First, let me say thank you both for all the info that you've shared about
the GZ34 rectifier, bias rectifier, and recommendations for
improvement, troubleshooting and fixes.

With that said, I still don't have an answer to my original question: Is a
high ramp up of current draw with a return to set bias normal for the
6CA7?

I realize that I didn't ask the question properly, mislead Patrick with a
reference to anode when I should have said cathode, and misunderstood
Chris's reference to replacing the selenium rectifier with a silicon one.

Sorry about that. Permit me an attempt at what I meant to convey and not
what I said. ;-)

1. I have a Dynaco ST-70 Power Amplifier. It's working fine with a silicon
rectifier in place of the original selenium one used for bias voltage.
I've also replaced the power tube's 15.6 ohm cathode resistor with a 10
ohm one. It uses EL-34's and there is a reference in the documentation
that 6CA7's can be used as well with the proper bias adjustment.

2. I recently obtained a second ST-70 that came with Sylvania 6CA7's and
wanted to see if they would indeed work in place of the Svetlana EL-34's
that I'm using in my working amp.

3. After installing them and powering up the amp I observed that the
voltage at the cathode (which would normally be 1.00 volt with my current
arrangement), during warmup went from 0 volts to 3.00 volts and then
returned to 0.5 volts (the lowest setting on my bias potentiometer).

There's my premise, now my question:

Is this normal for 6CA7's or are these tubes showing their age?

I ask this as the EL-34's go from 0 to 1.00 volt during warmup and don't
"ramp up" their voltage beyond that with a return to normal bias.

/SUMMATION MODE ON/

Chris mentioned that the tubes might be gassy, suggested that the selenium
rectifier used for bias as well as capacitors should be replaced with
modern equivalents and then got caught up in my unintentional misdirection
about replacing the rectifier tube with silicon.

Patrick ventured that the process of obtaining bias as used by the Dynaco
is inferior and suggested several ways to obtain the bias through more
reliable means as well as overvolt protection to prevent catastrophic
failure.

Other methods for checking bias circuit characteristics and a few
excellent hints about the bias and power rectification circuits' were also
brought up (this is very excellent info by the way, my head is going to
explode with all that raw input that I'll be researching in the next few
days).

/SUMMATION MODE OFF/

I probably missed a few points there, but I've archived all answers and
will most carefully look them over for whatever nuggets of information
that I can find about tube amps.

Again, thank you both. Is my question worded in a more understandable
format?

And please don't say "fourty-two". :-P

-Bart

bart wrote:

First, let me say thank you both for all the info that you've shared about
the GZ34 rectifier, bias rectifier, and recommendations for
improvement, troubleshooting and fixes.

With that said, I still don't have an answer to my original question: Is a
high ramp up of current draw with a return to set bias normal for the
6CA7?

I realize that I didn't ask the question properly, mislead Patrick with a
reference to anode when I should have said cathode, and misunderstood
Chris's reference to replacing the selenium rectifier with a silicon one.

Sorry about that. Permit me an attempt at what I meant to convey and not
what I said. ;-)

1. I have a Dynaco ST-70 Power Amplifier. It's working fine with a silicon
rectifier in place of the original selenium one used for bias voltage.
I've also replaced the power tube's 15.6 ohm cathode resistor with a 10
ohm one. It uses EL-34's and there is a reference in the documentation
that 6CA7's can be used as well with the proper bias adjustment.

2. I recently obtained a second ST-70 that came with Sylvania 6CA7's and
wanted to see if they would indeed work in place of the Svetlana EL-34's
that I'm using in my working amp.

3. After installing them and powering up the amp I observed that the
voltage at the cathode (which would normally be 1.00 volt with my current
arrangement), during warmup went from 0 volts to 3.00 volts and then
returned to 0.5 volts (the lowest setting on my bias potentiometer).

There's my premise, now my question:

Is this normal for 6CA7's or are these tubes showing their age?

Its not normal for any tubes to draw 3 times the normal bias currents during
warm up
as you suggest.

During warm up, check that there isn't a period of supersonic oscillations
causing large current consumption and the DC voltage you see across the
cathode R.

Chances are that the B+ rises to its peak surge maximum possible value
before the output tubes begin to conduct and when they do the B+ is way higher
than it is when
the amp has settled down so the cathode current will rise fast but it doesn't
explain why the bias
current is 3 times normal.

However, the driver amp also has to come on in warm up, and what may be
happening
is that one of the output tubes is turned on with slow moving dcV settlement
transients
causing saturation of that tube, hence 300mA.
if you had a pair of 10ohm currentent sensing R, one for each cathode, then
the
turn on behaviour of each tube could be tracked far better.

So I suggest replace the single common 10 ohms with a pair of 10 ohms.
And the big benefit is that you will be able to see if there is equal bias
current in reach tube,
something very necessary expecially as the tubes age.

But place a CRO on the cathode R on a sensitive voltage range to see if any RF
signals
occur during turn on.
Also check the turn off behaviour.

Some amps are unstable during turn on and turn off.



I ask this as the EL-34's go from 0 to 1.00 volt during warmup and don't
"ramp up" their voltage beyond that with a return to normal bias.

Old tubes tend to be slower to turn on and your symptoms are not that of aged
tubes.



/SUMMATION MODE ON/

Chris mentioned that the tubes might be gassy, suggested that the selenium
rectifier used for bias as well as capacitors should be replaced with
modern equivalents and then got caught up in my unintentional misdirection
about replacing the rectifier tube with silicon.

Gassy tubes that are old and past their use by date usually have positive
grids with respect to the bias
voltage at the 10k pots of the bias supply, hence my mention of measuring the
voltage across the 270k bias R.




Patrick ventured that the process of obtaining bias as used by the Dynaco
is inferior and suggested several ways to obtain the bias through more
reliable means as well as overvolt protection to prevent catastrophic
failure.

Such protection measures in older tube amps were never fitted because
transistors and scrs and zeners were yet to be invented or reliable, and when
they were the tubes
were retired and all the circuits went to SS.
Protection circuits using tubes were not deemed cost effective.
People were expected to service their amps regularly, but most didn't, and
they suffered blown tubes and a wrecked OPT.
Re-tubing was avoided because of the expense, and only the BBC and
the military ever bothered to log the hours of the tubes and replace them
to prevent failures.
But many ppl took time to blow a tube or two, and when they did
they found stereo SS had arrived so they dumped their mono tube amp systems.
In 1965, you could easily buy a new OPT and replacement tubes.



Other methods for checking bias circuit characteristics and a few
excellent hints about the bias and power rectification circuits' were also
brought up (this is very excellent info by the way, my head is going to
explode with all that raw input that I'll be researching in the next few
days).

/SUMMATION MODE OFF/

I probably missed a few points there, but I've archived all answers and
will most carefully look them over for whatever nuggets of information
that I can find about tube amps.

Again, thank you both. Is my question worded in a more understandable
format?

Maybe we are getting there, but the reasons for the
dc current turn on behaviour remain unknown.

Patrick Turner.



And please don't say "fourty-two". :-P

-Bart

Bart:

I have a rebuilt-to-stock ST-70 (excepting the silicon diode ILO the
selenium), and both sets of EL34 and 6CA7 tubes NIB and 'well-used'.

Let me know where you took your measurements and I will try to
duplicate them, if for no other reason than to try to determine which
of your two units is the outlyer. Did you read right off the Bias Point
or somewhere else? I will be measuring with a Fluke true-rms DMM.

I am sure that you are aware that the 5AR4/GZ34 is what is called a
'soft-start' rectifier. It can be replaced with diodes, but that
soft-start option should also be duplicated. There are several sources
for SS drop-in replacements that duplicate the action for far less than
a NOS US/UK unit.

Peter Wieck
Wyncote, PA

On Mon, 28 Nov 2005 01:55:40 -0500, bart wrote:

And please don't say "fourty-two". :-P

What's nine times seven?

Patrick covered a lot of outlying possibilities, just
in case ('cause cases happen), but first, just
replace the goll-durned selenium rectifier and retest.
Can't cost 25 cents, and ain't worth the drama.

Cathode current is completely and solely a simple,
known and repeatable function of grid voltage and
plate (= G2, in this case) voltage.

Plate voltage is pretty much independent of the
particular tubes/ valves personalities. Grid (G1)
voltage is potentially affected (adversely, for
our purposes) by the tubes/ valves gassiness.
(For practical non-zero DC source impedances).

That's it. That's the entire model.




ps: I still don't recommend American 6CA7's for
Dyna ST70's because their higher transconductance can
too often cause motorboating. A modern revision
or rebuild shouldn't have this potential issue.

Good fortune,


Chris Hornbeck

Chris Hornbeck wrote:

On Mon, 28 Nov 2005 01:55:40 -0500, bart wrote:


And please don't say "fourty-two". :-P


What's nine times seven?

sixty-three. You should buy a calculator
if you don't know your times tables.

What is six times seven?

hth

Andy

On Mon, 28 Nov 2005 09:41:27 +0000, Patrick Turner wrote:

Subject: attn: Chris Hornbeck and Patrick Turner A question (6CA7 bias current)
From: Patrick Turner
Newsgroups: rec.audio.tubes
Date: Mon, 28 Nov 2005 09:41:27 GMT



bart wrote:

First, let me say thank you both for all the info that you've shared about
the GZ34 rectifier, bias rectifier, and recommendations for
improvement, troubleshooting and fixes.

With that said, I still don't have an answer to my original question: Is a
high ramp up of current draw with a return to set bias normal for the
6CA7?

I realize that I didn't ask the question properly, mislead Patrick with a
reference to anode when I should have said cathode, and misunderstood
Chris's reference to replacing the selenium rectifier with a silicon one.

Sorry about that. Permit me an attempt at what I meant to convey and not
what I said. ;-)

1. I have a Dynaco ST-70 Power Amplifier. It's working fine with a silicon
rectifier in place of the original selenium one used for bias voltage.
I've also replaced the power tube's 15.6 ohm cathode resistor with a 10
ohm one. It uses EL-34's and there is a reference in the documentation
that 6CA7's can be used as well with the proper bias adjustment.

2. I recently obtained a second ST-70 that came with Sylvania 6CA7's and
wanted to see if they would indeed work in place of the Svetlana EL-34's
that I'm using in my working amp.

3. After installing them and powering up the amp I observed that the
voltage at the cathode (which would normally be 1.00 volt with my current
arrangement), during warmup went from 0 volts to 3.00 volts and then
returned to 0.5 volts (the lowest setting on my bias potentiometer).

There's my premise, now my question:

Is this normal for 6CA7's or are these tubes showing their age?

Its not normal for any tubes to draw 3 times the normal bias currents during
warm up
as you suggest.

During warm up, check that there isn't a period of supersonic oscillations
causing large current consumption and the DC voltage you see across the
cathode R.

Chances are that the B+ rises to its peak surge maximum possible value
before the output tubes begin to conduct and when they do the B+ is way higher
than it is when
the amp has settled down so the cathode current will rise fast but it doesn't
explain why the bias
current is 3 times normal.

However, the driver amp also has to come on in warm up, and what may be
happening
is that one of the output tubes is turned on with slow moving dcV settlement
transients
causing saturation of that tube, hence 300mA.
if you had a pair of 10ohm currentent sensing R, one for each cathode, then
the
turn on behaviour of each tube could be tracked far better.

So I suggest replace the single common 10 ohms with a pair of 10 ohms.
And the big benefit is that you will be able to see if there is equal bias
current in reach tube,
something very necessary expecially as the tubes age.

But place a CRO on the cathode R on a sensitive voltage range to see if any RF
signals
occur during turn on.
Also check the turn off behaviour.

Some amps are unstable during turn on and turn off.



I ask this as the EL-34's go from 0 to 1.00 volt during warmup and don't
"ramp up" their voltage beyond that with a return to normal bias.

Old tubes tend to be slower to turn on and your symptoms are not that of aged
tubes.



/SUMMATION MODE ON/

Chris mentioned that the tubes might be gassy, suggested that the selenium
rectifier used for bias as well as capacitors should be replaced with
modern equivalents and then got caught up in my unintentional misdirection
about replacing the rectifier tube with silicon.

Gassy tubes that are old and past their use by date usually have positive
grids with respect to the bias
voltage at the 10k pots of the bias supply, hence my mention of measuring the
voltage across the 270k bias R.




Patrick ventured that the process of obtaining bias as used by the Dynaco
is inferior and suggested several ways to obtain the bias through more
reliable means as well as overvolt protection to prevent catastrophic
failure.

Such protection measures in older tube amps were never fitted because
transistors and scrs and zeners were yet to be invented or reliable, and when
they were the tubes
were retired and all the circuits went to SS.
Protection circuits using tubes were not deemed cost effective.
People were expected to service their amps regularly, but most didn't, and
they suffered blown tubes and a wrecked OPT.
Re-tubing was avoided because of the expense, and only the BBC and
the military ever bothered to log the hours of the tubes and replace them
to prevent failures.
But many ppl took time to blow a tube or two, and when they did
they found stereo SS had arrived so they dumped their mono tube amp systems.
In 1965, you could easily buy a new OPT and replacement tubes.



Other methods for checking bias circuit characteristics and a few
excellent hints about the bias and power rectification circuits' were also
brought up (this is very excellent info by the way, my head is going to
explode with all that raw input that I'll be researching in the next few
days).

/SUMMATION MODE OFF/

I probably missed a few points there, but I've archived all answers and
will most carefully look them over for whatever nuggets of information
that I can find about tube amps.

Again, thank you both. Is my question worded in a more understandable
format?

Maybe we are getting there, but the reasons for the
dc current turn on behaviour remain unknown.

Patrick Turner.



And please don't say "fourty-two". :-P

-Bart


Okay. Not ignoring you, just working on attempting to understand all the
information provided and implement all the troubleshooting and measurement
suggestions. should be able to complete the testing and measurement
within a few days.

Thanks!

-Bart

On Tue, 29 Nov 2005 07:02:05 +0000, Chris Hornbeck wrote:

On Mon, 28 Nov 2005 01:55:40 -0500, bart wrote:

And please don't say "fourty-two". :-P

What's nine times seven?

Umm, an integer between fourty one and fourty three?


Patrick covered a lot of outlying possibilities, just
in case ('cause cases happen), but first, just
replace the goll-durned selenium rectifier and retest.
Can't cost 25 cents, and ain't worth the drama.

That's done. Tube still exhibits the ramp up characteristic. After
reading the comments written into the provided manual, it would appear
that the original owner experienced a bias problem with the original
6CA7's and replaced them, the 30-20-20-20 uf cap and the bias capacitors
in 1978. It didn't solve the problem and he left the cathode voltage at
1.71 volts vs 1.56 volts. I'll be warming up the original chassis
shortly, checking out the bias circuit and seeing what the tubes' behavior
is. I don't want to blow out my existing working one.

Also, this thing has been modded per the Jensen specification (1978).
Dunno if I should go back to the original 5AR4 to drop the voltage down or
not.


Cathode current is completely and solely a simple,
known and repeatable function of grid voltage and
plate (= G2, in this case) voltage.

Plate voltage is pretty much independent of the
particular tubes/ valves personalities. Grid (G1)
voltage is potentially affected (adversely, for
our purposes) by the tubes/ valves gassiness.
(For practical non-zero DC source impedances).

That's it. That's the entire model.

Okay, I'll keep it in mind.





ps: I still don't recommend American 6CA7's for
Dyna ST70's because their higher transconductance can
too often cause motorboating. A modern revision
or rebuild shouldn't have this potential issue.

Revision as in a zener at the -48 volt portion of the chain or something
more drastic? Have you any web available examples of this?


Good fortune,


Chris Hornbeck

On Tue, 29 Nov 2005 12:36:17 +0000, Andy Cowley wrote:

Chris Hornbeck wrote:

On Mon, 28 Nov 2005 01:55:40 -0500, bart wrote:


And please don't say "fourty-two". :-P


What's nine times seven?

sixty-three. You should buy a calculator
if you don't know your times tables.

What is six times seven?

Umm, is it equivalent to nine times seven minus three times seven? (I'm
feeling a bit low in the IQ department today).


hth

Andy

On Mon, 28 Nov 2005 11:37:17 -0800, pfjw wrote:

Bart:

I have a rebuilt-to-stock ST-70 (excepting the silicon diode ILO the
selenium), and both sets of EL34 and 6CA7 tubes NIB and 'well-used'.

Let me know where you took your measurements and I will try to
duplicate them, if for no other reason than to try to determine which
of your two units is the outlyer. Did you read right off the Bias Point
or somewhere else? I will be measuring with a Fluke true-rms DMM.

Right at the provided wiring point of the octal socket on the front of the
chassis (marked "BIASET 1.56V"). It shows the same characteristic whether
I use silicon or the 5AR4 for power rectification.


I am sure that you are aware that the 5AR4/GZ34 is what is called a
'soft-start' rectifier. It can be replaced with diodes, but that
soft-start option should also be duplicated. There are several sources
for SS drop-in replacements that duplicate the action for far less than
a NOS US/UK unit.

Are any examples of the circuitry used available on the web?


Peter Wieck
Wyncote, PA

bart wrote:

On Tue, 29 Nov 2005 07:02:05 +0000, Chris Hornbeck wrote:


On Mon, 28 Nov 2005 01:55:40 -0500, bart wrote:


And please don't say "fourty-two". :-P

What's nine times seven?


Umm, an integer between fourty one and fourty three?


What base are you working in?

Andy

(clue: In base 10 9x7=63)

Andy Cowley wrote:

bart wrote:

On Tue, 29 Nov 2005 07:02:05 +0000, Chris Hornbeck wrote:


On Mon, 28 Nov 2005 01:55:40 -0500, bart wrote:


And please don't say "fourty-two". :-P

What's nine times seven?


Umm, an integer between fourty one and fourty three?


What base are you working in?

Andy

(clue: In base 10 9x7=63)

Well I reckon you guys are talking in some secret coded spylogical amp
talk.

Is it about a secret airforce base?, ie, the base you workin in?

But all i know is that I am lucky if I get a 69'er.

Patrick Turner.

bart wrote

With that said, I still don't have an answer to my original
question: Is a
high ramp up of current draw with a return to set bias normal for
the
6CA7?

It's not the kind of thing that a healthy valve can do on its own.

Does the current surge happen every time you power up, even if you
switch off and then on again after a minute or so?

Perhaps the valves have run out of electrons, or rather whatever it
is in the cathode that helps to emit them. It normally leaches to
the surface to replenish what is lost. Probably too slow a process
to explain this though.

Or perhaps they warm up quicker, before the grid bias has ramped up.
Have you tracked the grid voltage (with a high impedance meter) and
observed the phenomenon there?

Sorry if it's already sorted...I don't think I have the whole
thread. Neither am I Chris Hornbeck *or* Patrick Turner, they will
be glad to know.

cheers, Ian

On Tue, 29 Nov 2005 12:36:17 GMT, Andy Cowley
wrote:

What is six times seven?

Life, the Universe, and Everything.

HTH,

Chris Hornbeck

On Wed, 30 Nov 2005 00:57:21 GMT, "Ian Iveson"
wrote:

question: Is a
high ramp up of current draw with a return to set bias normal for
the
6CA7?

It's not the kind of thing that a healthy valve can do on its own.

Bingo.


Have you tracked the grid voltage (with a high impedance meter) and
observed the phenomenon there?

By this, Ian means observing both the idealized bias
voltage at the bias pot's wiper tap and the actual voltage
at the tube/ valves grid (G1) pin (sorry, can't remember;
pin 5 for a 7S valve?).

"Gassy" valves conduct noticable current into G1, and in
a direction that *reduces* actual G1 voltage. Result: higher
cathode current than expected.

A good floating meter across the power valves' grid resistors
(100K?, 270K ohm? ; sorry, can't remember) might tell you
something interesting. Anything over a volt is interesting.

Also, I haven't seen anything yet posted (coulda missed it,
natch) about the coupling caps. The originals probably
voted for Nixon; they deserve replacement wholesale.
For testing purposes you can just disconnect 'em.


Sorry if it's already sorted...I don't think I have the whole
thread. Neither am I Chris Hornbeck *or* Patrick Turner, they will
be glad to know.

Lucky You in the former case, and possibly also in the latter,
if we can believe Patrick's tales of pussy drought. But then,
nothing's forever.

Thanks, as always,

Chris Hornbeck

"Ian Iveson" wrote in message
.uk...
bart wrote

With that said, I still don't have an answer to my original question: Is
a
high ramp up of current draw with a return to set bias normal for the
6CA7?

It's not the kind of thing that a healthy valve can do on its own.

Does the current surge happen every time you power up, even if you switch
off and then on again after a minute or so?

Perhaps the valves have run out of electrons, or rather whatever it is in
the cathode that helps to emit them. It normally leaches to the surface to
replenish what is lost. Probably too slow a process to explain this
though.

Or perhaps they warm up quicker, before the grid bias has ramped up. Have
you tracked the grid voltage (with a high impedance meter) and observed
the phenomenon there?

Sorry if it's already sorted...I don't think I have the whole thread.
Neither am I Chris Hornbeck *or* Patrick Turner, they will be glad to
know.

cheers, Ian

One item I didn't catch (coulda missed it) is if both sets of output tubes
behave exactly the same. If they do, then the problem may lie in a common
area to both channels, such as the bias supply as Ian suggests. Measuring
the grid voltage at an OT socket as the amp is turned on is one of the first
items I would look at.
Mark

Mark S wrote:

"Ian Iveson" wrote in message
.uk...
bart wrote

With that said, I still don't have an answer to my original question: Is
a
high ramp up of current draw with a return to set bias normal for the
6CA7?

It's not the kind of thing that a healthy valve can do on its own.

Does the current surge happen every time you power up, even if you switch
off and then on again after a minute or so?

Perhaps the valves have run out of electrons, or rather whatever it is in
the cathode that helps to emit them. It normally leaches to the surface to
replenish what is lost. Probably too slow a process to explain this
though.

Or perhaps they warm up quicker, before the grid bias has ramped up. Have
you tracked the grid voltage (with a high impedance meter) and observed
the phenomenon there?

Sorry if it's already sorted...I don't think I have the whole thread.
Neither am I Chris Hornbeck *or* Patrick Turner, they will be glad to
know.

cheers, Ian

One item I didn't catch (coulda missed it) is if both sets of output tubes
behave exactly the same. If they do, then the problem may lie in a common
area to both channels, such as the bias supply as Ian suggests. Measuring
the grid voltage at an OT socket as the amp is turned on is one of the first
items I would look at.
Mark

The guy has solid state grid bias rectifiers that should provide the full bias
voltage with a couple of seconds, and then the B+ would turn on slow from the
tube rectifier.
Still he gets a large current surge at turn on.

But all this has to be checked out carefully, including the driver amp turn on
behaviour as well, to see if any LF transient is applied to one or both output
tube
grids during power up.

Patrick Turner.

Patrick Turner wrote

One item I didn't catch (coulda missed it) is if both sets of
output tubes
behave exactly the same. If they do, then the problem may lie in
a common
area to both channels, such as the bias supply as Ian suggests.
Measuring
the grid voltage at an OT socket as the amp is turned on is one
of the first
items I would look at.

The guy has solid state grid bias rectifiers that should provide
the full bias
voltage with a couple of seconds, and then the B+ would turn on
slow from the
tube rectifier.
Still he gets a large current surge at turn on.


The time taken for grid bias to stabilise depends not just on the
type of rectifier, but also on the time constant of the bias
circuit. Mine, for example, takes nearly 10 seconds to level out. If
I take the timer out of the HT circuit, my output valves would
behave just like the 6CA7s in question.

The problem here though is that the EL34s seem OK. So I was trying
to think of a reason for a different time constant somewhere in the
valves themselves. Warming-up is the only one I know of with an
appropriate delay, comparable to another delay in the supporting
circuit: the rise time of the grid bias voltage. The circuit is not
in my hands, and all I have in my head is logic.

Hence the possibly improbable idea of running out of electrons. A
battery or an electrolytic cap can show a high voltage for an
instant before collapsing, if it is discharged and/or poorly, in a
similar fashion to the current in the 6CA7s. So I wondered about a
similar cause, in the chemistry of the cathode or the space around
it.

Thing is, I've never seen a valve that is dying a natural death.
They have always been abuse victims, or have inexplicably crumbled
and shorted somewhere. So tell me people, in your experience, do
valves ever develop a fault whereby upon warming up and given a
constant bias voltage, they race away but then wilt to half the
expected current, within a few seconds? It seems unlikely to me,
because I guess the metal chemistry would be too slow, and space
chemistry too quick.

But all this has to be checked out carefully, including the driver
amp turn on
behaviour as well, to see if any LF transient is applied to one or
both output
tube
grids during power up.

As several of us are saying.

cheers, Ian

Ian Iveson wrote:

Patrick Turner wrote

One item I didn't catch (coulda missed it) is if both sets of
output tubes
behave exactly the same. If they do, then the problem may lie in
a common
area to both channels, such as the bias supply as Ian suggests.
Measuring
the grid voltage at an OT socket as the amp is turned on is one
of the first
items I would look at.

The guy has solid state grid bias rectifiers that should provide
the full bias
voltage with a couple of seconds, and then the B+ would turn on
slow from the
tube rectifier.
Still he gets a large current surge at turn on.

The time taken for grid bias to stabilise depends not just on the
type of rectifier, but also on the time constant of the bias
circuit. Mine, for example, takes nearly 10 seconds to level out. If
I take the timer out of the HT circuit, my output valves would
behave just like the 6CA7s in question.

But a bias circuit can be made to be very noise free and still be able
to power up in less than 3 seconds.

The heaters slow the turn on of the OP tubes.

Hence I have asked thre poster to measure the time things take to happen

during power up.




The problem here though is that the EL34s seem OK. So I was trying
to think of a reason for a different time constant somewhere in the
valves themselves. Warming-up is the only one I know of with an
appropriate delay, comparable to another delay in the supporting
circuit: the rise time of the grid bias voltage. The circuit is not
in my hands, and all I have in my head is logic.

Tube age may make the effective R part of the time constant longer
than tubes which turn on faster than old ones.
But where is the C element for a time constant to exist?

Just be careful of that logic in your head....



Hence the possibly improbable idea of running out of electrons. A
battery or an electrolytic cap can show a high voltage for an
instant before collapsing, if it is discharged and/or poorly, in a
similar fashion to the current in the 6CA7s. So I wondered about a
similar cause, in the chemistry of the cathode or the space around
it.

I can safely say that I have NEVER seen the kind of behaviour as
repeorted in any
amps I have measured.

Its either a burst of oscillations at HF during power up,
or a LF transient applied to the output tubes differentially or common
mode
by the drive amp; The ST70 has a concertina phase inverter whose anode
voltage
starts high and falls as the tube warms, and a cathode voltage that does
the opposite, so
the output tube being driven from a cap from the CPI cathode if TC of
the cap and bias R
is long may get a LF transient.


Thing is, I've never seen a valve that is dying a natural death.
They have always been abuse victims, or have inexplicably crumbled
and shorted somewhere.

Spontaneous malfunction is a natural death, like death caused in us
by cancer, rather than very old age and gradual enfeeblement of all
facilities.

I have seen much natural death.

So tell me people, in your experience, do
valves ever develop a fault whereby upon warming up and given a
constant bias voltage, they race away but then wilt to half the
expected current, within a few seconds? It seems unlikely to me,
because I guess the metal chemistry would be too slow, and space
chemistry too quick.

I have not winessed the surging of Ia after turn on unless the tubes are

hot to start with.



But all this has to be checked out carefully, including the driver
amp turn on
behaviour as well, to see if any LF transient is applied to one or
both output
tube
grids during power up.

As several of us are saying.

I didn't see too many raise the issue.

But I have not read every single thread.

Patrick Turner.



cheers, Ian