation in the US
but in the EU out of some thirty or so commercial transformer
manufacturers, there seem to be only two that can work to such a high
standard.


Lundahl and Sowter?

Indeed.


I have just nearly completed reforming a VAC 7070 amp with a quad of
300B per channel.

The brand of OPT used is unknown, but they do seem to have a big
simularity
to something made by Hammond, which is the brand used by many
hobbyists in the US and elsewhere.

Hammond ain't the cream though.

They seem to have several different standards of transformers for differing
requirements. A chap I once spoke to at an audio fair, who made
very handsome custom-built Williamsons told me that he had never
been able to make one work properly with a Hammond.

The guy probably had hardly any idea about critical damping.

Hardly anyone does.

I replaced the Chinese OPT in a Jolida 520 with Hammond 1650P and they
worked fine.

The original 520 wasn't very stable. Not much GNFB.

Many Chinese made amps cannot use a high amount of GNFB because their
OPT are so bleedin awful,
and because the makers have no idea about critical damping. Production
is controlled by dumb arse
entrepreneurs, and the marketting ensures hordes of equally dumb arse
customers roll in to spend up big.
Jolida cost over $3,500 in shops here and retail price is at least 20dB
above the cost of production at the sweatshops.

In the 520, I kept the original amount of NFB and stabilised it just
fine.
The Hammond were only marginally better, ie, hardly much better at all
compared to the
original Chinese OPT, one of which had shorted turns on one side of its
vertically divided bobbin.
I pulled the chinese OPT to peices to find the ****ed up turns and the
burnt wires.

What a ****ing mess i found!!! The ambition of joint venture
American-Chinese management was to make a nice OPT,
but in the making, the fuctard Chinese underpaid overworked cretins in
chinese slave labour sweatshops
failed miserably to maintain quality, and the layers soon became
jumbled, plain insulation tape was used for
P-S insulation, and frankly, the Chinese product was absolute crap.
No real quality control.

With the Hammond, sure, things get fiddly-diddly if you insist on 20dB
GNFB, like Willy did in 1947.
Like the Chinese Jolida OPT, Hammond doesn't use much interleaving in
that particular OPT.

Hammond have released a new range of OPT which entirely avoids having to
adjust
winding connections to get the 4, 8 or 16 ohm outlets. The new range
just has one S winding
with the full S = 16, 0.7 of the winding = 8 ohms, and the CT = 4 ohms,
without the ability
to parallel each 1/2 sec to get low losses and leakage with 4 ohms.

But even ARC use the one S winding suits all approach.

So they have worsened the quality!!!
Why?, because buyers want things simple, and cannot understand anything
even slightly complex any more.

But even with a tranny with 270kHz of BW, you still have troubles with
stability.

The troubles mean oscillation happens at a higher F than it does with an
OPT with a poorer BW.



I have used
Hammond iron (mainly mains xformers), and found them to be OK,
but their regulation is not as good as the more expensive makers.

You pay more for more interleaving less shunt C, less leakage, and
quality.


Feel free to wind anything you fancy at my website such as OPT No1.

An acquaintance of mine has been looking for a coilwinding machine for
several years. He come accross all kinds of junk, but never a hig-quality
machine. AVO in the UK used to make one, I think it was marketed
under the name McCade or something similar. It could wind four
bobbins simultaneously. I have never seen one, especially for sale.

I have a friend in Sydney who recently bought a winding lathe made in
Germany
50 years or more ago from an 80 yr old who finally retired.
He says he's able to put on 10,000 turns of fine wire in an ESL step up
tranny
in neat layers before morning tea time. And he never breaks a wire
and it traverses perfectly.

There are winding machines out there.

I couldn't find one either so I made my own.
Its slow, because there isn't any auto traversing mech, but I get a
perfect coil
with persistance.

There is more to making an OPT or PT or choke than just the winding.
There is the designing, the winding, the finding parts, wire, core,
insulations, then
assembly, varnishing, potting, testing and terminations.
There are risks and dangers of ****ing up your work at
each and every process.





I have a feeling that the Radford machines were built in-house. I
rememember on a visit to the Ashton Vale factory seeing three or
four machines.

Clever dicks will do a lot better than I have.

I doubt that many people have enen an inclination to try
Those of use who fettle tube amps have little enough spare time
as it is, without dabbling in the black arts as well:-)

OPTs are NOT a black art.

Its just plain simple engineering and trade work that has been mainly
forgotten in western nations as a direct result of using SS which can be
direct coupled to
any load.

The Chinese are now making nearly all the transformers used in western
countries
because the pay rate in China is $2 per day, but in London or Berlin or
new York its $100 per day.
And because workers in the west mainly don't actually work, but sit in
front of screens
and get fat arses.

Everyone who buys chinese goods supports the foul unjust social economic
status quo
where the workers of China are screwed into the dust.

The Chinese are going to try to get rich and in doing so will blacken
the planet with their
soot and the West doesn't want to pay them a fair wage to fund anti CO2
emission measures.

Its a completely fuct up world.

It should be just as economically viable and evironmentally friendly to
set up a factory to anything
in any country, because each worker of the world should get the same pay
for the same work.

Communism couldn't provide that.

Capitalism can't either.

I just wind all my own OPT for my own projects, **** the rest of the
world; it can do whatever, I don't give a ****.

I also give all my secrets away freely. But they are NOT really MY
secrets, but just common good
knowledge widely known by competent tranny winders of the past, present
and future.

Patrick Turner.





Regards to all
Iain

[Iain Churches[_2_]]

"Patrick Turner" wrote in message
...


Iain Churches wrote:

"Patrick Turner" wrote in message
...


I've never seen the actual winding plans for the Radford amps.
Old secret stuff. Bah, let them keep thie bloody secrets, I don't care,
what I will wind will be better anyway.

All good coilwinders are very careful about giving away too
much information. I suppose that's understandable really. It's
their living and they have had to learn what they know the
hard way.

I once talked to Dr. Gavin Sowter about cloning a Radford
OPT. He said, "Hmm, quite a challenge. It is extremely
complex.

Bull****.
I have heard this bull**** often.

I have dealt with Sowter since the mid 60s, and know many
of the people there personally. BS is not something in which
they indulge.

But, I do remember John Widgery telling me that the Radford
STA100 primary had twenty separate sections in series and
parallel, some with reversed polarity.´That sounds pretty
complex to me!

Its all bull**** so the guy can avoid treading on toes, or taking any
time to talk to anyone.

Sowter won't talk to us right here, and what's he worried about?

Is he a snob?

Dr Gavin Sowter ?? A gentleman, but certainly not a snob.
One of the nicest men you could have wished to have met.
He is now in the great transformer winding shop in the sky.

In the spec for the STA100 which Radford wrote
for the BBC, he stated, "the amplifier can run
indefinitely at full power into an open or short-circuit
load. I have never seen such a claim from any other
maker!

I have never seen any tube amp able to sustain high levels into a short
circuit.

Of course, the above statement you quote is a lie, and grossly
misleading.

You can be sure the people at the Beeb amd many others, tried it.
Arthur Radford would have been foolish indeed to make such a
statement if it could be proved to be true. I will try to find the
amp spec and copy it for you.

Full PO of 100 watts cannot ever be sustained into an open circuit, or a
short circuit!!!!!

I have such an amplifier, but its value it such that I would be unwilling
to abuse it in any way.

Personally it makes little difference to me either way. I certainly
would not get hot under the collar about it:-)

Lundahl and Sowter?

Indeed.


I have just nearly completed reforming a VAC 7070 amp with a quad of
300B per channel.

A chap I once spoke to at an audio fair, who made
very handsome custom-built Williamsons told me that he had never
been able to make one work properly with a Hammond.

The guy probably had hardly any idea about critical damping.

Hardly anyone does.

IIRC he mentioned that excess leakage inductance was the problem
with Hammond iron at that time. The situation may or may not have
improved. I cannot say.

Regards to all
Iain

[Ian Thompson-Bell]

IOt appears the original Audio Cyclopedia is highly thought of, so I
have asked my ocal library to see if they can find me a copy.

There appears to be a modern replacement called The New Audio
Cyclopedia. Is it any good?

Cheers

Ian

[Iain Churches[_2_]]

"Ian Thompson-Bell" wrote in message
...


It appears the original Audio Cyclopedia is highly thought of, so I have
asked my ocal library to see if they can find me a copy.

The second edition was ubiquitous, and used by studios and broadcasting
authorities as a teaching manual. Your library should be able to find you
a copy. I have seen it also in the UK for sale second hand for about
UKP 35

I actually have two copies of the second edition. One of these,
almost mint, I keep by my easy chair in the music room is for
relaxed reading. The other copy is next to my bench. It got left
out in the snow last winter. Don't asky why and how, it's a
long story!!

There appears to be a modern replacement called The New Audio Cyclopedia.
Is it any good?

Is this written by Tremaine? Or is it someone trying to cash in
on his reputation ?

The Morgan Jones books are excellent also. I find my second
edition of "Valve Amplifiers" most useful. He also has a book
on the practical aspects of building, called "Building Valve Amplifiers"
One of the chapters is entitled "Metalwork for Poets" ! Morgan Jones
has a first class BBC pedigree, and a very enjoyable writing style.

Iain

[Patrick Turner]

Iain Churches wrote:

"Patrick Turner" wrote in message
...


Iain Churches wrote:

"Patrick Turner" wrote in message
...


I've never seen the actual winding plans for the Radford amps.
Old secret stuff. Bah, let them keep thie bloody secrets, I don't care,
what I will wind will be better anyway.

All good coilwinders are very careful about giving away too
much information. I suppose that's understandable really. It's
their living and they have had to learn what they know the
hard way.

I once talked to Dr. Gavin Sowter about cloning a Radford
OPT. He said, "Hmm, quite a challenge. It is extremely
complex.

Bull****.
I have heard this bull**** often.

I have dealt with Sowter since the mid 60s, and know many
of the people there personally. BS is not something in which
they indulge.

But, I do remember John Widgery telling me that the Radford
STA100 primary had twenty separate sections in series and
parallel, some with reversed polarity.´That sounds pretty
complex to me!

So bloody what?

20 P sections is NOT especially complex for any OPT.

Usually, some winders say all sorts of crap to ppl to make sure they
think only the
High Priests of Tube Audio are entitled to come up with valid OPT
designs and
be able to wind them properly.

What is usually meant by a "section" isn't something you are familiar
with,
and Widgery did a grand job in pulling wool right over your eyes.

A section of a Primary winding is one that is defined as being all of a
group of turns which are located between
Secondary sections on each side.

So when some buffon say, "Aw gees mate, there are 20 sections that there
primary.."
it implies there would be 21 sections of secondary in a bobbin winding
section with consecutive sections arranged as
SPSPSPSPSPSPSPSPSPSPSPSPSPSPSPSPSPSPSPSPSPS.

This would give the OPT disgustingly high C shunt, and be very poor
practice.

21 P sections when mentioned without definitions that you should have
insisted upon like I would,
could mean that there were 21 primary layers of wire which happens to be
a very common number of P layers.

These layers, you may have discovered had you got the whole truth and
nothing but the truth from Widgery,
might have been divided into say 5 sections if you were lucky, say 4L +
4L + 5L + 4L + 4L,
with the CT for the B+ 1/2 way along the 5L centre section of P.
Its highly likely there are 4 single layer sections of secondary,
perhaps divided into sub sections,
but all very ordinary good practice that i am 100% sure Radford wasn't
going to try to exceed lest he
saddle himself with terrible quality control problems.
But in fact with decent winding machinery and in a mass production
regime, winding complexity
isn't all that difficult if you have the right tradeswomen who've been
well trained after years of
winding all kinds of stuff.

Perhaps the Primary sections of 4 layers each were subdivided into
layers connected "strangly" with regard to winding direction
and sequencing to reduce shunt capacitance.

But Widgery did nothing except deliberatly conceal the truth if indeed
he ever really knew it, and
he bull****ted to boost his ego and aura with the audience that
comprised yourself.



Its all bull**** so the guy can avoid treading on toes, or taking any
time to talk to anyone.

Sowter won't talk to us right here, and what's he worried about?

Is he a snob?

Dr Gavin Sowter ?? A gentleman, but certainly not a snob.
One of the nicest men you could have wished to have met.
He is now in the great transformer winding shop in the sky.

I can't disagree. I never met him, and being dead is a fair excuse for
not being present here.

But we never hear from anyone who IS at Sowter now.

I don't expect them to wade in here because they'd have zero to gain
from the experience.



In the spec for the STA100 which Radford wrote
for the BBC, he stated, "the amplifier can run
indefinitely at full power into an open or short-circuit
load. I have never seen such a claim from any other
maker!

I have never seen any tube amp able to sustain high levels into a short
circuit.

Of course, the above statement you quote is a lie, and grossly
misleading.

You can be sure the people at the Beeb amd many others, tried it.
Arthur Radford would have been foolish indeed to make such a
statement if it could be proved to be true. I will try to find the
amp spec and copy it for you.

Full PO of 100 watts cannot ever be sustained into an open circuit, or a
short circuit!!!!!

I have such an amplifier, but its value it such that I would be unwilling
to abuse it in any way.

Personally it makes little difference to me either way. I certainly
would not get hot under the collar about it:-)


Maybe you see my point.

I am NOT so gullible to believe idiotic claims and half baked
explanations.

If you have an input sine wave equal to that which causes clipping into
8 ohms,
and the outlet has been designed for 8 ohms, then the amp shouldn't
overheat,
ie, Pda should not quite get up to the rating for the tubes, ie, 42
watts for each KT88.

But lemme tellya, if ya run any tube amp with that input voltage, and a
shorted output,
then expect red hot anodes and maybe a ****ed OPT within 15 minutes if
the mains fuses don't blow.

Do the calculations for the Pda per KT88, and you'll find the average
tube current x average
tube voltage is way over 42 watts.

Radford amps like all that old stuff do not have active protection, and
rely on fuses.
Sometimes the fuses work when there is gross overload with a sine wave
because the
Idc needing to be supplied to the OPT CT goes way high in a fault
condition.
But it is very easy to blow up many amps with a shorted output with
music, and not have any fuses blow.
Twice last year I repaired a Quad-II amp driving an ESL57 with an
intermittently shorting midrange panel.
Boy whatta mess!

Radford amps would **** up under the same situation.



Lundahl and Sowter?

Indeed.


I have just nearly completed reforming a VAC 7070 amp with a quad of
300B per channel.

A chap I once spoke to at an audio fair, who made
very handsome custom-built Williamsons told me that he had never
been able to make one work properly with a Hammond.

The guy probably had hardly any idea about critical damping.

Hardly anyone does.

IIRC he mentioned that excess leakage inductance was the problem
with Hammond iron at that time. The situation may or may not have
improved. I cannot say.

The LL with a 1650P is a bit too high. I know, I used a pair in a Jolida
502 3 mths ago
to replace the horrible Chinese OPTs, one of which had got shorted turns
after 8 years.

The Chinese didn't get the critical damping right, but I did.

Leak made OPT with 50mH of LL and boy they were crap, but you can get
around these defects
if you know how, see my pages on it.

I routinely wind OPT with less than 5mH for the same applications.
Instead of Leaks woeful low amount of interleaving, I use at least a 5S
x 4P section arrangement.
So before you praise up what Radford done, define exactly what he really
did do.

If you asked Leak about his trannies, you'd get all this hogwash about
how ****in marvellous they were.
But really, Leak OPT were ****in horrid, and worse than the Chinese
crap.

Sure these things are touted to be "complex" and hard to wind.

BS, most OPT are not complex at all, but the winders always crap on
about their difficulties,
to make themselves look like heroes who should be knighted by QE2.

Some need sending to Siberia to teach them about the truth ahd hard
work.

They talk utter BS and never reveal the REAL details of what they do.

The original Williamson design is one that is defined OPENLY and FREELY
in TRANSPARENT
detail in RDH4 had two side by side indentical bobbins with a lot of
complexity
if that is what it was. Take a look and the simple 1 paragraph
description of it.
Get anything like that paragraph from Widgery?

I rest my case.

Patrick Turner.





Regards to all
Iain

[John Byrns]

In article ,
"Iain Churches" wrote:

The Morgan Jones books are excellent also. I find my second
edition of "Valve Amplifiers" most useful. He also has a book
on the practical aspects of building, called "Building Valve Amplifiers"
One of the chapters is entitled "Metalwork for Poets" ! Morgan Jones
has a first class BBC pedigree, and a very enjoyable writing style.

I find it hard to think positively of Morgan Jones ever since his goof
with the concertina cathode build out resistor in a magazine article and
the first edition of his book. IIRC he flip flopped in the second
edition but he should have known better in the first place. It just
shows that he didn't bother doing the math, which you should do when
making a recommendation that flies in the face of years conventional
practice.


Regards,

John Byrns

--
Regards,

John Byrns

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

[Iain Churches[_2_]]

"flipper" wrote in message
...
On Mon, 04 Aug 2008 10:18:06 -0500, John Byrns
wrote:

In article ,
"Iain Churches" wrote:

The Morgan Jones books are excellent also. I find my second
edition of "Valve Amplifiers" most useful. He also has a book
on the practical aspects of building, called "Building Valve Amplifiers"
One of the chapters is entitled "Metalwork for Poets" ! Morgan Jones
has a first class BBC pedigree, and a very enjoyable writing style.

I find it hard to think positively of Morgan Jones ever since his goof
with the concertina cathode build out resistor in a magazine article and
the first edition of his book. IIRC he flip flopped in the second
edition but he should have known better in the first place. It just
shows that he didn't bother doing the math, which you should do when
making a recommendation that flies in the face of years conventional
practice.

Could you elaborate on what the goof was?

In the preface to the second edition, Morgan Jones makes a reference
to a number of howlers "for which the author can only humbly aplogize"

Its a 500 page book packed with information, tips, schematics
and formulae. That there were a few errors, now corrected,
is not a surprise.

Most still regard his books as some of the most important in
sustaining the healthy interest in thermionic audio.

Regards to all
Iain

[Ian Thompson-Bell]

Iain Churches wrote:
"Ian Thompson-Bell" wrote in message
...


It appears the original Audio Cyclopedia is highly thought of, so I have
asked my ocal library to see if they can find me a copy.

The second edition was ubiquitous, and used by studios and broadcasting
authorities as a teaching manual. Your library should be able to find you
a copy. I have seen it also in the UK for sale second hand for about
UKP 35


I checked it out at Amazon and the cheapest I could find was 185USD.
Where in the UK have you seen it for 35 quid?

I actually have two copies of the second edition. One of these,
almost mint, I keep by my easy chair in the music room is for
relaxed reading. The other copy is next to my bench. It got left
out in the snow last winter. Don't asky why and how, it's a
long story!!

There appears to be a modern replacement called The New Audio Cyclopedia.
Is it any good?

Is this written by Tremaine? Or is it someone trying to cash in
on his reputation ?


Apparently it includes some of Tremaines work, but is extended to
include more recent stuf e.g.digital. I believe Tremaine is NOT credited
though.

The Morgan Jones books are excellent also. I find my second
edition of "Valve Amplifiers" most useful.

Yes, I have that too. Some useful stuff in there not found elsewhere.

He also has a book
on the practical aspects of building, called "Building Valve Amplifiers"
One of the chapters is entitled "Metalwork for Poets" ! Morgan Jones
has a first class BBC pedigree, and a very enjoyable writing style.

Iain


Cheers

Ian

[Iain Churches[_2_]]

"Ian Thompson-Bell" wrote in message
...
Iain Churches wrote:
"Ian Thompson-Bell" wrote in message
...


It appears the original Audio Cyclopedia is highly thought of, so I have
asked my ocal library to see if they can find me a copy.

The second edition was ubiquitous, and used by studios and broadcasting
authorities as a teaching manual. Your library should be able to find
you
a copy. I have seen it also in the UK for sale second hand for about
UKP 35


I checked it out at Amazon and the cheapest I could find was 185USD. Where
in the UK have you seen it for 35 quid?

At a bookshop in Twickenham, just a few months ago.
The mint copy I have is marked £25-95 on the inside cover.
I paid a little more than that for it.


There appears to be a modern replacement called The New Audio
Cyclopedia. Is it any good?

Is this written by Tremaine? Or is it someone trying to cash in
on his reputation ?


Apparently it includes some of Tremaines work, but is extended to include
more recent stuf e.g.digital. I believe Tremaine is NOT credited though.

Hmm. Curious!

The Morgan Jones books are excellent also. I find my second
edition of "Valve Amplifiers" most useful.

Yes, I have that too. Some useful stuff in there not found elsewhere.


By the way Ian, did you manage to get the Feedback Instruments
phase shift meter you were thinking to bid for on e-Bay?

Regards
Iain

[Ian Thompson-Bell]

Iain Churches wrote:
"Ian Thompson-Bell" wrote in message
...
Iain Churches wrote:
"Ian Thompson-Bell" wrote in message
...


It appears the original Audio Cyclopedia is highly thought of, so I have
asked my ocal library to see if they can find me a copy.
The second edition was ubiquitous, and used by studios and broadcasting
authorities as a teaching manual. Your library should be able to find
you
a copy. I have seen it also in the UK for sale second hand for about
UKP 35

I checked it out at Amazon and the cheapest I could find was 185USD. Where
in the UK have you seen it for 35 quid?

At a bookshop in Twickenham, just a few months ago.
The mint copy I have is marked £25-95 on the inside cover.
I paid a little more than that for it.


Up here in the wilds of Norfolk the only 2nd hand bookshops we have
stock arty farty crap - very little technical stuff and virtually nil
electronics. I visit them regularly but balways come away disappointed.

There appears to be a modern replacement called The New Audio
Cyclopedia. Is it any good?
Is this written by Tremaine? Or is it someone trying to cash in
on his reputation ?

Apparently it includes some of Tremaines work, but is extended to include
more recent stuf e.g.digital. I believe Tremaine is NOT credited though.

Hmm. Curious!

The Morgan Jones books are excellent also. I find my second
edition of "Valve Amplifiers" most useful.
Yes, I have that too. Some useful stuff in there not found elsewhere.


By the way Ian, did you manage to get the Feedback Instruments
phase shift meter you were thinking to bid for on e-Bay?

Unfortunately not - had to go out and missed the end of the auction -
wish I had just paid the buy it now price - ah well no matter you can
use the oscillator on its own to measure phase.

Cheers

ian

[Iain Churches[_2_]]

"flipper" wrote in message
...
On Mon, 4 Aug 2008 22:18:00 +0300, "Iain Churches"
wrote:



"flipper" wrote in message
. ..
On Mon, 04 Aug 2008 10:18:06 -0500, John Byrns
wrote:

In article ,
"Iain Churches" wrote:

The Morgan Jones books are excellent also. I find my second
edition of "Valve Amplifiers" most useful. He also has a book
on the practical aspects of building, called "Building Valve
Amplifiers"
One of the chapters is entitled "Metalwork for Poets" ! Morgan Jones
has a first class BBC pedigree, and a very enjoyable writing style.

I find it hard to think positively of Morgan Jones ever since his goof
with the concertina cathode build out resistor in a magazine article and
the first edition of his book. IIRC he flip flopped in the second
edition but he should have known better in the first place. It just
shows that he didn't bother doing the math, which you should do when
making a recommendation that flies in the face of years conventional
practice.

Could you elaborate on what the goof was?

In the preface to the second edition, Morgan Jones makes a reference
to a number of howlers "for which the author can only humbly aplogize"

Its a 500 page book packed with information, tips, schematics
and formulae. That there were a few errors, now corrected,
is not a surprise.

Most still regard his books as some of the most important in
sustaining the healthy interest in thermionic audio.

I'm not concerned with throwing stones.


No of course not.

I just wanted to be awares in
case I run across a site repeating or referencing it.

This book is now into its third edition second or third
printing. The error to which John refers was in the
1st edition, so unless you buy a second hand copy
of the book you will not come across them.

Regards to all
Iain

[Patrick Turner]

John Byrns wrote:

In article ,
"Iain Churches" wrote:

The Morgan Jones books are excellent also. I find my second
edition of "Valve Amplifiers" most useful. He also has a book
on the practical aspects of building, called "Building Valve Amplifiers"
One of the chapters is entitled "Metalwork for Poets" ! Morgan Jones
has a first class BBC pedigree, and a very enjoyable writing style.

I find it hard to think positively of Morgan Jones ever since his goof
with the concertina cathode build out resistor in a magazine article and
the first edition of his book. IIRC he flip flopped in the second
edition but he should have known better in the first place. It just
shows that he didn't bother doing the math, which you should do when
making a recommendation that flies in the face of years conventional
practice.

Regards,

John Byrns

I have a copy of Morgan Jones 'Valve Amplifiers', second edition,
and its riddled with errors and sloppy maths without derivations or
examples or any depth.
Eg, see page 107, with a schematic of a white cathode follower.
Why is the anode R of the top triode = 62k????

It looks like a nice oscillator though

Away from the mistakes, its got some good stuff.

Patrick Turner.

--
Regards,

John Byrns

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

[Iain Churches[_2_]]

"Patrick Turner" wrote in message
...

I have a copy of Morgan Jones 'Valve Amplifiers', second edition,
and its riddled with errors and sloppy maths without derivations or
examples or any depth.


Eg, see page 107, with a schematic of a white cathode follower.
Why is the anode R of the top triode = 62k????

Perhaps this is a misprint? Has it been corected in the 3rd edition
I wonder?

Have you collected together the errors and sent a list of them to
the publishers, for acknowledgement and correction in a future
edition? This would be of greater use to the tube-audio fraternity
that simply denouncing the book here on this group.

Iain

[Patrick Turner]

Iain Churches wrote:

"Patrick Turner" wrote in message
...

I have a copy of Morgan Jones 'Valve Amplifiers', second edition,
and its riddled with errors and sloppy maths without derivations or
examples or any depth.

Eg, see page 107, with a schematic of a white cathode follower.
Why is the anode R of the top triode = 62k????

Perhaps this is a misprint? Has it been corected in the 3rd edition
I wonder?

Have you collected together the errors and sent a list of them to
the publishers, for acknowledgement and correction in a future
edition? This would be of greater use to the tube-audio fraternity
that simply denouncing the book here on this group.

Iain

Sigh, If I was retired, and had all day to do nothing but be "creative"
and not worry about earning money, and not having to interrupt my work
with digging up
drain pipes and repairing /cleaning clogged pipes, and replacing kitchen
plumbing now
46 years old, and doing all this stuff a man has to damn well do besides
be a good tubologist,
then I might get time to properly write a long critique of Mr Jones's
book
which I feel wasn't proofed well enough by himself or anyone else and
then nobody would worry
I was just being grumpy on a news group but then I reckon I have gone
past the age of 40 and have earned the do-as-you-bloody-well-like
license, and now been awarded and official permit from the GOT to be
grumpy, because I cannot do as I like
because the older I get the better I woz.

:-)


Errors bedevil authors severly if they don't re read all their work 10
times slowly, and ask
what a dumb clod would make of what was being written.

I recall writing the last edition of my website over a period of 4
months full time,
and boy, did I make some errors!
Some got past my editediteditediting, and some good ppl
sent me their concerns. There were mainly trivial R or C values, or a
wrong line of calculations.
I fixed the errors as soon as i was told.
I like giving fully worked examples, like RDH4.

People like me who don't believe anything they are told, or anything
they read unless proof is offered,
know how to read books like Jones's.

We have a little chuckle now and then, and move right along, realizing
that nobody, not even ourselves, are perfect.

Patrick Turner.

[John Byrns]

In article ,
flipper wrote:

On Mon, 04 Aug 2008 10:18:06 -0500, John Byrns
wrote:

In article ,
"Iain Churches" wrote:

The Morgan Jones books are excellent also. I find my second
edition of "Valve Amplifiers" most useful. He also has a book
on the practical aspects of building, called "Building Valve Amplifiers"
One of the chapters is entitled "Metalwork for Poets" ! Morgan Jones
has a first class BBC pedigree, and a very enjoyable writing style.

I find it hard to think positively of Morgan Jones ever since his goof
with the concertina cathode build out resistor in a magazine article and
the first edition of his book. IIRC he flip flopped in the second
edition but he should have known better in the first place. It just
shows that he didn't bother doing the math, which you should do when
making a recommendation that flies in the face of years conventional
practice.

Could you elaborate on what the goof was?

The goof was not a typo or an error in a formula, his idea of placing a
build out resistor in the cathode of a concertina phase inverter to
equalize the source impedances of the plate and cathode circuits was
simply a goofy idea. It was a "bright" idea intended to fix an imagined
problem that didn't actually exist, that instead created a real problem.

I saw the magazine article and the first edition of his book where he
presented this goofy idea, I have never seen the second edition of his
book to see how he extricated himself from the predicament he created
for himself, although I have been told by people that have seen the
second edition that he did somehow extricate himself.

--
Regards,

John Byrns

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

[Iain Churches[_2_]]

"Patrick Turner" wrote in message
...


Iain Churches wrote:

"Patrick Turner" wrote in message
...

I have a copy of Morgan Jones 'Valve Amplifiers', second edition,
and its riddled with errors and sloppy maths without derivations or
examples or any depth.

Eg, see page 107, with a schematic of a white cathode follower.
Why is the anode R of the top triode = 62k????

Perhaps this is a misprint? Has it been corected in the 3rd edition
I wonder?

Have you collected together the errors and sent a list of them to
the publishers, for acknowledgement and correction in a future
edition? This would be of greater use to the tube-audio fraternity
that simply denouncing the book here on this group.


Sigh, If I was retired, and had all day to do nothing but be "creative"
and not worry about earning money, and not having to interrupt my work
with digging up
drain pipes and repairing /cleaning clogged pipes, and replacing kitchen
plumbing now
46 years old, and doing all this stuff a man has to damn well do besides
be a good tubologist,
then I might get time to properly write a long critique of Mr Jones's
book
which I feel wasn't proofed well enough by himself or anyone else and
then nobody would worry
I was just being grumpy on a news group but then I reckon I have gone
past the age of 40 and have earned the do-as-you-bloody-well-like
license, and now been awarded and official permit from the GOT to be
grumpy, because I cannot do as I like
because the older I get the better I woz.

:-)

Understood: And taken in the spirit in which I am sure it
was intended:-)


We have a little chuckle now and then, and move right along, realizing
that nobody, not even ourselves, are perfect.

Erm. Don't you mean "realizing that nobody, not even
ourselves, IS perfect" ?

Cheers:-)

Iain

[Iain Churches[_2_]]

"John Byrns" wrote in message
...
In article ,
flipper wrote:

I find it hard to think positively of Morgan Jones ever since his goof
with the concertina cathode build out resistor in a magazine article and
the first edition of his book. IIRC he flip flopped in the second
edition but he should have known better in the first place. It just
shows that he didn't bother doing the math, which you should do when
making a recommendation that flies in the face of years conventional
practice.

Could you elaborate on what the goof was?

The goof was not a typo or an error in a formula, his idea of placing a
build out resistor in the cathode of a concertina phase inverter to
equalize the source impedances of the plate and cathode circuits was
simply a goofy idea. It was a "bright" idea intended to fix an imagined
problem that didn't actually exist, that instead created a real problem.

I saw the magazine article and the first edition of his book where he
presented this goofy idea, I have never seen the second edition of his
book to see how he extricated himself from the predicament he created
for himself, although I have been told by people that have seen the
second edition that he did somehow extricate himself.


John, can you get access to the 2nd edition of Valve Amplifiers?
It would be of interest to know if this refers to page 279 of the
paragraph which starts "Output resistance with both terminals
equally loaded" ?

Is not the large difference in output impedance
between anode and cathode an important
shortcoming of the concertina?

Iain

[Patrick Turner]

Iain Churches wrote:

"Patrick Turner" wrote in message
...


Iain Churches wrote:

"Patrick Turner" wrote in message
...

I have a copy of Morgan Jones 'Valve Amplifiers', second edition,
and its riddled with errors and sloppy maths without derivations or
examples or any depth.

Eg, see page 107, with a schematic of a white cathode follower.
Why is the anode R of the top triode = 62k????

Perhaps this is a misprint? Has it been corected in the 3rd edition
I wonder?

Have you collected together the errors and sent a list of them to
the publishers, for acknowledgement and correction in a future
edition? This would be of greater use to the tube-audio fraternity
that simply denouncing the book here on this group.

Sigh, If I was retired, and had all day to do nothing but be "creative"
and not worry about earning money, and not having to interrupt my work
with digging up
drain pipes and repairing /cleaning clogged pipes, and replacing kitchen
plumbing now
46 years old, and doing all this stuff a man has to damn well do besides
be a good tubologist,
then I might get time to properly write a long critique of Mr Jones's
book
which I feel wasn't proofed well enough by himself or anyone else and
then nobody would worry
I was just being grumpy on a news group but then I reckon I have gone
past the age of 40 and have earned the do-as-you-bloody-well-like
license, and now been awarded and official permit from the GOT to be
grumpy, because I cannot do as I like
because the older I get the better I woz.

:-)

Understood: And taken in the spirit in which I am sure it
was intended:-)


We have a little chuckle now and then, and move right along, realizing
that nobody, not even ourselves, are perfect.

Erm. Don't you mean "realizing that nobody, not even
ourselves, IS perfect" ?

Well nobody plus one self would be plural, not singular, because there
are a heck of a lotta nobodies around.
If you had a somebody and you were quite out of your cotton pickin'
mind, then that'd be singular.
Well, single again, after getting unmarried yet again.

Patrick Turner.

Cheers:-)

Iain

[Patrick Turner]

Iain Churches wrote:

"John Byrns" wrote in message
...
In article ,
flipper wrote:

I find it hard to think positively of Morgan Jones ever since his goof
with the concertina cathode build out resistor in a magazine article and
the first edition of his book. IIRC he flip flopped in the second
edition but he should have known better in the first place. It just
shows that he didn't bother doing the math, which you should do when
making a recommendation that flies in the face of years conventional
practice.

Could you elaborate on what the goof was?

The goof was not a typo or an error in a formula, his idea of placing a
build out resistor in the cathode of a concertina phase inverter to
equalize the source impedances of the plate and cathode circuits was
simply a goofy idea. It was a "bright" idea intended to fix an imagined
problem that didn't actually exist, that instead created a real problem.

I saw the magazine article and the first edition of his book where he
presented this goofy idea, I have never seen the second edition of his
book to see how he extricated himself from the predicament he created
for himself, although I have been told by people that have seen the
second edition that he did somehow extricate himself.

John, can you get access to the 2nd edition of Valve Amplifiers?
It would be of interest to know if this refers to page 279 of the
paragraph which starts "Output resistance with both terminals
equally loaded" ?

Is not the large difference in output impedance
between anode and cathode an important
shortcoming of the concertina?

Not in most amps.

While the loads driven by the CPI remain equal, and nothing clips or
saturates, ie, as in a Williamson amp,
it matters not one bit if one phase is a current source and the other is
a voltage source.

Patrick Turner.

Iain

[John Byrns]

In article ,
"Iain Churches" wrote:

"John Byrns" wrote in message
...
In article ,
flipper wrote:

I find it hard to think positively of Morgan Jones ever since his goof
with the concertina cathode build out resistor in a magazine article and
the first edition of his book. IIRC he flip flopped in the second
edition but he should have known better in the first place. It just
shows that he didn't bother doing the math, which you should do when
making a recommendation that flies in the face of years conventional
practice.

Could you elaborate on what the goof was?

The goof was not a typo or an error in a formula, his idea of placing a
build out resistor in the cathode of a concertina phase inverter to
equalize the source impedances of the plate and cathode circuits was
simply a goofy idea. It was a "bright" idea intended to fix an imagined
problem that didn't actually exist, that instead created a real problem.

I saw the magazine article and the first edition of his book where he
presented this goofy idea, I have never seen the second edition of his
book to see how he extricated himself from the predicament he created
for himself, although I have been told by people that have seen the
second edition that he did somehow extricate himself.


John, can you get access to the 2nd edition of Valve Amplifiers?
It would be of interest to know if this refers to page 279 of the
paragraph which starts "Output resistance with both terminals
equally loaded" ?

Hi Iain,

I no longer have easy access to any of the Morgan Jones magazine
articles or books. The balance of the concertina phase inverter was
extensively discussed here in this group at least once in the past. The
simplest way to look at the balance question is to consider that
ignoring the slight effect caused by the grid to anode and grid to
cathode capacitances the same current flows through both the anode and
cathode circuits, therefore if the cathode and anode loads are made
equal the two output voltages will also be equal since the current in
each is the same.

How does the sentence you quote above, "Output resistance with both
terminals equally loaded", end and what does the next sentence say? I
think most everyone here eventually understood that the concertina was
inherently balanced irrespective of what the output resistances may be.
Relative to the output resistances I think Henry Pasternack and I
disagreed, although the disagreement was not about balance. I took a
conventional view that the two output impedances were different but that
the two output signals ended up equal because the source voltages for
the anode and cathode also differed and exactly compensated for the
differing output impedances, I think I posted all the relevant equations
supporting this view at the time. I basically looked at the anode and
cathode source impedances independently, although you do have to
consider the actual total anode and cathode loads when doing this
analysis.

On the other hand IIRC Henry Pasternack took the view that the anode and
cathode impedances were actually identical in operation because of the
way the signals appear on both the anode an cathode terminals. This
view does sort of workout, but you can't test the anode and cathode
impedances independently, you have to apply equal and opposite test
signals to each when measuring the impedance of either terminal.

Is not the large difference in output impedance
between anode and cathode an important
shortcoming of the concertina?

Yes, the large difference in output impedance is important, but only
when the concertina is driven into clipping, where the dynamically
changing anode and cathode load impedances cause very bizarre clipping
behavior. When the grid of the tube driven from the anode conducts, the
gain at the cathode doesn't change very much, however when the grid of
the tube being driven by the cathode conducts, the gain as seen at the
anode increases very greatly leading to grossly asymmetrical waveforms
at the two outputs. Morgan Jones build out resistor probably helps
moderate this bad clipping behavior, at the expense of balance below
clipping level. I had the impression that his point with the build out
resistor was to equalize the two source impedances so that the high
frequency response would not fall off faster on the anode due to
capacitive loading, although in reality the build out resistor actually
worsens the balance. Perhaps I misunderstood the reason he suggested
the build out resistor and that its real purpose was to mitigate the
poor clipping behavior to some extent. What he actually says in the
second edition may shed some light on what his original intention was.
It would also be interesting to know how he presents the balance issue
in the second edition, does he take the Pasternack or the Byrns approach
to the analysis? My favorite phase inverter is the floating paraphase,
think of it as an inverting opamp in a unity gain configuration.

--
Regards,

John Byrns

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

[Patrick Turner]

flipper wrote:

On Wed, 06 Aug 2008 12:10:02 GMT, Patrick Turner
wrote:



Iain Churches wrote:

"John Byrns" wrote in message
...
In article ,
flipper wrote:

I find it hard to think positively of Morgan Jones ever since his goof
with the concertina cathode build out resistor in a magazine article and
the first edition of his book. IIRC he flip flopped in the second
edition but he should have known better in the first place. It just
shows that he didn't bother doing the math, which you should do when
making a recommendation that flies in the face of years conventional
practice.

Could you elaborate on what the goof was?

The goof was not a typo or an error in a formula, his idea of placing a
build out resistor in the cathode of a concertina phase inverter to
equalize the source impedances of the plate and cathode circuits was
simply a goofy idea. It was a "bright" idea intended to fix an imagined
problem that didn't actually exist, that instead created a real problem.

I saw the magazine article and the first edition of his book where he
presented this goofy idea, I have never seen the second edition of his
book to see how he extricated himself from the predicament he created
for himself, although I have been told by people that have seen the
second edition that he did somehow extricate himself.

John, can you get access to the 2nd edition of Valve Amplifiers?
It would be of interest to know if this refers to page 279 of the
paragraph which starts "Output resistance with both terminals
equally loaded" ?

Is not the large difference in output impedance
between anode and cathode an important
shortcoming of the concertina?

Not in most amps.

While the loads driven by the CPI remain equal, and nothing clips or
saturates, ie, as in a Williamson amp,

Are you sure about 'nothing clips' in a Williamson?

I mean, with 20dB of NFB, when the output tubes clip the voltage amp
is going to try going 100 times the drive voltage and I don't see how
the LTP is going to manage that without going into positive grid drive
itself.

OK, so the amp begins to clip, so there isn't any increasing NFB signal
sent
to the V1 cathode, so V1 Vgk suddenly begins to increase A times, and
the balanced amp which is already having to make
maybe 40Vrms to drive the OP grids is asked to double that but into OP
tube grid current.
This causes the balanced amp to saturate with its own grid I and the
concertina gets into trouble.
But before the onset of OP tube clipping, the CPI only has to make about
2.5 Vrms to each grid of the
balanced amp, and because there isn't any saturation or clipping of
anything, the CPI
works really well with no strange artifact production.

Si while the OP tubes don't clip in a Wiliamson, none of the 6SN7 input
or driver tubes will.
The Willy amp was designed so that first the OP tubes clip, then the
balanced amp and then the CPI and then input.

The balanced amp in the W act as a buffer against the micro capacitive
FB from OP anodes to the OP grids.
So the CPI sees identical a and k loadings to a very high F.
But eventually capacitance catches up with the CPI, and anode output
falls before the cathode output does.
To ensure equal bandwidth at the CPI a and the k, some small value
trimmer C across the cathode R of the CPI is advisable.
15pF to 47pF is about right, and the value used is chosen to get a
symetrical square wave overshoot at the OPT sec.

The CPI is a stage with a lot of local current NFB, and as such acts as
a buffer between
the input tube and the Miller C of the balanced amp input.
The bandwidth of the Willy input/driver amp with 2 x 6SN7 is over 250kHz
at the OP tube grids, and is damned excellent.

I prefer the use of a single input driving an LTP into one side and with
with common cathode CCS.
Its not quite as "fast" as a Willy amp, but its slightly simpler, and
just as good when NFB is used.

But in some recent re-engineering of ARC and Manley amps, I have used an
input LTP with cathode CCS
to take the input to one grid and NFB to the other, then the output from
this
goes to a second LTP but with a common Rk taken to -120V, and this
worked very well indeed,
with almost no 2H in the distortion character that you get with V1 tube
as an SET.

So it does not matter too much what you use in an input stage and driver
stage
as long as the noise is SFA, bandwidth is over 150kHz for the output
voltage level needed to cause clipping
of the OP stage, and without shelving networks applied,
and distortion is less than 0.5% at the clipping level for the OP tubes,
and the voltage ability
is preferably twice what you need for OP tube clipping before the driver
amp clips itself
without the OP tubes in place; ie loaded by just the bias R cap coupled.

If you can easily satisfy all these design requirements, maybe your
input driver amp is OK.

There are several ways to build a tube amp which has little sonic
signature and
just presents us the music without smear, glare, coldness, hardness, or
damned "tube sound"
with all its faults from compromised engineering.

So, I can repeat this about the CPI, if used wisely,
"it matters not one bit if one phase is a current source and the other
is
a voltage source."

JB and HP had a long running brawl of words about the CPI way back in
2000
when I was told never to mention OPTs or leakage inductance on r.a.t
because it'd lead to a flame war
all over again.

In an ST70, a bootstrapped high gain input pentode must produce slightly
more signal that what
appears at each OP grid.
The CPI between input pentode and OP grids must be able to produce the
OP grid to grid signal,
ie, twice the OP grid signal, and if that was 30Vrms, then the CPI must
make 60Vrms.

When the OP tubes clip and draw grid I then the CPI is immediately in
serious trouble,
and the clipping is asymetrical and ugly.

But so darn what?

The ST70 is a miserly way to build an amplifier; a bean counter special;
something designed by accountants, and to an audiophile, barely
acceptable as entry level gear.
No matter, its purpose is for hi-fi, and the level was never ever meant
to be taken right
up to clipping. Teenagers were never invited to parties, or allowed near
the hi-fi controls
where they immediately press the loudness button, and turn up the bass
boost to max, then run everything
at window breking volume, which includes clipping, and fusing the
tweeters at least.

One of The Best sounding low power amplifiers under 12 watts is made
with a single
12AX7 with 1/2 as an input triode, and 1/2 as a CPI, and the OP tubes
are a pair of EL84 in 40% UL.
15dB GNFB used.

Countless commercial amps like Star etc used this basic recipe.
The EL84/6BQ5 is a very easy tube to drive.

Its a pity that in most integrated chassis where you'll find this
circuit for each channel,
the line stage amp, tone control amp and phono amp are Z grade, noisy,
inaccurate, poor BW, highly distorting
and always degrading the sound well before it hits the power amps.
Mainly because of bean counters trying to use less tubes, less chassis
iron, and less sockets
and less PSU filtering, and generally cheating Joe Public out of what he
paid for.

Patrick Turner.









Patrick Turner.

Iain

[tubegarden]

On Aug 7, 5:05�am, Patrick Turner wrote:

Mainly because of bean counters trying to use less tubes, less chassis
iron, and less sockets
and less PSU filtering, and generally cheating Joe Public out of what he
paid for.

Patrick Turner.



Hi RATs!

We each find something of interest in at least some portion of tube
amps

I like hearing what they sound like when I ...

What passes for Audio in the marketplace is very similar to what
passes for sex - in the marketplace.

"Can't buy me love ..."

Money and minds rarely mix ... Bill Gates is not the funniest software
(nor hardware) guy I ever met

Enjoy your experiments, they are joys beyond measure, once in a
while

Happy Ears!
Al

[Patrick Turner]

tubegarden wrote:

On Aug 7, 5:05�am, Patrick Turner wrote:

Mainly because of bean counters trying to use less tubes, less chassis
iron, and less sockets
and less PSU filtering, and generally cheating Joe Public out of what he
paid for.

Patrick Turner.

Hi RATs!

We each find something of interest in at least some portion of tube
amps

I like hearing what they sound like when I ...

What passes for Audio in the marketplace is very similar to what
passes for sex - in the marketplace.

"Can't buy me love ..."

Money and minds rarely mix ... Bill Gates is not the funniest software
(nor hardware) guy I ever met

Enjoy your experiments, they are joys beyond measure, once in a
while

Happy Ears!
Al

Hmm, I only get paid if my experiments give the best sound possible.

Sometimes the experiments are painful; they take such a long time and
the pay is very low - Hundreds of hours for just one pair of monoblocs.

In the market place the girls don't love you so the sex is cold and dull
and propelled by illusions.

Good sound must be found after a search for reality without
the chill of dull illusion.

Patrick Turner.

[John Byrns]

In article ,
Patrick Turner wrote:

The balanced amp in the W act as a buffer against the micro capacitive
FB from OP anodes to the OP grids.
So the CPI sees identical a and k loadings to a very high F.
But eventually capacitance catches up with the CPI, and anode output
falls before the cathode output does.
To ensure equal bandwidth at the CPI a and the k, some small value
trimmer C across the cathode R of the CPI is advisable.
15pF to 47pF is about right, and the value used is chosen to get a
symetrical square wave overshoot at the OPT sec.

Could you elaborate on this, why it is necessary, is it to compensate
for the effects of the capacitances of the CPI grid?

JB and HP had a long running brawl of words about the CPI way back in
2000
when I was told never to mention OPTs or leakage inductance on r.a.t
because it'd lead to a flame war
all over again.

As I remember it we both agreed that the circuit was basically balanced,
so the brawl must have been over the actual source impedance of the
anode and cathode circuits. Thinking about it over the last 24 hours, I
suspect that what HP was actually measuring/calculating was the source
impedance seen by a load driven as if it were connected between the
anode and cathode, or actually half of that value. I don't believe that
HP's contrived methodology actually demonstrates that the anode and
cathode source impedances are equal as he claimed.

One of The Best sounding low power amplifiers under 12 watts is made
with a single
12AX7 with 1/2 as an input triode, and 1/2 as a CPI, and the OP tubes
are a pair of EL84 in 40% UL.
15dB GNFB used.

Countless commercial amps like Star etc used this basic recipe.
The EL84/6BQ5 is a very easy tube to drive.

Its a pity that in most integrated chassis where you'll find this
circuit for each channel,
the line stage amp, tone control amp and phono amp are Z grade, noisy,
inaccurate, poor BW, highly distorting
and always degrading the sound well before it hits the power amps.
Mainly because of bean counters trying to use less tubes, less chassis
iron, and less sockets
and less PSU filtering, and generally cheating Joe Public out of what he
paid for.

Surely if the basic pair of EL84s and an ECC83 make a good sounding
power amplifier, there must be a complimentary "bean counter" approved
phono amp and tone control amp to go with it? I assume a "bean counter"
would allow two sockets in each stereo channel to be used in providing
these functions.

--
Regards,

John Byrns

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

[John Byrns]

In article ,
flipper wrote:

On Thu, 07 Aug 2008 16:20:50 -0500, John Byrns
wrote:

In article ,
Patrick Turner wrote:

The balanced amp in the W act as a buffer against the micro capacitive
FB from OP anodes to the OP grids.
So the CPI sees identical a and k loadings to a very high F.
But eventually capacitance catches up with the CPI, and anode output
falls before the cathode output does.
To ensure equal bandwidth at the CPI a and the k, some small value
trimmer C across the cathode R of the CPI is advisable.
15pF to 47pF is about right, and the value used is chosen to get a
symetrical square wave overshoot at the OPT sec.

Could you elaborate on this, why it is necessary, is it to compensate
for the effects of the capacitances of the CPI grid?

JB and HP had a long running brawl of words about the CPI way back in
2000
when I was told never to mention OPTs or leakage inductance on r.a.t
because it'd lead to a flame war
all over again.

As I remember it we both agreed that the circuit was basically balanced,
so the brawl must have been over the actual source impedance of the
anode and cathode circuits. Thinking about it over the last 24 hours, I
suspect that what HP was actually measuring/calculating was the source
impedance seen by a load driven as if it were connected between the
anode and cathode, or actually half of that value. I don't believe that
HP's contrived methodology actually demonstrates that the anode and
cathode source impedances are equal as he claimed.


I've seen the classic calculations showing the 'different' anode and
cathode source impedances, and it sounds logical, but it doesn't
behave that way.

What do you mean that "it doesn't behave that way", are you saying that
the classic calculations are incorrect?

I got into checking that with the 13FD7 'mini Williamson' amp because
I'm using large 'grid stoppers' as a HF roll off so the 'different'
source impedances became an issue, since they would be 'in series'
with the grid stopper resistors, but when I spiced it the thing acts
as if the source impedances are identical. I.E. there is no difference
in HF roll off (with equal value grid stoppers). To double check I
also spiced a standalone concertina with just capacitive loading and
got the same results.

The fact that "there is no difference in HF roll off" doesn't imply that
the anode and cathode source impedances are equal.

If, however, you accept that the concertina is balanced as long as the
two loads are equal then it's an 'of course' the roll off is the same
because the loads are equal, they just vary (equally) with frequency.

The response vs. frequency at the two outputs of the concertina are
balanced, the source impedances are not.

If you keep the 'different impedance' analysis the thing is as cathode
impedance drops anode gain increases in exact proportion to the drop
across the 'larger' anode impedance.

Or, if you use a single common source voltage model (grid), the effect
of FB is to make the anode and cathode source impedances appear
equal.

Only HP and you believe that the anode and cathode source impedances
appear equal. The "classic calculations" give the correct results, and
show that the anode and cathode source impedances are not equal, that
does not prevent the two outputs from being balanced when driving equal
loads.

You are ignoring a couple of important points that must be considered
when using the "classic calculations", if you want to get the correct
answer.

First when calculating the source impedance at the cathode you must
include the total anode load in the calculation, this means not just the
plate resistor, but also the coupling capacitor, the following grid
resistor, the shunt and miller capacitances of the following tube, et
al. Similarly when calculating the source impedance at the anode you
must include the total cathode load in the calculations.

Second, it isn't sufficient to calculate only the source impedances, you
must also calculate the source voltages at the anode and cathode.

The trick is that the anode and cathode source voltages vary with
frequency in such a way that they cancel the effects of the differing
source impedances driving the two loads, the result being that the two
output voltages remain balanced even as frequency varies.

I'm not good enough to do the math but George E. Jones Jr did it in a
1951 paper on the subject.

http://www.diybanter.com/attachment....2&d=1213179423

I will take a look at that article, does it come to the conclusion that
the anode and source impedances are equal?

--
Regards,

John Byrns

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

[Patrick Turner]

John Byrns wrote:

In article ,
Patrick Turner wrote:

The balanced amp in the W act as a buffer against the micro capacitive
FB from OP anodes to the OP grids.
So the CPI sees identical a and k loadings to a very high F.
But eventually capacitance catches up with the CPI, and anode output
falls before the cathode output does.
To ensure equal bandwidth at the CPI a and the k, some small value
trimmer C across the cathode R of the CPI is advisable.
15pF to 47pF is about right, and the value used is chosen to get a
symetrical square wave overshoot at the OPT sec.

Could you elaborate on this, why it is necessary, is it to compensate
for the effects of the capacitances of the CPI grid?

Its not absoluely necessary. It just makes a prettier looking wave above
50khZ.

But if you ever do measure a Wiliamson type amp, or one with a CPU drive
to the OP tubes,
then you'll see that the HF response of the anode output of the CPI sags
before the cathode.
Some slight C added to the cathode Rk will make the sag even on both
anode and cathode.

The Ren7070 VAC amp I have just re-wired for a guy had 47pF across a 22k
in the CPI.
It also had cross coupled neutralising C from anode connections to the
OPT back to the
oposite grid of the OP 300B. This boosts the HF response with what is
positive FB.
I abolished this utter BS because the amp has selectable amounts of NFB
and
in the 10dB FB selection, there was very lousy HF stability.

But I left the 47pF across the CPI cathode Rk.



JB and HP had a long running brawl of words about the CPI way back in
2000
when I was told never to mention OPTs or leakage inductance on r.a.t
because it'd lead to a flame war
all over again.

As I remember it we both agreed that the circuit was basically balanced,
so the brawl must have been over the actual source impedance of the
anode and cathode circuits.

At the time, you graciously suffered daily broadsides of being called a
complete idiot.
And of course we know you ain't, and Pasternak is nowhere to be seen.


Thinking about it over the last 24 hours, I
suspect that what HP was actually measuring/calculating was the source
impedance seen by a load driven as if it were connected between the
anode and cathode, or actually half of that value. I don't believe that
HP's contrived methodology actually demonstrates that the anode and
cathode source impedances are equal as he claimed.

If you measure the Rout of a CPI SEPARATELY at its anode, its high, and
at the cathode ITS LOW.
but because at AF the load at a and k remain the same, so too does the
balance.



One of The Best sounding low power amplifiers under 12 watts is made
with a single
12AX7 with 1/2 as an input triode, and 1/2 as a CPI, and the OP tubes
are a pair of EL84 in 40% UL.
15dB GNFB used.

Countless commercial amps like Star etc used this basic recipe.
The EL84/6BQ5 is a very easy tube to drive.

Its a pity that in most integrated chassis where you'll find this
circuit for each channel,
the line stage amp, tone control amp and phono amp are Z grade, noisy,
inaccurate, poor BW, highly distorting
and always degrading the sound well before it hits the power amps.
Mainly because of bean counters trying to use less tubes, less chassis
iron, and less sockets
and less PSU filtering, and generally cheating Joe Public out of what he
paid for.

Surely if the basic pair of EL84s and an ECC83 make a good sounding
power amplifier, there must be a complimentary "bean counter" approved
phono amp and tone control amp to go with it? I assume a "bean counter"
would allow two sockets in each stereo channel to be used in providing
these functions.

You have not had the displeasure to fully examine the woeful attempts by
the army of bean counters
who controlled what the public bought in 1960.

I have, and in nearly every crummy integrated amp, receiver or power +
pre set the
performance left a lot to be desired.
I don't get many sets from 1960 to re-engineer now as I used to 10 years
ago.
Their owners are usually getting old and deaf, and need to be
re-engineered
themselves, just to stay alive.

Most MM phono stages in 1960 consisted of ONE EF86, as in Quad22,
bleedin awful,
or two 1/2 of a 12AX7, which was better IMHO, because there was more NFB
and the response was more
predictable. But most had only nominal adherance to RIAA eq, and actual
responses were
+/- 3dB and different for each channel. Non deleatable tone controls
with similar defects and which
had a passive network driven with 1/2 a 12AX7 were horrid.

Bean counters ensured that only one tube socket and tube was ever to be
used for a single phono channel,
and definately not two sockets and tubes.

Using a µ-follower stage with a single 12AX7 was utterly out of the
question, and made bean counters
have a fit of apoplexy and foam at the mouth when some innocent young
engineer trotted
into the office with prospective design for next year's models.

It was with great glee that bean counters adopted solid state devices to
replace tubes.
They replaced the music as well with something else.

Patrick Turner.

--
Regards,

John Byrns

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

[Patrick Turner]

The Willy amp was designed so that first the OP tubes clip, then the
balanced amp and then the CPI and then input.

I think that pretty much goes for any reasonably designed topology.
You don't want the voltage stages/PS clipping before the OP tubes,
unless maybe you're making a compressor or intentionally creating
distortion like in an overdriven guitar preamp or pedal.

You got it.

The balanced amp in the W act as a buffer against the micro capacitive
FB from OP anodes to the OP grids.
So the CPI sees identical a and k loadings to a very high F.
But eventually capacitance catches up with the CPI, and anode output
falls before the cathode output does.

Are you speaking of inter electrode capacitances in the CPI itself or
the balanced amp grids? Because, if you're speaking of the balanced
amp grids the roll off is the same despite the seemingly different CPI
anode/cathode impedances.

The variations in load at the OP tube anodes cause a change in OP tube
gain.
The Miller C varies, so the driver stage experiences a change in load as
well.

If the load on the OP anodes is say low, and then gain is low, so Miller
C is low,
so the load from Miller makes driver stage gain higher.
This might be especially true of a beam tetrode output stage driving
ESL speakers where the C of the speaker means the load goes below 2
ohms, at 18kHz, as in Quad ESL57.

In a W amp, the CPI is largely buffered from the effects of the Miller C
of the OP stage and
the change of gain or change of output balance in the balanced drive amp
would not have a huge effect on the CPI, let alone on the input stage.

To ensure equal bandwidth at the CPI a and the k, some small value
trimmer C across the cathode R of the CPI is advisable.
15pF to 47pF is about right, and the value used is chosen to get a
symetrical square wave overshoot at the OPT sec.

The CPI is a stage with a lot of local current NFB, and as such acts as
a buffer between
the input tube and the Miller C of the balanced amp input.
The bandwidth of the Willy input/driver amp with 2 x 6SN7 is over 250kHz
at the OP tube grids, and is damned excellent.

I didn't say it wasn't 'excellent'. Just said it clips.

I prefer the use of a single input driving an LTP into one side and with
with common cathode CCS.
Its not quite as "fast" as a Willy amp, but its slightly simpler, and
just as good when NFB is used.

But in some recent re-engineering of ARC and Manley amps, I have used an
input LTP with cathode CCS
to take the input to one grid and NFB to the other, then the output from
this
goes to a second LTP but with a common Rk taken to -120V, and this
worked very well indeed,
with almost no 2H in the distortion character that you get with V1 tube
as an SET.

So it does not matter too much what you use in an input stage and driver
stage
as long as the noise is SFA, bandwidth is over 150kHz for the output
voltage level needed to cause clipping
of the OP stage, and without shelving networks applied,
and distortion is less than 0.5% at the clipping level for the OP tubes,
and the voltage ability
is preferably twice what you need for OP tube clipping before the driver
amp clips itself
without the OP tubes in place; ie loaded by just the bias R cap coupled.

If you can easily satisfy all these design requirements, maybe your
input driver amp is OK.

There are several ways to build a tube amp which has little sonic
signature and
just presents us the music without smear, glare, coldness, hardness, or
damned "tube sound"
with all its faults from compromised engineering.

So, I can repeat this about the CPI, if used wisely,
"it matters not one bit if one phase is a current source and the other
is
a voltage source."

It may not matter with excellent OPTs but with the less than ideal
OPTs I use CPI clipping can cause HF bursts and the 'excellent'
bandwidth along with OP triodes makes it more difficult to control
than a 'plain Jane' CPI into a pair of 6BQ5s. Mainly, I think, because
that lovely low impedance triode FB more readily shoves the phase
shifted crap through the OPT.

The low Ra or low driving source resistance of the OP triodes mean that
the the second order filtering effect of the poorer OPT
becomes undamped, and the response is peaked at the sec, has a high
phase shift
and any FB becomes positive all too easily.

Triode OP stage are wonderful sounding things, but have bothers with
damping.
The VAC 7070 I have just worked on is no exception with a quad of 300B
driving an OPT which is no better than a Hammond with regard LL and
Cshunt,
so its very necessary to reduce the OLG of the whole input drive amp
with a Zobel across the V1 anode load, and have a Zobel at the OPT sec,
to get unconditional stability
with 12dB GNFB.

Now I have a pair of Audion horrors to fix.
Same sort of problems.
Un-optimised circuitry, N&D way too high, poor stability.


JB and HP had a long running brawl of words about the CPI way back in
2000
when I was told never to mention OPTs or leakage inductance on r.a.t
because it'd lead to a flame war
all over again.

In an ST70, a bootstrapped high gain input pentode must produce slightly
more signal that what
appears at each OP grid.
The CPI between input pentode and OP grids must be able to produce the
OP grid to grid signal,
ie, twice the OP grid signal, and if that was 30Vrms, then the CPI must
make 60Vrms.

When the OP tubes clip and draw grid I then the CPI is immediately in
serious trouble,
and the clipping is asymetrical and ugly.

Yep, if you don't do something about it.


But so darn what?

Oscillation bursts is potentially what.

Squegging, infact...


The ST70 is a miserly way to build an amplifier; a bean counter special;
something designed by accountants, and to an audiophile, barely
acceptable as entry level gear.
No matter, its purpose is for hi-fi, and the level was never ever meant
to be taken right
up to clipping. Teenagers were never invited to parties, or allowed near
the hi-fi controls
where they immediately press the loudness button, and turn up the bass
boost to max, then run everything
at window breking volume, which includes clipping, and fusing the
tweeters at least.

I hate to ruin your favorite whipping boy but 'bean counters' don't
'pick parts' or perform any other design function. They simply 'count
beans' and tell you the sum.

Bean counters are employed to say no to engineers.
They keep on saying no until the engineers present a design that's
cheaper than the competitors are making.

The engineers go to Confession before Communion,

"Bless me father, for I have Sinned, I sold my ****in soul to the
Devil..."

" No need to swear my son, three Our Fathers, and six Hail Marys "

And so it was that the people of America were jilted at the shop when
they bought anything.


One of The Best sounding low power amplifiers under 12 watts is made
with a single
12AX7 with 1/2 as an input triode, and 1/2 as a CPI, and the OP tubes
are a pair of EL84 in 40% UL.
15dB GNFB used.

I wonder what would happen if they tried 20dB of GNFB and are they
using grid stoppers on the OP tubes?

The grid stoppers don't do a lot in polite hi-fi amps.
Quad-II don't have them between EF86 drivers and KT66 OP tubes.
And really only need to be used to prevent excessive grid currents, or
as a kind of limiter
to prevent rapid rise of bias in the coupling caps on over drive,
something never experienced
in hi-fi listening.
OK, you see them in an Ampeg SVT at 47k per grid of 6550, but the
6550 are in beam tet mode so Miller is low, and there's a high anode
voltage for huge PO levels
and anything can and does happen with guitar amps when driven into heavy
overdrive.



Countless commercial amps like Star etc used this basic recipe.
The EL84/6BQ5 is a very easy tube to drive.

I know. I did the same thing in my 'Stealth AX; amp, except I'm using
the electrically identical 6GK6..

Its a pity that in most integrated chassis where you'll find this
circuit for each channel,
the line stage amp, tone control amp and phono amp are Z grade, noisy,
inaccurate, poor BW, highly distorting
and always degrading the sound well before it hits the power amps.
Mainly because of bean counters trying to use less tubes, less chassis
iron, and less sockets

Bean counters count beans, they do not design a damn thing..

and less PSU filtering, and generally cheating Joe Public out of what he
paid for.

In a free market Joe Public gets exactly what he paid for and is free
to pay more if he wants more.

My point is that he doesn't get what he paid for.

He wants this and that, and reads the sales blurbs, get's fooled into
thinking
the sales blurbs cover what he thought he wanted, and pays out even in
doubt because
its the same sorry story at every brandname.

Excellence was never wasted upon the masses.

There was remarkably good electronics in the moon shot rockets from 1969
onwards,
but it didn't extend to Dynacos and Fords.

Shareholders, CEOs, marketeers, and bean counters all have to get a big
slice of the pie, not leaving
much for Joe Consumer.

Take a packet of chips from the supermarket.

You'd stave to death if that's all you ate, and probably be poisoned.

BS rules, OK.

Its the modern western capitalist way, bloody awful, until you think of
the alternatives.

In 1960, I doubt there was a Russian amplifier that was worth buying,
and there was,
you'd have to wait years for one, or know someone in the Party to get
one,
and it was a similar deal with the "amazingly innovative and economical
Trabant" East German limozine :-) !!!!

Patrick Turner.





Patrick Turner.

[Henry Pasternack[_2_]]

Flipper, do you have a real email address?

-Henry

moc.ncr@kcanretsaph

[John Byrns]

In article ,
flipper wrote:

On Fri, 08 Aug 2008 10:11:02 -0500, John Byrns
wrote:

In article ,
flipper wrote:

On Thu, 07 Aug 2008 16:20:50 -0500, John Byrns
wrote:

In article ,
Patrick Turner wrote:

The balanced amp in the W act as a buffer against the micro capacitive
FB from OP anodes to the OP grids.
So the CPI sees identical a and k loadings to a very high F.
But eventually capacitance catches up with the CPI, and anode output
falls before the cathode output does.
To ensure equal bandwidth at the CPI a and the k, some small value
trimmer C across the cathode R of the CPI is advisable.
15pF to 47pF is about right, and the value used is chosen to get a
symetrical square wave overshoot at the OPT sec.

Could you elaborate on this, why it is necessary, is it to compensate
for the effects of the capacitances of the CPI grid?

JB and HP had a long running brawl of words about the CPI way back in
2000
when I was told never to mention OPTs or leakage inductance on r.a.t
because it'd lead to a flame war
all over again.

As I remember it we both agreed that the circuit was basically balanced,
so the brawl must have been over the actual source impedance of the
anode and cathode circuits. Thinking about it over the last 24 hours, I
suspect that what HP was actually measuring/calculating was the source
impedance seen by a load driven as if it were connected between the
anode and cathode, or actually half of that value. I don't believe that
HP's contrived methodology actually demonstrates that the anode and
cathode source impedances are equal as he claimed.


I've seen the classic calculations showing the 'different' anode and
cathode source impedances, and it sounds logical, but it doesn't
behave that way.

What do you mean that "it doesn't behave that way", are you saying that
the classic calculations are incorrect?

I explained what I meant in the subsequent text.

I got into checking that with the 13FD7 'mini Williamson' amp because
I'm using large 'grid stoppers' as a HF roll off so the 'different'
source impedances became an issue, since they would be 'in series'
with the grid stopper resistors, but when I spiced it the thing acts
as if the source impedances are identical. I.E. there is no difference
in HF roll off (with equal value grid stoppers). To double check I
also spiced a standalone concertina with just capacitive loading and
got the same results.

The fact that "there is no difference in HF roll off" doesn't imply that
the anode and cathode source impedances are equal.

I think most people would think it implies just that and in most
circumstances it does. For example, if you're designing an RIAA filter
you include the source impedance of the driving stage in the
calculations and if that is 'different' then so would be the
compensation. Same thing with a guitar tone stack. If you design a
Marshal style tone stack driven by a low impedance cathode follower
and then, instead, take from the anode output you get a 'different'
response because you have a 'different' source impedance.


If, however, you accept that the concertina is balanced as long as the
two loads are equal then it's an 'of course' the roll off is the same
because the loads are equal, they just vary (equally) with frequency.

The response vs. frequency at the two outputs of the concertina are
balanced, the source impedances are not.

Makes calculating 1/f=2Pi RC a bitch, don't it?

If you keep the 'different impedance' analysis the thing is as cathode
impedance drops anode gain increases in exact proportion to the drop
across the 'larger' anode impedance.

Or, if you use a single common source voltage model (grid), the effect
of FB is to make the anode and cathode source impedances appear
equal.

Only HP and you believe that the anode and cathode source impedances
appear equal.

I said no such thing. I talked about both models but made not one
comment about which I 'believed' or preferred.

A more reasonable interpretation would be that I suggested both views
are simply different ways of looking at the same thing.

The "classic calculations" give the correct results, and
show that the anode and cathode source impedances are not equal, that
does not prevent the two outputs from being balanced when driving equal
loads.

You are ignoring a couple of important points that must be considered
when using the "classic calculations", if you want to get the correct
answer.

First when calculating the source impedance at the cathode you must
include the total anode load in the calculation, this means not just the
plate resistor,

I'm not ignoring anything and my referencing the classic equations
should have given you a clue I'm including all of them.

but also the coupling capacitor, the following grid
resistor, the shunt and miller capacitances of the following tube, et
al.

They ain't there in the stand alone CPI test I ran.

Nor does it make any difference to the question at hand because
they're equal and covered by the caveat of equal load impedances. A
theorem I verified by running both a 'whole amp' and standalone CPI
simulation.

Similarly when calculating the source impedance at the anode you
must include the total cathode load in the calculations.

Second, it isn't sufficient to calculate only the source impedances, you
must also calculate the source voltages at the anode and cathode.

The trick is that the anode and cathode source voltages vary with
frequency in such a way that they cancel the effects of the differing
source impedances driving the two loads, the result being that the two
output voltages remain balanced even as frequency varies.

Did you bother to read what I said? Because I said exactly that for
the 'different impedance' model.

However, the whole point to 'equivalent impedance' calculations is so
the 'simplified' term can be used in subsequent calculations, like in
the RIAA filter I mentioned above. But, in the special case of an
equally loaded CPI, it doesn't provide any useful differentiation
because, while you can argue all day long that the impedances are
'different', there is no difference in the behavior.

As I said, and to which you decided to 'argue' with by saying the same
thing, the 'trick' is "as cathode impedance drops anode gain increases
in exact proportion to the drop across the 'larger' anode impedance."
I.E. the 'source voltages' change.

However, another way to look at it is, with equal load impedances, you
have a local feedback loop that lowers the 'effective impedance' of
the anode and that's not a particularly odd view as it's conceivably a
similar mechanism to that used in lowering amplifier 'effective output
impedance' with GNFB. Or, conversely (again), if one insists on saying
the reflected pentode impedance 'is' the (actual) impedance
(equivalent to arguing the CPI source impedances are, 'in fact',
different) then the 'source voltage' is changing to compensate for the
varying load impedance.

The mechanism, in either case, falls apart if the load impedances are
not the same, like if the following grid goes positive (since they
don't do so at the same time) and then the 'different' source
impedances becomes seemingly apparent. Or the feedback loop goes
cockeyed (as does GNFB at clipping) because the special case it
depends on is no longer valid.

I say "seemingly" for the 'different impedance' model because, again,
it does not behave as one would expect from a 'simplified' impedance
calculation since, when the cathode goes +ve into the following grid,
the anode signal changes dramatically even though there is no
'difference' in the load *it is seeing* (at that point in time). And,
no, I'm not 'ignoring' that the anode source impedance has changed
because of the change in cathode impedance, I'm simply saying that
things fall apart so these 'simplified' equations (calculated for the
equal load case) no longer hold.


I'm not good enough to do the math but George E. Jones Jr did it in a
1951 paper on the subject.

http://www.diybanter.com/attachment....2&d=1213179423

I will take a look at that article, does it come to the conclusion that
the anode and source impedances are equal?

He wasn't dealing with that 'argument', although the 'different source
impedance' model would seem to clearly be what leads people to the
erroneous 'imbalanced at HF' conclusion, but I suppose it depends on
how you look at the results. The roll off can be approximated with
cathode source impedance and the roll off is the same for the plate
side so one might conclude it has the same effective impedance,
because that is the apparent behavior.


Flipper, you seem to be flopping all around, trying to walk both sides
of the street, or perhaps trying to swim both sides of the pier would be
more accurate.

The anode and cathode source impedances are either the same or they
aren't the same, you can't have it both ways, you have to pick one or
the other, they can't both be true. Once you have made your choice,
then you can work on trying to justify it.

I still don't see how the fact that the two output voltages remain equal
as the frequency is varied, in anyway implies that the anode and cathode
source impedances are equal?

Of course the effects of grid to anode and grid to cathode capacitance
do unbalance the concertina at very high frequencies, but I leave that
for Patrick to worry about, from my perspective it isn't a problem worth
troubling over.

--
Regards,

John Byrns

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

[Patrick Turner]

John Byrns wrote:

In article ,
flipper wrote:

On Fri, 08 Aug 2008 10:11:02 -0500, John Byrns
wrote:

In article ,
flipper wrote:

On Thu, 07 Aug 2008 16:20:50 -0500, John Byrns
wrote:

In article ,
Patrick Turner wrote:

The balanced amp in the W act as a buffer against the micro capacitive
FB from OP anodes to the OP grids.
So the CPI sees identical a and k loadings to a very high F.
But eventually capacitance catches up with the CPI, and anode output
falls before the cathode output does.
To ensure equal bandwidth at the CPI a and the k, some small value
trimmer C across the cathode R of the CPI is advisable.
15pF to 47pF is about right, and the value used is chosen to get a
symetrical square wave overshoot at the OPT sec.

Could you elaborate on this, why it is necessary, is it to compensate
for the effects of the capacitances of the CPI grid?

JB and HP had a long running brawl of words about the CPI way back in
2000
when I was told never to mention OPTs or leakage inductance on r.a.t
because it'd lead to a flame war
all over again.

As I remember it we both agreed that the circuit was basically balanced,
so the brawl must have been over the actual source impedance of the
anode and cathode circuits. Thinking about it over the last 24 hours, I
suspect that what HP was actually measuring/calculating was the source
impedance seen by a load driven as if it were connected between the
anode and cathode, or actually half of that value. I don't believe that
HP's contrived methodology actually demonstrates that the anode and
cathode source impedances are equal as he claimed.


I've seen the classic calculations showing the 'different' anode and
cathode source impedances, and it sounds logical, but it doesn't
behave that way.

What do you mean that "it doesn't behave that way", are you saying that
the classic calculations are incorrect?

I explained what I meant in the subsequent text.

I got into checking that with the 13FD7 'mini Williamson' amp because
I'm using large 'grid stoppers' as a HF roll off so the 'different'
source impedances became an issue, since they would be 'in series'
with the grid stopper resistors, but when I spiced it the thing acts
as if the source impedances are identical. I.E. there is no difference
in HF roll off (with equal value grid stoppers). To double check I
also spiced a standalone concertina with just capacitive loading and
got the same results.

The fact that "there is no difference in HF roll off" doesn't imply that
the anode and cathode source impedances are equal.

I think most people would think it implies just that and in most
circumstances it does. For example, if you're designing an RIAA filter
you include the source impedance of the driving stage in the
calculations and if that is 'different' then so would be the
compensation. Same thing with a guitar tone stack. If you design a
Marshal style tone stack driven by a low impedance cathode follower
and then, instead, take from the anode output you get a 'different'
response because you have a 'different' source impedance.


If, however, you accept that the concertina is balanced as long as the
two loads are equal then it's an 'of course' the roll off is the same
because the loads are equal, they just vary (equally) with frequency.

The response vs. frequency at the two outputs of the concertina are
balanced, the source impedances are not.

Makes calculating 1/f=2Pi RC a bitch, don't it?

If you keep the 'different impedance' analysis the thing is as cathode
impedance drops anode gain increases in exact proportion to the drop
across the 'larger' anode impedance.

Or, if you use a single common source voltage model (grid), the effect
of FB is to make the anode and cathode source impedances appear
equal.

Only HP and you believe that the anode and cathode source impedances
appear equal.

I said no such thing. I talked about both models but made not one
comment about which I 'believed' or preferred.

A more reasonable interpretation would be that I suggested both views
are simply different ways of looking at the same thing.

The "classic calculations" give the correct results, and
show that the anode and cathode source impedances are not equal, that
does not prevent the two outputs from being balanced when driving equal
loads.

You are ignoring a couple of important points that must be considered
when using the "classic calculations", if you want to get the correct
answer.

First when calculating the source impedance at the cathode you must
include the total anode load in the calculation, this means not just the
plate resistor,

I'm not ignoring anything and my referencing the classic equations
should have given you a clue I'm including all of them.

but also the coupling capacitor, the following grid
resistor, the shunt and miller capacitances of the following tube, et
al.

They ain't there in the stand alone CPI test I ran.

Nor does it make any difference to the question at hand because
they're equal and covered by the caveat of equal load impedances. A
theorem I verified by running both a 'whole amp' and standalone CPI
simulation.

Similarly when calculating the source impedance at the anode you
must include the total cathode load in the calculations.

Second, it isn't sufficient to calculate only the source impedances, you
must also calculate the source voltages at the anode and cathode.

The trick is that the anode and cathode source voltages vary with
frequency in such a way that they cancel the effects of the differing
source impedances driving the two loads, the result being that the two
output voltages remain balanced even as frequency varies.

Did you bother to read what I said? Because I said exactly that for
the 'different impedance' model.

However, the whole point to 'equivalent impedance' calculations is so
the 'simplified' term can be used in subsequent calculations, like in
the RIAA filter I mentioned above. But, in the special case of an
equally loaded CPI, it doesn't provide any useful differentiation
because, while you can argue all day long that the impedances are
'different', there is no difference in the behavior.

As I said, and to which you decided to 'argue' with by saying the same
thing, the 'trick' is "as cathode impedance drops anode gain increases
in exact proportion to the drop across the 'larger' anode impedance."
I.E. the 'source voltages' change.

However, another way to look at it is, with equal load impedances, you
have a local feedback loop that lowers the 'effective impedance' of
the anode and that's not a particularly odd view as it's conceivably a
similar mechanism to that used in lowering amplifier 'effective output
impedance' with GNFB. Or, conversely (again), if one insists on saying
the reflected pentode impedance 'is' the (actual) impedance
(equivalent to arguing the CPI source impedances are, 'in fact',
different) then the 'source voltage' is changing to compensate for the
varying load impedance.

The mechanism, in either case, falls apart if the load impedances are
not the same, like if the following grid goes positive (since they
don't do so at the same time) and then the 'different' source
impedances becomes seemingly apparent. Or the feedback loop goes
cockeyed (as does GNFB at clipping) because the special case it
depends on is no longer valid.

I say "seemingly" for the 'different impedance' model because, again,
it does not behave as one would expect from a 'simplified' impedance
calculation since, when the cathode goes +ve into the following grid,
the anode signal changes dramatically even though there is no
'difference' in the load *it is seeing* (at that point in time). And,
no, I'm not 'ignoring' that the anode source impedance has changed
because of the change in cathode impedance, I'm simply saying that
things fall apart so these 'simplified' equations (calculated for the
equal load case) no longer hold.


I'm not good enough to do the math but George E. Jones Jr did it in a
1951 paper on the subject.

http://www.diybanter.com/attachment....2&d=1213179423

I will take a look at that article, does it come to the conclusion that
the anode and source impedances are equal?

He wasn't dealing with that 'argument', although the 'different source
impedance' model would seem to clearly be what leads people to the
erroneous 'imbalanced at HF' conclusion, but I suppose it depends on
how you look at the results. The roll off can be approximated with
cathode source impedance and the roll off is the same for the plate
side so one might conclude it has the same effective impedance,
because that is the apparent behavior.

Flipper, you seem to be flopping all around, trying to walk both sides
of the street, or perhaps trying to swim both sides of the pier would be
more accurate.

The anode and cathode source impedances are either the same or they
aren't the same, you can't have it both ways, you have to pick one or
the other, they can't both be true. Once you have made your choice,
then you can work on trying to justify it.

I still don't see how the fact that the two output voltages remain equal
as the frequency is varied, in anyway implies that the anode and cathode
source impedances are equal?

Of course the effects of grid to anode and grid to cathode capacitance
do unbalance the concertina at very high frequencies, but I leave that
for Patrick to worry about, from my perspective it isn't a problem worth
troubling over.

Well, at very HF over say 50kHz, capacitances begin to lower the
way the CPI is loaded at its anode and cathode.

The cathode output tends to have a higher F pole than at the anode.

There us *some* Miller C in the CPI, because the anode
signal is inverted, and effectively, the C between g and a is more than
between g and k.

So the CPI becomes loaded with a lower load at the anode than cathodeand
hence the difference in output levels at HF.

Taken to extreme, consider what happens at 1MHz.

If Cga was say 10pF, ZC = 16k, and so there is the V1 anode Rout in
series with
16k in series with RLa of the CPI triode, and the Cgk is lower, but the
cathode output
is from "follower" action, and anode output falls before cathode output.

The capacitances involved manage to become such low impedances as F
rises they saturate the opeation of the tubes
and the triodes clip. But for ordinary amp builders,
they needn't be too worried and any old CPI works quite well as long as
you don't expect
too much voltage swing or you overload it or expect it to have F2 too
high.


I prefer having an SET input tube using a paralleled twin triode such as
6CG7 or 6DJ8
and then drive an LTP woth a pair of triodes such as two 6CG7 tubes, and
have a CCS cathode sink.
This is best set up so it makes 50% more voltage than needed for output
stage clipping,
and you'll find it overloads symetrically, or at + and - voltage peaks
together.

The Williamson done right sounds just great though.

About the only improvements anyone needs to think about is adding a
small C across the
Rk of the CPI, and then having a long tail resistor from the balanced
amp to say a -120V rail;
about 13k is OK if the Ia in each 1/2 of the balanced amp = 5mA.

This long tail makes the two phases of the balanced driver amp stay
equal
and ensures that much better cancelation of 2H occurs and stops 2H
appearing at each
anode of the balanced amp. The balanced amp creates the most distortion,
and should be a small quantity
of mainly 3H.

Patrick Turner.



--
Regards,

John Byrns

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