[West]

AC filaments mean less components (diodes, caps) and from what I read way
back are superior for certain tubes. Can DC can cause uneven cathode
stripping or filament wear? Is this true? If so, how can we protect our DC
wired tubes? Are there certain tubes that fare better with AC for their
filaments? I know this topic has been beaten to death before, but please
indulge me, if you will. Thanks.

west

[Iain Churches[_2_]]

"West" wrote in message news:%3LGi.347$6o2.294@trnddc05...
AC filaments mean less components (diodes, caps) and from what I read way
back are superior for certain tubes. Can DC can cause uneven cathode
stripping or filament wear? Is this true? If so, how can we protect our DC
wired tubes? Are there certain tubes that fare better with AC for their
filaments? I know this topic has been beaten to death before, but please
indulge me, if you will. Thanks.


Hi West. I am in contact with a very interesting tube amp builder (who
probably doesn't want his name mentioned here) He is a former BBC
man, with published material. He knows what he is talking about.

He states that heaters run from AC have a longer working life,
and that DC heaters always fail at the same end, (Would
that be the positive end?) He has also told me that heaters should be
biased above the cathode (for which you need a floating heater supply)

This is particularly so in two-section topology like the cascode or
mu-follower where the higher cathode may be 140V or so above ground.
In this case the heaters should be biased within the heater to cathode
voltage limits stated by the manufacturer. 90V in such a case is common.

In a discussion about heater regulators, he was of the opinion that it is
the *current* which needs to be regulated and not the voltage, as a supply
within 5% tolerance is quite acceptable. The cold heater draws several
times its nominal current.

Best regards
Iain





west

[West]

"Iain Churches" wrote in message
ti.fi...

"West" wrote in message
news:%3LGi.347$6o2.294@trnddc05...
AC filaments mean less components (diodes, caps) and from what I read
way
back are superior for certain tubes. Can DC can cause uneven cathode
stripping or filament wear? Is this true? If so, how can we protect our
DC
wired tubes? Are there certain tubes that fare better with AC for their
filaments? I know this topic has been beaten to death before, but please
indulge me, if you will. Thanks.


Hi West. I am in contact with a very interesting tube amp builder (who
probably doesn't want his name mentioned here) He is a former BBC
man, with published material. He knows what he is talking about.

He states that heaters run from AC have a longer working life,
and that DC heaters always fail at the same end, (Would
that be the positive end?) He has also told me that heaters should be
biased above the cathode (for which you need a floating heater supply)

This is particularly so in two-section topology like the cascode or
mu-follower where the higher cathode may be 140V or so above ground.
In this case the heaters should be biased within the heater to cathode
voltage limits stated by the manufacturer. 90V in such a case is common.

In a discussion about heater regulators, he was of the opinion that it is
the *current* which needs to be regulated and not the voltage, as a supply
within 5% tolerance is quite acceptable. The cold heater draws several
times its nominal current.

Best regards
Iain

Thanks for the interesting post, Iain. This is what I am alluding to. But I
would like to eventually know if this problem with DC is for all type tubes
(Output,vs. small signal, Direct vs. Indirect, etc.). Also tube life can be
hastened with filament voltage above and below spec.I thought that if you
lower the voltage a tad it will prolong tube life. Perhaps there are 2
schools of thought.
A cold filament can draw more than 7 times its steady state current (ouch).
This also hastens the tube's life. Finally, if you change polarity from time
to time on DC filaments, I believe you will resolve the uneven stripping.
Nice hearing again from you.

Cordially,
west

west

[Andre Jute]

On Sep 14, 11:36 pm, "West" wrote:
AC filaments mean less components (diodes, caps) and from what I read way
back are superior for certain tubes. Can DC can cause uneven cathode
stripping or filament wear? Is this true? If so, how can we protect our DC
wired tubes? Are there certain tubes that fare better with AC for their
filaments? I know this topic has been beaten to death before, but please
indulge me, if you will. Thanks.

west

Google the Sound List, Westley. This is the list of experts often
referred to as the Joelist. The archive is kept at Harvard. About ten
years ago, maybe 1996/7, there was a long learned discussion of the
subject, and it was decided AC fils plus a raised DC voltage (I seem
to remember 50-90V) plus occasional fil polarity swapping would do the
trick. No one had ever seen cathode stripping... Personally, while I
agree with everything Iain reports his friend as saying, I always show
the filament lift in my designs but often leave it out of what I build
for myself; nor can I remember even swapping fil ends; yet I have
never seen a tube ruined by it. We can be too careful. Nontheless, a
necessary subject, because we can also take too much for granted.

We should get a formula down for determining the exact voltage that
will do the trick. Patrick, this is a job for the resident guru.

Andre Jute
Visit Jute on Amps at http://members.lycos.co.uk/fiultra/
"wonderfully well written and reasoned information
for the tube audio constructor"
John Broskie TubeCAD & GlassWare
"an unbelievably comprehensive web site
containing vital gems of wisdom"
Stuart Perry Hi-Fi News & Record Review

[mick]

On Sat, 15 Sep 2007 08:09:32 -0700, Andre Jute wrote:

snip

We should get a formula down for determining the exact voltage that will
do the trick. Patrick, this is a job for the resident guru.



Hi Andre!
It shouldn't need a formula. RMS has the same heating effect as the
equivalent DC voltage, so 6.3vDC works just as well as 6.3vAC.

To extend heater life the most important thing is to restrict the initial
inrush current (like any light bulb). Series resistors lose too many
volts though. You could use a series resistor or diodes to drop a volt or
so and short it out with a relay after a time delay, I suppose. This
could be useful with SS rectification, starting the heaters on low power
(but hot) then switching to full when HT is applied. Saves the electric
bill a bit too. :-)

Semiconductor inrush current-limiters should help on the primary of the
mains tranny but may have too high a hot resistance for use directly on
the heaters themselves.

I really can't understand why indirectly-heated valve heater life should
be shorter on DC supplies, providing that there aren't too many volts
between the heater and cathode. The heater doesn't emit any quantity of
electrons - it isn't designed to, the material is all wrong. If anything,
I would suspect that AC would actually shorten the life due to vibration
within local magnetic fields. I also suspect that all this is pseudo-
scientific speculation. ;-)

We have to remember that the valve manufacturers designed the heaters to
be able to handle switch-on surges to some extent.

--
Mick (Working in a M$-free zone!)
Web: http://www.nascom.info http://mixpix.batcave.net

[Iain Churches[_2_]]

"Andre Jute" wrote in message
ups.com...
On Sep 14, 11:36 pm, "West" wrote:
AC filaments mean less components (diodes, caps) and from what I read way
back are superior for certain tubes. Can DC can cause uneven cathode
stripping or filament wear? Is this true? If so, how can we protect our
DC
wired tubes? Are there certain tubes that fare better with AC for their
filaments? I know this topic has been beaten to death before, but please
indulge me, if you will. Thanks.

Hello Andre

Google the Sound List, Westley. This is the list of experts often
referred to as the Joelist.

Can you point us to this list, please?

Personally, while I
agree with everything Iain reports his friend as saying, I always show
the filament lift in my designs but often leave it out of what I build
for myself; nor can I remember even swapping fil ends;

In some circuits, like cascodes, and mu followers where the top
cathode is at say half the B+, the allowable heater to cathode voltage
will be exceed if the filament lift is not applied. We call it elevated
heaters, or heater biasing in this part of the globe. It may be known
by some other term in the US. John Byrns will soon shout at me if
it is:-))

yet I have
never seen a tube ruined by it. We can be too careful. Nontheless, a
necessary subject, because we can also take too much for granted.

Hmm. I recently went through a very large box of B9A "pulls"
The most common fault was open circuit heater followed by
cathode to heater short

We should get a formula down for determining the exact voltage that
will do the trick. Patrick, this is a job for the resident guru.

I would be interested in this also, and in addition a current regulator
at say 1.2A

Andre, West. It's nice to be talking tube audio with you again.

Best regards to each
Iain

[John Byrns]

In article i,
"Iain Churches" wrote:

In some circuits, like cascodes, and mu followers where the top
cathode is at say half the B+, the allowable heater to cathode voltage
will be exceed if the filament lift is not applied. We call it elevated
heaters, or heater biasing in this part of the globe. It may be known
by some other term in the US. John Byrns will soon shout at me if
it is:-))

I reserve my shouting for those few who really deserve it, what reason
would I have for shouting at you Iain?

I don't even know what we call it here in this part of the world (US),
IIRC I have mostly always referred to it as "biased heaters", or
sometimes as "elevated heaters".


Regards,

John Byrns

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

[Andre Jute]

On Sep 15, 9:11 am, mick wrote:
On Sat, 15 Sep 2007 08:09:32 -0700, Andre Jute wrote:

snip



We should get a formula down for determining the exact voltage that will
do the trick. Patrick, this is a job for the resident guru.

Hi Andre!
It shouldn't need a formula. RMS has the same heating effect as the
equivalent DC voltage, so 6.3vDC works just as well as 6.3vAC.

You're overcomplicating things, Mick. I just meant a formula to
calculate how high the voltage must be for say an SRPP.

To extend heater life the most important thing is to restrict the initial
inrush current (like any light bulb). Series resistors lose too many
volts though. You could use a series resistor or diodes to drop a volt or
so and short it out with a relay after a time delay, I suppose. This
could be useful with SS rectification, starting the heaters on low power
(but hot) then switching to full when HT is applied. Saves the electric
bill a bit too. :-)

Semiconductor inrush current-limiters should help on the primary of the
mains tranny but may have too high a hot resistance for use directly on
the heaters themselves.

I really can't understand why indirectly-heated valve heater life should
be shorter on DC supplies, providing that there aren't too many volts
between the heater and cathode. The heater doesn't emit any quantity of
electrons - it isn't designed to, the material is all wrong. If anything,
I would suspect that AC would actually shorten the life due to vibration
within local magnetic fields. I also suspect that all this is pseudo-
scientific speculation. ;-)

We have to remember that the valve manufacturers designed the heaters to
be able to handle switch-on surges to some extent.

Compared to modern lightbulbs, which are in many instances cheap ****
despite the "LIFETIME GUARANTEE" printed on the box, thermionic tubes
are industrial strength artifacts.

--
Mick (Working in a M$-free zone!)
Web:http://www.nascom.infohttp://mixpix.batcave.net

Great to see you back where you belong, Mick.

Andre Jute
"You don't need global feedback to build a good-sounding amplifier."
-- Henry Pasternack

[Andre Jute]

On Sep 15, 9:18 am, "Iain Churches" wrote:
"Andre Jute" wrote in message

ups.com... On Sep 14, 11:36 pm, "West" wrote:
AC filaments mean less components (diodes, caps) and from what I read way
back are superior for certain tubes. Can DC can cause uneven cathode
stripping or filament wear? Is this true? If so, how can we protect our
DC
wired tubes? Are there certain tubes that fare better with AC for their
filaments? I know this topic has been beaten to death before, but please
indulge me, if you will. Thanks.

Hello Andre

Google the Sound List, Westley. This is the list of experts often
referred to as the Joelist.

Can you point us to this list, please?

The archive is he
http://www.nmr.mgh.harvard.edu/~reese/joetest/
and this is a page from it
http://www.nmr.mgh.harvard.edu/~rees...les_v01_j.html

Copy what you want as I don't imagine that archive will be there
forever.


Personally, while I
agree with everything Iain reports his friend as saying, I always show
the filament lift in my designs but often leave it out of what I build
for myself; nor can I remember even swapping fil ends;

In some circuits, like cascodes, and mu followers where the top
cathode is at say half the B+, the allowable heater to cathode voltage
will be exceed if the filament lift is not applied. We call it elevated
heaters, or heater biasing in this part of the globe. It may be known
by some other term in the US. John Byrns will soon shout at me if
it is:-))

yet I have
never seen a tube ruined by it. We can be too careful. Nontheless, a
necessary subject, because we can also take too much for granted.

Hmm. I recently went through a very large box of B9A "pulls"
The most common fault was open circuit heater followed by
cathode to heater short

We should get a formula down for determining the exact voltage that
will do the trick. Patrick, this is a job for the resident guru.

I would be interested in this also, and in addition a current regulator
at say 1.2A

Andre, West. It's nice to be talking tube audio with you again.

Yes, isn't it a breath of fresh air.

Best regards to each
Iain

Keep the thermionic faith, bro'

Andre Jute
Visit Jute on Amps at http://members.lycos.co.uk/fiultra/
"wonderfully well written and reasoned information
for the tube audio constructor"
John Broskie TubeCAD & GlassWare
"an unbelievably comprehensive web site
containing vital gems of wisdom"
Stuart Perry Hi-Fi News & Record Review

[Iain Churches[_2_]]

"John Byrns" wrote in message
...
In article i,
"Iain Churches" wrote:

In some circuits, like cascodes, and mu followers where the top
cathode is at say half the B+, the allowable heater to cathode voltage
will be exceed if the filament lift is not applied. We call it elevated
heaters, or heater biasing in this part of the globe. It may be known
by some other term in the US. John Byrns will soon shout at me if
it is:-))

I reserve my shouting for those few who really deserve it, what reason
would I have for shouting at you Iain?

I added a smiley John.

I don't even know what we call it here in this part of the world (US),
IIRC I have mostly always referred to it as "biased heaters", or
sometimes as "elevated heaters".

Then we do use the same term
Do you use biased heaters in your amps?

Regards
Iain

[Chris Hornbeck]

On Sat, 15 Sep 2007 09:51:50 +0300, "Iain Churches"
wrote:

He states that heaters run from AC have a longer working life,
and that DC heaters always fail at the same end, (Would
that be the positive end?)

This is true anecdotally but I've never read a really
convincing explanation. Of course, not all true
explanations could be explained down to my level.

There's plenty of history, FWIW.

And in very expensive and high power transmitting
tube circuits, a filament polarity switch is done
automatically.

He has also told me that heaters should be
biased above the cathode (for which you need a floating heater supply)

There are two separate unrelated issues here. One is the
insulation between the very hot and uninsulated filament
and the hot and uninsulated cathode (all sorta/kinda true).

The other issue is the circuit noise immunity from
heater voltage. Leakage here is minimized by keeping
the cathode voltage below heater voltage. (Otherwise
it gets grand delusions of being an anode and drawing
electrons to itself. The little bitch.)

In a discussion about heater regulators, he was of the opinion that it is
the *current* which needs to be regulated and not the voltage, as a supply
within 5% tolerance is quite acceptable. The cold heater draws several
times its nominal current.

Start-up and normal run should properly be considered as
two separate gigs. Start-up current limiting might be
considered lily-gilding these days, but is appropriate
in best-effort designs.

Run operation should be at rated voltage within +/- 5%
for power stages, +0/ -10% for low work-factor stages.

Rough numbers, but not too far off the mark.

Thanks, as always,

Chris Hornbeck

[mick]

On Sat, 15 Sep 2007 16:21:33 -0700, Andre Jute wrote:

snip

You're overcomplicating things, Mick. I just meant a formula to
calculate how high the voltage must be for say an SRPP.


Sorry - misunderstood you! Shirley that's just the same as the top
valve's cathode, isn't it? I would guess that a floating supply would be
ideal, but any small voltage difference shouldn't matter. As I said, the
heater simply doesn't have freely available electrons so, providing that
we are talking about indirect heating here, you aren't going to peel away
the surface of the heater if it is at a similar potential to the cathode.
Any electron movement between the heater & cathode is irrelevant from an
operational point of view anyway.

snip

We have to remember that the valve manufacturers designed the heaters
to be able to handle switch-on surges to some extent.

Compared to modern lightbulbs, which are in many instances cheap ****
despite the "LIFETIME GUARANTEE" printed on the box, thermionic tubes
are industrial strength artifacts.


AMEN! ;-)

As an example of extreme cost cutting (this has to be a loss-leader) our
local Morrissons has been selling "high efficiency" bulbs for 39p each...

snip

Great to see you back where you belong, Mick.


:-)

--
Mick (Working in a M$-free zone!)
Web: http://www.nascom.info http://mixpix.batcave.net

[Ian Iveson]

flipper wrote:

The other issue is the circuit noise immunity from
heater voltage. Leakage here is minimized by keeping
the cathode voltage below heater voltage. (Otherwise
it gets grand delusions of being an anode and drawing
electrons to itself. The little bitch.)

I've wondered why that matters. I mean, in one case
there's flow out
and, in the other, flow in.

Possibly not. It's a diode, maybe.

Ok, so why is one 'quieter' than the other?

If the diode is biased greater than signal + noise voltage,
then it should block both?

Ian

[John Byrns]

In article ,
"Iain Churches" wrote:

"John Byrns" wrote in message
...
In article i,
"Iain Churches" wrote:

In some circuits, like cascodes, and mu followers where the top
cathode is at say half the B+, the allowable heater to cathode voltage
will be exceed if the filament lift is not applied. We call it elevated
heaters, or heater biasing in this part of the globe. It may be known
by some other term in the US. John Byrns will soon shout at me if
it is:-))

I reserve my shouting for those few who really deserve it, what reason
would I have for shouting at you Iain?

I added a smiley John.

I don't even know what we call it here in this part of the world (US),
IIRC I have mostly always referred to it as "biased heaters", or
sometimes as "elevated heaters".

Then we do use the same term
Do you use biased heaters in your amps?

No, biased heaters are only required in weird audiophile circuit designs.


Regards,

John Byrns

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

[Ian Iveson]

Chris Hornbeck wrote

There are two separate unrelated issues here. One is the
insulation between the very hot and uninsulated filament
and the hot and uninsulated cathode (all sorta/kinda
true).

The other issue is the circuit noise immunity from
heater voltage. Leakage here is minimized by keeping
the cathode voltage below heater voltage. (Otherwise
it gets grand delusions of being an anode and drawing
electrons to itself. The little bitch.)

They may not be unrelated.

I've never used whatever I learned about corrosion, so I'm a
little rusty, but it may be that if the heater becomes a
cathode, then it can suffer from cathode poisoning. Perhaps
thinning of the most cathode-like end of the heater wire
results in a hot spot.

Corrosion in general is an interesting subject, and a read
around should provide some insight.

cheers, Ian

[John Byrns]

In article ,
Chris Hornbeck wrote:

On Sat, 15 Sep 2007 09:51:50 +0300, "Iain Churches"
wrote:

He states that heaters run from AC have a longer working life,
and that DC heaters always fail at the same end, (Would
that be the positive end?)

This is true anecdotally but I've never read a really
convincing explanation. Of course, not all true
explanations could be explained down to my level.

There's plenty of history, FWIW.

And in very expensive and high power transmitting
tube circuits, a filament polarity switch is done
automatically.

How many post early 30s high power transmitting tube circuits actually
used a DC filament supply? Can you point to examples? Also can you
provide an example of a high power transmitting tube circuit that had an
automatic filament polarity switch? All of which also raises the
question of what "high power" is? Sorry for all the questions but
enquiring minds want to know


Regards,

John Byrns

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

[Iain Churches[_2_]]

"John Byrns" wrote in message
...
In article ,
"Iain Churches" wrote:

"John Byrns" wrote in message
...
In article i,
"Iain Churches" wrote:

In some circuits, like cascodes, and mu followers where the top
cathode is at say half the B+, the allowable heater to cathode voltage
will be exceed if the filament lift is not applied. We call it
elevated
heaters, or heater biasing in this part of the globe. It may be known
by some other term in the US. John Byrns will soon shout at me if
it is:-))

I reserve my shouting for those few who really deserve it, what reason
would I have for shouting at you Iain?

I added a smiley John.

I don't even know what we call it here in this part of the world (US),
IIRC I have mostly always referred to it as "biased heaters", or
sometimes as "elevated heaters".

Then we do use the same term
Do you use biased heaters in your amps?

No, biased heaters are only required in weird audiophile circuit designs.

Does SRPP come within that classification?

Iain

[John Byrns]

In article i,
"Iain Churches" wrote:

"John Byrns" wrote in message
...
In article ,
"Iain Churches" wrote:

"John Byrns" wrote in message
...
In article i,
"Iain Churches" wrote:

In some circuits, like cascodes, and mu followers where the top
cathode is at say half the B+, the allowable heater to cathode voltage
will be exceed if the filament lift is not applied. We call it
elevated
heaters, or heater biasing in this part of the globe. It may be known
by some other term in the US. John Byrns will soon shout at me if
it is:-))

I reserve my shouting for those few who really deserve it, what reason
would I have for shouting at you Iain?

I added a smiley John.

I don't even know what we call it here in this part of the world (US),
IIRC I have mostly always referred to it as "biased heaters", or
sometimes as "elevated heaters".

Then we do use the same term
Do you use biased heaters in your amps?

No, biased heaters are only required in weird audiophile circuit designs.

Does SRPP come within that classification?

Yes, obviously, on both counts.


Regards,

John Byrns

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

[Iain Churches[_2_]]

"mick" wrote in message
. uk...
On Sat, 15 Sep 2007 16:21:33 -0700, Andre Jute wrote:

snip

You're overcomplicating things, Mick. I just meant a formula to
calculate how high the voltage must be for say an SRPP.


Sorry - misunderstood you! Shirley that's just the same as the top
valve's cathode, isn't it? I would guess that a floating supply would be
ideal, but any small voltage difference shouldn't matter. As I said, the
heater simply doesn't have freely available electrons so, providing that
we are talking about indirect heating here, you aren't going to peel away
the surface of the heater if it is at a similar potential to the cathode.
Any electron movement between the heater & cathode is irrelevant from an
operational point of view anyway.

My own experience is that SRPP circuits with DC heaters
between 6V3 and ground seem to go noisy after some
months of use. Replacing the tubes (same brand) and
floating the heater at say 90V seems to solve the problem.
Who knows why?

Iain

[Iain Churches[_2_]]

"Bret Ludwig" wrote in message
oups.com...

Cascodes and mu followers and such things may require or at least
benefit from the set having multiple heater supplies, differently
biased. Where they're the same tube that's of course impossible.


Indeed separate supplies would be the right way to do this.

As a pretty good compormise, designers seem often to
set the bias at 90V which is within both the maximum
positive differential and the maximum negative differential
of the top and bottom cathode respectively.

Regards to all
Iain

[Iain Churches[_2_]]

"John Byrns" wrote in message
...
In article i,
"Iain Churches" wrote:

"John Byrns" wrote in message
...
In article ,
"Iain Churches" wrote:

"John Byrns" wrote in message
...
In article i,
"Iain Churches" wrote:

In some circuits, like cascodes, and mu followers where the top
cathode is at say half the B+, the allowable heater to cathode
voltage
will be exceed if the filament lift is not applied. We call it
elevated
heaters, or heater biasing in this part of the globe. It may be
known
by some other term in the US. John Byrns will soon shout at me if
it is:-))

I reserve my shouting for those few who really deserve it, what
reason
would I have for shouting at you Iain?

I added a smiley John.

I don't even know what we call it here in this part of the world
(US),
IIRC I have mostly always referred to it as "biased heaters", or
sometimes as "elevated heaters".

Then we do use the same term
Do you use biased heaters in your amps?

No, biased heaters are only required in weird audiophile circuit
designs.

Does SRPP come within that classification?

Yes, obviously, on both counts.

Hmm. That seems to rule out a lot of interesting topologies.
So what would you regard as "not" being weird audiophile
circuit designs?

Regards
Iain

[mick]

On Sun, 16 Sep 2007 22:35:05 +0300, Iain Churches wrote:

"John Byrns" wrote in message
...
In article i, "Iain
Churches" wrote:

"John Byrns" wrote in message
...
In article ,
"Iain Churches" wrote:

"John Byrns" wrote in message
...
In article i,
"Iain Churches" wrote:

In some circuits, like cascodes, and mu followers where the top
cathode is at say half the B+, the allowable heater to cathode
voltage
will be exceed if the filament lift is not applied. We call it
elevated
heaters, or heater biasing in this part of the globe. It may be
known
by some other term in the US. John Byrns will soon shout at me
if it is:-))

I reserve my shouting for those few who really deserve it, what
reason
would I have for shouting at you Iain?

I added a smiley John.

I don't even know what we call it here in this part of the world
(US),
IIRC I have mostly always referred to it as "biased heaters", or
sometimes as "elevated heaters".

Then we do use the same term
Do you use biased heaters in your amps?

No, biased heaters are only required in weird audiophile circuit
designs.

Does SRPP come within that classification?

Yes, obviously, on both counts.

Hmm. That seems to rule out a lot of interesting topologies. So what
would you regard as "not" being weird audiophile circuit designs?

Regards
Iain


common emitter...


gets coat

--
Mick (Working in a M$-free zone!)
Web: http://www.nascom.info http://mixpix.batcave.net

[John Byrns]

In article i,
"Iain Churches" wrote:

"John Byrns" wrote in message
...
In article i,
"Iain Churches" wrote:

"John Byrns" wrote in message
...
In article ,
"Iain Churches" wrote:

"John Byrns" wrote in message
...
In article i,
"Iain Churches" wrote:

In some circuits, like cascodes, and mu followers where the top
cathode is at say half the B+, the allowable heater to cathode
voltage
will be exceed if the filament lift is not applied. We call it
elevated
heaters, or heater biasing in this part of the globe. It may be
known
by some other term in the US. John Byrns will soon shout at me if
it is:-))

I reserve my shouting for those few who really deserve it, what
reason
would I have for shouting at you Iain?

I added a smiley John.

I don't even know what we call it here in this part of the world
(US),
IIRC I have mostly always referred to it as "biased heaters", or
sometimes as "elevated heaters".

Then we do use the same term
Do you use biased heaters in your amps?

No, biased heaters are only required in weird audiophile circuit
designs.

Does SRPP come within that classification?

Yes, obviously, on both counts.

Hmm. That seems to rule out a lot of interesting topologies.
So what would you regard as "not" being weird audiophile
circuit designs?

Audiophile circuit designs are always overly complex, I first became
aware of this syndrome with the "cross coupled" phase inverter, I think
it may also have another name, that was popular with audiophiles in the
1950s. Modern day audiophiles carry on the tradition with "SRPP" and
the like.

Circuits that are "not" audiophile emphasize simplicity and reliability
over needless complexity.


Regards,

John Byrns

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

[Andre Jute]

On Sep 16, 4:12 pm, John Byrns wrote:
In article i,
"Iain Churches" wrote:
Do you use biased heaters in your amps?

No, biased heaters are only required in weird audiophile circuit
designs.

Does SRPP come within that classification?

Yes, obviously, on both counts.

Hmm. That seems to rule out a lot of interesting topologies.
So what would you regard as "not" being weird audiophile
circuit designs?

Audiophile circuit designs are always overly complex, I first became
aware of this syndrome with the "cross coupled" phase inverter, I think
it may also have another name, that was popular with audiophiles in the
1950s.

Modern day audiophiles carry on the tradition with "SRPP" and
the like.

Circuits that are "not" audiophile emphasize simplicity and reliability
over needless complexity.

Audiophile circuits too, if you know the right audiophiles.

The entire SET audiophile movement is, at least theoretically,
dedicated to simplicity, purity and, admittedly a little way down the
line, reliability. The ultrafidelista build amps so simple there is
almost nothing to go wrong, and their amps are by default extremely
reliable, among other reasons because the unltrafidelista, if they're
true to their principles, run their tubes very conservatively, only on
the flattest part of the transfer curve, away from the extremes.

Regards,

John Byrns

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

Andre Jute
Visit Jute on Amps at http://members.lycos.co.uk/fiultra/
"wonderfully well written and reasoned information
for the tube audio constructor"
John Broskie TubeCAD & GlassWare
"an unbelievably comprehensive web site
containing vital gems of wisdom"
Stuart Perry Hi-Fi News & Record Review

[Chris Hornbeck]

On Sun, 16 Sep 2007 14:20:44 GMT, John Byrns
wrote:

And in very expensive and high power transmitting
tube circuits, a filament polarity switch is done
automatically.

How many post early 30s high power transmitting tube circuits actually
used a DC filament supply? Can you point to examples? Also can you
provide an example of a high power transmitting tube circuit that had an
automatic filament polarity switch? All of which also raises the
question of what "high power" is? Sorry for all the questions but
enquiring minds want to know

I heard about this from an old friend who runs the local
CBS affiliate TV transmitter. These use a final tube the
size of a dinner plate, and have a lot of what, to us,
is pretty amazing tube protection. Lightning protection?
You bet, although I don't know the details.

Next time I see him, I'll get more info on the polarity
switching.

Thanks, as always,

Chris Hornbeck

[John Byrns]

In article ,
Chris Hornbeck wrote:

On Sun, 16 Sep 2007 14:20:44 GMT, John Byrns
wrote:

And in very expensive and high power transmitting
tube circuits, a filament polarity switch is done
automatically.

How many post early 30s high power transmitting tube circuits actually
used a DC filament supply? Can you point to examples? Also can you
provide an example of a high power transmitting tube circuit that had an
automatic filament polarity switch? All of which also raises the
question of what "high power" is? Sorry for all the questions but
enquiring minds want to know

I heard about this from an old friend who runs the local
CBS affiliate TV transmitter. These use a final tube the
size of a dinner plate, and have a lot of what, to us,
is pretty amazing tube protection. Lightning protection?
You bet, although I don't know the details.

Next time I see him, I'll get more info on the polarity
switching.

I suspect that when you talk to your friend you will find that the CBS
affiliate TV transmitter uses a polarity reversal system that flips the
filament polarity of the dinner plate sized tube 120 times a second by
what is really a very simple mechanism.

When I was in college many years ago I worked for an AM-FM-TV station.
The final tubes in the TV transmitter were quite a bit larger than a
dinner plate, their scale was more that of man sized, about my size
IIRC. The cathode was heated by a 5,000 volt supply IIRC. The cathode
of the main amplifier portion of the tube was also the anode of a vacuum
diode which had the aforementioned 5,000 volt potential applied between
its cathode and the anode/cathode. The voltage on the filament of the
diode was reversed automatically 120 times a second. This vacuum diode
heater system was called a "bombarder".


Regards,

John Byrns

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

[Chris Hornbeck]

On Mon, 17 Sep 2007 03:11:25 GMT, John Byrns
wrote:

Next time I see him, I'll get more info on the polarity
switching.

I suspect that when you talk to your friend you will find that the CBS
affiliate TV transmitter uses a polarity reversal system that flips the
filament polarity of the dinner plate sized tube 120 times a second by
what is really a very simple mechanism.

Could turn out to be true, but my memory is of a simple relay
arrangement that swapped polarity only at (rare) turn-on's.

This is from memory, so I will need to check with Jay to
have any real info for ya.


When I was in college many years ago I worked for an AM-FM-TV station.
The final tubes in the TV transmitter were quite a bit larger than a
dinner plate, their scale was more that of man sized, about my size
IIRC. The cathode was heated by a 5,000 volt supply IIRC. The cathode
of the main amplifier portion of the tube was also the anode of a vacuum
diode which had the aforementioned 5,000 volt potential applied between
its cathode and the anode/cathode. The voltage on the filament of the
diode was reversed automatically 120 times a second. This vacuum diode
heater system was called a "bombarder".

Cool beans. Never seen one personally. I did see a spare final tube
at the tower and it was about the size of a dinner plate, forced-air
cooled of course. Larger dimensions would be a problem for a VHF
transmitter; channel 11 is what? two channels of 6MHz each down from
216MHz, so 198-204MHz.

One strange thing that I remember is that the (linear) final also
amplifies the video carrier. You'd think they'd use a separate
class-C carrier final and mix 'em together, but nope. Sound
carrier was separate, IIRC.


Thanks, as always,

Chris Hornbeck

[Patrick Turner]

Andre Jute wrote:

On Sep 14, 11:36 pm, "West" wrote:
AC filaments mean less components (diodes, caps) and from what I read way
back are superior for certain tubes. Can DC can cause uneven cathode
stripping or filament wear? Is this true? If so, how can we protect our DC
wired tubes? Are there certain tubes that fare better with AC for their
filaments? I know this topic has been beaten to death before, but please
indulge me, if you will. Thanks.

west

Google the Sound List, Westley. This is the list of experts often
referred to as the Joelist. The archive is kept at Harvard. About ten
years ago, maybe 1996/7, there was a long learned discussion of the
subject, and it was decided AC fils plus a raised DC voltage (I seem
to remember 50-90V) plus occasional fil polarity swapping would do the
trick. No one had ever seen cathode stripping... Personally, while I
agree with everything Iain reports his friend as saying, I always show
the filament lift in my designs but often leave it out of what I build
for myself; nor can I remember even swapping fil ends; yet I have
never seen a tube ruined by it. We can be too careful. Nontheless, a
necessary subject, because we can also take too much for granted.

We should get a formula down for determining the exact voltage that
will do the trick. Patrick, this is a job for the resident guru.

I have used many IDH tubes with DC heaters for low noise.

I have never found any reliablity problems.

The benefits of low noise outweigh any disadvantages of shortened tube
life.

In CF or µ-follower circuits the heaters should always be within 90 pk
volts of the cathoded,
and if a CF cathode is typicallly at +150V, then having the low side of
12V heaters at +75V is pretty standard practice.
CJ has been doing it awhile, and a PV2 amp I have just re-wired
completely uses
regulated 12.3Vdc floating at +76Vdc.

Slow turn on of heaters is good practice, but usually where you have a
rectifier,
and place a small resistance in series between diodes and the reservoir
cap in the DC heater supply
you will effectively slow down the Vdc applied to the heaters in lieu of
having
the heaters powered from cold from ac windings.

Cathode stripping in home hi-fi amps is rarely a problem, and only could
be one where
the tubes are turned on very often while hot without bias and they
saturate momentarily.

The power supply does well to have the dc heater supply shunt regulated
so that in the event of a longer than usual power voltage surge the
heaters don't
cop too much which can damage heaters. Its unlikely, and if mains rose
50%
usually a mains fuse will blow fairly quick.

With DH triodes, the same could be said, except that the heater supply
must
always be at the cathode voltage. I'd done it with 300B amps without
problems,
and it ppl feel uneasy, then reverse the dc direction every so often,
or arrange a sequential relay circuit to do it automatically each time
the amp
is switched on and off.

Most DH cathodes and AC heaters are designed to produce a wanted ideal
temperature
for the emission, and RDH4 says design voltage should be about within
+/- 10%,
but its better probably to be slightly lower.
With preamp tubes, you never need the cathode to produce its full
ability
for current production, so running 6V instead of 6.3V is quite OK.

Chinese soldering irons
designed for 220V and sold in Oz last a month if you are lucky, and run
way too hot
because we have 250V here often. I have a steel box with some series R
and a switch
for cool, medium, and some control of temp
and the irons last 18mths plus, not bad for $20 worth of junk,
wheras Weller mains irons don't last that long despite their US design
legacy.
Chinese tubes can sometimes however have an appalling failure rate from
a variety of reasons, and try as they do,
they keep trying to sell crap way off spec and with worse made heaters
than were made in the good old days. But not all chinese tubes. Some
seem to last very well,
but maybe its luck, feng shoie....

I once put Russian, British and American preamp input and driver tubes
into a circuit
as replacements, and guess which tube developed a heater-cathode short
soon
after the amp went back to its owner?

The bloody American tube of course.

They always veto what the other tubes try to do.

Patrick Turner.





Andre Jute
Visit Jute on Amps at http://members.lycos.co.uk/fiultra/
"wonderfully well written and reasoned information
for the tube audio constructor"
John Broskie TubeCAD & GlassWare
"an unbelievably comprehensive web site
containing vital gems of wisdom"
Stuart Perry Hi-Fi News & Record Review

[John Byrns]

In article ,
Chris Hornbeck wrote:

On Mon, 17 Sep 2007 03:11:25 GMT, John Byrns
wrote:

Next time I see him, I'll get more info on the polarity
switching.

I suspect that when you talk to your friend you will find that the CBS
affiliate TV transmitter uses a polarity reversal system that flips the
filament polarity of the dinner plate sized tube 120 times a second by
what is really a very simple mechanism.

Could turn out to be true, but my memory is of a simple relay
arrangement that swapped polarity only at (rare) turn-on's.

This is from memory, so I will need to check with Jay to
have any real info for ya.

When I was in college many years ago I worked for an AM-FM-TV station.
The final tubes in the TV transmitter were quite a bit larger than a
dinner plate, their scale was more that of man sized, about my size
IIRC. The cathode was heated by a 5,000 volt supply IIRC. The cathode
of the main amplifier portion of the tube was also the anode of a vacuum
diode which had the aforementioned 5,000 volt potential applied between
its cathode and the anode/cathode. The voltage on the filament of the
diode was reversed automatically 120 times a second. This vacuum diode
heater system was called a "bombarder".

Cool beans. Never seen one personally. I did see a spare final tube
at the tower and it was about the size of a dinner plate, forced-air
cooled of course. Larger dimensions would be a problem for a VHF
transmitter; channel 11 is what? two channels of 6MHz each down from
216MHz, so 198-204MHz.

Yeah, the larger tubes were used at UHF frequencies.

One strange thing that I remember is that the (linear) final also
amplifies the video carrier. You'd think they'd use a separate
class-C carrier final and mix 'em together, but nope. Sound
carrier was separate, IIRC.

I'm not sure what you are trying to say above, the writing seems a bit
muddled and the meaning is unclear. In the way-back days of the
transmitter I was talking about, the vision and sound transmitters were
independent, with two completely independent amplifiers, one for the
vision signal and one for the sound. It was during the time I was
working at the station that the FCC changed their regulations and
reduced the level of the sound carrier, I remember we had to reduce the
FM sound carrier power from 50% of the peak vision power down to 20%. I
don't know for sure but I get the impression that with the reduced sound
carrier power many current transmitters are able to run the FM sound
carrier through the same linear amplifier used for the AM vision
carrier, but the old transmitters weren't designed to be able to do this.


Regards,

John Byrns

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

[Chris Hornbeck]

On Mon, 17 Sep 2007 07:53:20 -0500, John Byrns
wrote:

One strange thing that I remember is that the (linear) final also
amplifies the video carrier. You'd think they'd use a separate
class-C carrier final and mix 'em together, but nope. Sound
carrier was separate, IIRC.

I'm not sure what you are trying to say above, the writing seems a bit
muddled and the meaning is unclear.

Sorry. The video final operates in class AB, and is burdened by
linearly amplifying the carrier along with the upper sideband.
The carrier could easily be supplied by a class C final, and
the overall efficiency improved, but there were apparently
other considerations.

Thanks, as always,

Chris Hornbeck

[Ian Iveson]

Flipper said

If the diode is biased greater than signal + noise
voltage,
then it should block both?

If one is going to postulate a 'diode' then it seems
logical that the
thing made to emit electrons, I.E. the cathode, would be,
well, the
cathode, and the other object, I.E. the heater, would be
the anode.

It's a relative thing, see rambling below.

In that case, to 'block' flow you'd make the heater
(anode) negative
but the recommendation is to make it positive (cathode
voltage below
heater voltage). Which, if the above 'diode' description
is correct,
would induce current flow.

I figured that with the heater being the hottest thing in
there that
it would probably like to emit electrons too and, so,
current could
also flow the other way, with 'how much' left undefined.
Although, I
presume the heater emits less or else there'd be little
reason for the
cathode but, if so, that means the recommended heater bias
is in the
direction of maximizing electron flow to it.

I was just hoping someone could explain it to me.

But, trying to answer my own question, I suppose it could
be an
impedance thing. Like the cathode having so many 'spare'
electrons
that a modest amount going to the heater is
inconsequential (due to
the low impedance) but that the heater trying to emit more
into the
mix creates a problem because the cathode sure doesn't
want them (high
impedance) and, so, has more effect on the plate..


John Byrns wrote something a bit interesting:

"The final tubes in the TV transmitter were quite a bit
larger than a
dinner plate, their scale was more that of man sized, about
my size
IIRC. The cathode was heated by a 5,000 volt supply IIRC.
The cathode
of the main amplifier portion of the tube was also the anode
of a vacuum
diode which had the aforementioned 5,000 volt potential
applied between
its cathode and the anode/cathode. The voltage on the
filament of the
diode was reversed automatically 120 times a second. This
vacuum diode
heater system was called a "bombarder"."

I wonder if the diode John refers to here was a diode in the
sense we generally assume, ie that it would conduct only in
one direction?

I wonder what would happen if you put two heated cathode
assemblies in a valve and connected AC between them? Would
conduction take place in both directions, or neither? If one
was cooler than the other, or one was of a metal more
inclined to lose its electrons at the same temperature, how
would that effect the outcome? If you now introduced, behind
one of the 'cathodes', a negative electrode, so as to offset
the 'electron cloud' of the nearest 'cathode', how would
that affect conduction between the two cathode assemblies?

Maybe this would all be clearer if I could think in terms of
waves rather than particles. Sigh. I have some difficulty
imagining exactly how a cloud of electrons behaves, and
particularly what potential it appears to be at from the
point of view of an electrode.

It should be possible with one theory to explain all of the
maybe-myths of aberrant valve behaviour, e.g.:

* Heaters should not be left full on without HT present for
long periods of time for fear of damaging the cathode.

* HT should not be present when the heaters are not up to
temperature for fear of damaging the cathode.

* Heaters should be referenced to a voltage above the
cathode for best sound and lowest noise.

I've just been staring at my 6CH6 which have holes in so you
can see inside. It's a while since I did this in the dark, I
don't remember the blue glow being so bright. Anyway, heater
wires are very pale yellow as they enter the "insulation",
so they must be seriously white hot in there. The oxide
layer appears to be stuffed tightly to fill all the space
inside the orange-hot cathode. "Insulation", incidentally,
doesn't seem quite the right word...a vacuum is an insulator
too...but the oxide is a better heat-conductor. Does it also
need to be impervious, in the sense that there are no holes
big enough for an electron to fly through...to counter
thermionic emission?

White hot tungsten may be a stronger emitter of electrons
than red hot cathode material, in the sense that its emitted
electrons are of a higher average energy? Perhaps its little
cloud is consequently more negative than that of the
cathode, even when the cathode's cloud is being pushed by
the grid on the other side. How many volts would it take to
overcome such a bias? I guess it boils down to what kind of
voltage it takes to defeat the emission of electrons from a
particular material at a particular temperature.

Just musing obviously. There seems to be no shortage of
one-way, all or nothing kinds of interactions between
dissimilar materials and electricity: thermocouples,
batteries, electrolytic caps, semiconductors, etc.

Fortunately I don't have to worry in a practical way. These
valves have been running for years and don't seem to have
changed much, with heaters strung out over 25V in a
current-regulated string. They're cheap as chips, too.
People with expensive valves and/or more sensitive
speaker/room/ear combinations may have more cause to care.

Ian

[West]

"Chris Hornbeck" wrote in message
...
On Sat, 15 Sep 2007 09:51:50 +0300, "Iain Churches"
wrote:

He states that heaters run from AC have a longer working life,
and that DC heaters always fail at the same end, (Would
that be the positive end?)

This is true anecdotally but I've never read a really
convincing explanation. Of course, not all true
explanations could be explained down to my level.

There's plenty of history, FWIW.

And in very expensive and high power transmitting
tube circuits, a filament polarity switch is done
automatically.

How do they make an automatic polarity switch. I can not find anything on
this and tend to think it doesn't exist.

He has also told me that heaters should be
biased above the cathode (for which you need a floating heater supply)

Isn't this mostly for hum control?

There are two separate unrelated issues here. One is the
insulation between the very hot and uninsulated filament
and the hot and uninsulated cathode (all sorta/kinda true).

The other issue is the circuit noise immunity from
heater voltage. Leakage here is minimized by keeping
the cathode voltage below heater voltage. (Otherwise
it gets grand delusions of being an anode and drawing
electrons to itself. The little bitch.)

The main issue is which ac or dc is better for tube longevity and why.

In a discussion about heater regulators, he was of the opinion that it is
the *current* which needs to be regulated and not the voltage, as a
supply
within 5% tolerance is quite acceptable. The cold heater draws several
times its nominal current.

I'm not sure how you can current regulate a device that draws about 7X more
curent when cold.
It's like trying to regulate a thermister.

Start-up and normal run should properly be considered as
two separate gigs. Start-up current limiting might be
considered lily-gilding these days, but is appropriate
in best-effort designs.

Run operation should be at rated voltage within +/- 5%
for power stages, +0/ -10% for low work-factor stages.

Rough numbers, but not too far off the mark.

Thanks, as always,

Chris Hornbeck

I took you to task on several of your issues, Chris. However, switching
polarities can be the most difficult issue. I don't see how in very
expensive and high transmitting equippment, as you stated, have an automatic
polarity switch. Easier said than done, unless you have some first hand
knowledge, please share. Thank you.

west

[West]

"John Byrns" wrote in message
...
In article ,
Chris Hornbeck wrote:

On Sat, 15 Sep 2007 09:51:50 +0300, "Iain Churches"
wrote:

He states that heaters run from AC have a longer working life,
and that DC heaters always fail at the same end, (Would
that be the positive end?)

This is true anecdotally but I've never read a really
convincing explanation. Of course, not all true
explanations could be explained down to my level.

There's plenty of history, FWIW.

And in very expensive and high power transmitting
tube circuits, a filament polarity switch is done
automatically.

How many post early 30s high power transmitting tube circuits actually
used a DC filament supply? Can you point to examples? Also can you
provide an example of a high power transmitting tube circuit that had an
automatic filament polarity switch? All of which also raises the
question of what "high power" is? Sorry for all the questions but
enquiring minds want to know

I would also like to know as I mentioed in an above post. I have never seen
an automatic polarity switch. The only way you can do it is manually AFAIK.

west


Regards,

John Byrns

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

[West]

"John Byrns" wrote in message
...
In article ,
Chris Hornbeck wrote:

On Sun, 16 Sep 2007 14:20:44 GMT, John Byrns
wrote:

And in very expensive and high power transmitting
tube circuits, a filament polarity switch is done
automatically.

How many post early 30s high power transmitting tube circuits actually
used a DC filament supply? Can you point to examples? Also can you
provide an example of a high power transmitting tube circuit that had
an
automatic filament polarity switch? All of which also raises the
question of what "high power" is? Sorry for all the questions but
enquiring minds want to know

I heard about this from an old friend who runs the local
CBS affiliate TV transmitter. These use a final tube the
size of a dinner plate, and have a lot of what, to us,
is pretty amazing tube protection. Lightning protection?
You bet, although I don't know the details.

Next time I see him, I'll get more info on the polarity
switching.

I suspect that when you talk to your friend you will find that the CBS
affiliate TV transmitter uses a polarity reversal system that flips the
filament polarity of the dinner plate sized tube 120 times a second by
what is really a very simple mechanism.

Wouldn't that defeat the characteristics of the DC itself. You might as well
leave it AC filament at 60X/Sec. What is your so called simple mechanism?
Sorry but I can't bight without some elaboration.

When I was in college many years ago I worked for an AM-FM-TV station.
The final tubes in the TV transmitter were quite a bit larger than a
dinner plate, their scale was more that of man sized, about my size
IIRC. The cathode was heated by a 5,000 volt supply IIRC. The cathode
of the main amplifier portion of the tube was also the anode of a vacuum
diode which had the aforementioned 5,000 volt potential applied between
its cathode and the anode/cathode. The voltage on the filament of the
diode was reversed automatically 120 times a second. This vacuum diode
heater system was called a "bombarder".

How did they do this and why?

west

Regards,

John Byrns

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

[West]

"Chris Hornbeck" wrote in message
...
On Mon, 17 Sep 2007 03:11:25 GMT, John Byrns
wrote:

Next time I see him, I'll get more info on the polarity
switching.

I suspect that when you talk to your friend you will find that the CBS
affiliate TV transmitter uses a polarity reversal system that flips the
filament polarity of the dinner plate sized tube 120 times a second by
what is really a very simple mechanism.

Could turn out to be true, but my memory is of a simple relay
arrangement that swapped polarity only at (rare) turn-on's.

This willnot work unless you have a battery for memory or a large impulse
relay.
This is from memory, so I will need to check with Jay to
have any real info for ya.


When I was in college many years ago I worked for an AM-FM-TV station.
The final tubes in the TV transmitter were quite a bit larger than a
dinner plate, their scale was more that of man sized, about my size
IIRC. The cathode was heated by a 5,000 volt supply IIRC. The cathode
of the main amplifier portion of the tube was also the anode of a vacuum
diode which had the aforementioned 5,000 volt potential applied between
its cathode and the anode/cathode. The voltage on the filament of the
diode was reversed automatically 120 times a second. This vacuum diode
heater system was called a "bombarder".

Cool beans. Never seen one personally. I did see a spare final tube
at the tower and it was about the size of a dinner plate, forced-air
cooled of course. Larger dimensions would be a problem for a VHF
transmitter; channel 11 is what? two channels of 6MHz each down from
216MHz, so 198-204MHz.

One strange thing that I remember is that the (linear) final also
amplifies the video carrier. You'd think they'd use a separate
class-C carrier final and mix 'em together, but nope. Sound
carrier was separate, IIRC.


Thanks, as always,

Chris Hornbeck

[West]

"Patrick Turner" wrote in message
...


Andre Jute wrote:

On Sep 14, 11:36 pm, "West" wrote:
AC filaments mean less components (diodes, caps) and from what I read
way
back are superior for certain tubes. Can DC can cause uneven cathode
stripping or filament wear? Is this true? If so, how can we protect
our DC
wired tubes? Are there certain tubes that fare better with AC for
their
filaments? I know this topic has been beaten to death before, but
please
indulge me, if you will. Thanks.

west

Google the Sound List, Westley. This is the list of experts often
referred to as the Joelist. The archive is kept at Harvard. About ten
years ago, maybe 1996/7, there was a long learned discussion of the
subject, and it was decided AC fils plus a raised DC voltage (I seem
to remember 50-90V) plus occasional fil polarity swapping would do the
trick. No one had ever seen cathode stripping... Personally, while I
agree with everything Iain reports his friend as saying, I always show
the filament lift in my designs but often leave it out of what I build
for myself; nor can I remember even swapping fil ends; yet I have
never seen a tube ruined by it. We can be too careful. Nontheless, a
necessary subject, because we can also take too much for granted.

We should get a formula down for determining the exact voltage that
will do the trick. Patrick, this is a job for the resident guru.

I have used many IDH tubes with DC heaters for low noise.

I have never found any reliablity problems.

The benefits of low noise outweigh any disadvantages of shortened tube
life.

In CF or µ-follower circuits the heaters should always be within 90 pk
volts of the cathoded,
and if a CF cathode is typicallly at +150V, then having the low side of
12V heaters at +75V is pretty standard practice.
CJ has been doing it awhile, and a PV2 amp I have just re-wired
completely uses
regulated 12.3Vdc floating at +76Vdc.

Slow turn on of heaters is good practice, but usually where you have a
rectifier,
and place a small resistance in series between diodes and the reservoir
cap in the DC heater supply
you will effectively slow down the Vdc applied to the heaters in lieu of
having
the heaters powered from cold from ac windings.

Cathode stripping in home hi-fi amps is rarely a problem, and only could
be one where
the tubes are turned on very often while hot without bias and they
saturate momentarily.

The power supply does well to have the dc heater supply shunt regulated
so that in the event of a longer than usual power voltage surge the
heaters don't
cop too much which can damage heaters. Its unlikely, and if mains rose
50%
usually a mains fuse will blow fairly quick.

With DH triodes, the same could be said, except that the heater supply
must
always be at the cathode voltage. I'd done it with 300B amps without
problems,
and it ppl feel uneasy, then reverse the dc direction every so often,
or arrange a sequential relay circuit to do it automatically each time
the amp
is switched on and off.

This is easier said than done Patrick. Care to expand on what type of relay
& how it's hooked up?

west

Most DH cathodes and AC heaters are designed to produce a wanted ideal
temperature
for the emission, and RDH4 says design voltage should be about within
+/- 10%,
but its better probably to be slightly lower.
With preamp tubes, you never need the cathode to produce its full
ability
for current production, so running 6V instead of 6.3V is quite OK.

Chinese soldering irons
designed for 220V and sold in Oz last a month if you are lucky, and run
way too hot
because we have 250V here often. I have a steel box with some series R
and a switch
for cool, medium, and some control of temp
and the irons last 18mths plus, not bad for $20 worth of junk,
wheras Weller mains irons don't last that long despite their US design
legacy.
Chinese tubes can sometimes however have an appalling failure rate from
a variety of reasons, and try as they do,
they keep trying to sell crap way off spec and with worse made heaters
than were made in the good old days. But not all chinese tubes. Some
seem to last very well,
but maybe its luck, feng shoie....

I once put Russian, British and American preamp input and driver tubes
into a circuit
as replacements, and guess which tube developed a heater-cathode short
soon
after the amp went back to its owner?

The bloody American tube of course.

They always veto what the other tubes try to do.

Patrick Turner.





Andre Jute
Visit Jute on Amps at http://members.lycos.co.uk/fiultra/
"wonderfully well written and reasoned information
for the tube audio constructor"
John Broskie TubeCAD & GlassWare
"an unbelievably comprehensive web site
containing vital gems of wisdom"
Stuart Perry Hi-Fi News & Record Review

[Patrick Turner]

With DH triodes, the same could be said, except that the heater supply
must
always be at the cathode voltage. I'd done it with 300B amps without
problems,
and it ppl feel uneasy, then reverse the dc direction every so often,
or arrange a sequential relay circuit to do it automatically each time
the amp
is switched on and off.

This is easier said than done Patrick. Care to expand on what type of relay
& how it's hooked up?

It should be easy with a memory chip and 12V relay.

But a press button switch like that used for a bedside lamp, a press
gives on, another gives off, and so on, and operated by a solenoid that
gets a pulse at turn on should also
work OK, but an electronic switch would be maybe more reliable.
I have never done this, so i don't know exactly how.

But a manual switch some where for an owner to work whenever he
remembers to
make the switch would do better than nothing.


Patrick Turner.

[Iain Churches[_2_]]

"West" wrote in message news:c%IHi.2074$oc2.997@trnddc04...

"Chris Hornbeck" wrote in message
...
On Sat, 15 Sep 2007 09:51:50 +0300, "Iain Churches"
wrote:


He has also told me that heaters should be
biased above the cathode (for which you need a floating heater supply)

Isn't this mostly for hum control?

No. If the DC on the heaters is to be elevated, it must float,
so that one side is at say 90V and the other at 96.3V

There are two separate unrelated issues here. One is the
insulation between the very hot and uninsulated filament
and the hot and uninsulated cathode (all sorta/kinda true).

The other issue is the circuit noise immunity from
heater voltage. Leakage here is minimized by keeping
the cathode voltage below heater voltage. (Otherwise
it gets grand delusions of being an anode and drawing
electrons to itself. The little bitch.)

The main issue is which ac or dc is better for tube longevity
and why.

For longevity, I would put my money on AC for the reasons
mentioned earlier, but in many applications, DC is required if
you want low levels of hum.

In a discussion about heater regulators, he was of the opinion that it
is
the *current* which needs to be regulated and not the voltage, as a
supply
within 5% tolerance is quite acceptable. The cold heater draws several
times its nominal current.

I'm not sure how you can current regulate a device that draws about 7X
more
curent when cold.
It's like trying to regulate a thermister.

I have seen a current regulator with an LM317, which has a series resistor
on the output. The desired current it set by the value of the
resistor which is found by dividing the base voltage (usually 1.25V) by the
current required. So, if we want say 900mA we get 1.25/0.9 = 1.39 Ohms
The resistor needs to be of sufficient wattage to dissipate the heat.
You need 10V DC on the input, and an electrolytic between OP and
ground.

Regards
Iain

[John Byrns]

In article ,
Chris Hornbeck wrote:

On Mon, 17 Sep 2007 07:53:20 -0500, John Byrns
wrote:

One strange thing that I remember is that the (linear) final also
amplifies the video carrier. You'd think they'd use a separate
class-C carrier final and mix 'em together, but nope. Sound
carrier was separate, IIRC.

I'm not sure what you are trying to say above, the writing seems a bit
muddled and the meaning is unclear.

Sorry. The video final operates in class AB, and is burdened by
linearly amplifying the carrier along with the upper sideband.

Thanks Chris, that clarifies what you were talking about, I completely
misinterpreted what you were saying.

The carrier could easily be supplied by a class C final, and
the overall efficiency improved, but there were apparently
other considerations.

It's not obvious to me that this could be easily done, what are your
ideas on how to do it? Keep in mind that in Television the modulation
extends down to DC which means that the sideband amplifier is going to
have to be able to produce a carrier component which may be either in
phase with, or out of phase with, the carrier coming from the class C
amplifier depending on the scene being transmitted. If we simply "add"
the two signals the apparent load impedance seen by the two amplifiers
is going to vary a great deal with modulation which will doubtless cause
all sorts of problems. This problem could be eliminated by using a
hybrid combiner to add the two signals, but then we dump half the power
into a dummy load which would negate any efficiency gains, and probably
then some.

A Doherty Linear as used in AM radio broadcasting might be a better way
to go, the Doherty circuit puts the impedance variation to useful work
when combining the output of two amplifiers.


Regards,

John Byrns

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

[John Byrns]

In article ,
"Ian Iveson" wrote:

John Byrns wrote something a bit interesting:

"The final tubes in the TV transmitter were quite a bit
larger than a
dinner plate, their scale was more that of man sized, about
my size
IIRC. The cathode was heated by a 5,000 volt supply IIRC.
The cathode
of the main amplifier portion of the tube was also the anode
of a vacuum
diode which had the aforementioned 5,000 volt potential
applied between
its cathode and the anode/cathode. The voltage on the
filament of the
diode was reversed automatically 120 times a second. This
vacuum diode
heater system was called a "bombarder"."

I wonder if the diode John refers to here was a diode in the
sense we generally assume, ie that it would conduct only in
one direction?

Yes, the bombarder was a diode in the ordinary sense, at least as far as
I know. Also as far as I know the diode could only conduct in one
direction, as with any ordinary thermionic diode. The bombarder had a
filamentary cathode, the current through the diode heated the diode
plate to a high temperature, the diode plate in turn heated the cathode
of the main tube, which was part of the same structure. I think I may
have overstated the voltage that the bombarder operated on when I said
5,000 volts, now that I think about it, it was probably more like 3,000
volts, or maybe even slightly less.


Regards,

John Byrns

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