I ask this question because earlier in the evening I wired up a single
triode gain stage based on a 5610.

http://www.nj7p.org/Tube1.php?tube=5610

With a 5.6k plate resistor and 130 B+ the plate sees 78v at 10ma. The
cathode resistor was 220r and grid leak 470k. A 1v input gave 5v out
with a nice sinewave. Adding a 220uf or 1000uf cap on the cathode
resistor caused a noticable change on the bottom peaks, of course the
gain was higher at 7v out.

Using a generic red LED on the cathode the output was again 7v and plate
current was now 12ma. The output looked much cleaner and had little
change from 20hz to over 100kz.

Other than the stigma of using silicon in the signal path why isn't this
used in more designs?


Adam

Adam Stouffer wrote:

I ask this question because earlier in the evening I wired up a single
triode gain stage based on a 5610.

http://www.nj7p.org/Tube1.php?tube=5610

With a 5.6k plate resistor and 130 B+ the plate sees 78v at 10ma. The
cathode resistor was 220r and grid leak 470k. A 1v input gave 5v out
with a nice sinewave. Adding a 220uf or 1000uf cap on the cathode
resistor caused a noticable change on the bottom peaks, of course the
gain was higher at 7v out.

Using a generic red LED on the cathode the output was again 7v and plate
current was now 12ma. The output looked much cleaner and had little
change from 20hz to over 100kz.

Other than the stigma of using silicon in the signal path why isn't this
used in more designs?

Adam

I never use Si diodes for biasing signal tubes because the diode
does not help to regulate the bias current, since it acts as a fixed
voltage,
so the stage is acting under fixed bias.

But 1,000 uF is only 16 ohms impedance at 10Hz,
and is what i routinely use for bypassing in power and preamps,
sometimes with a 0.47 polyprop also across it.
I don't see how R&C bypassing could give any worse sine waves.

If the diode has some finite "on" impedance,
perhaps its non linear, and across a wide f range but
a 1,000 uF has falling Z as F rises....

I don't use battery bias either.

Patrick turner.

Adam Stouffer wrote:

I ask this question because earlier in the evening I wired up a single
triode gain stage based on a 5610.

http://www.nj7p.org/Tube1.php?tube=5610

With a 5.6k plate resistor and 130 B+ the plate sees 78v at 10ma. The
cathode resistor was 220r and grid leak 470k. A 1v input gave 5v out
with a nice sinewave. Adding a 220uf or 1000uf cap on the cathode
resistor caused a noticable change on the bottom peaks, of course the
gain was higher at 7v out.

Using a generic red LED on the cathode the output was again 7v and plate
current was now 12ma. The output looked much cleaner and had little
change from 20hz to over 100kz.

Other than the stigma of using silicon in the signal path why isn't this
used in more designs?

Because they are all based on old thinking ?

Graham

Adam Stouffer wrote:

Other than the stigma of using silicon in the signal path why isn't this
used in more designs?

p.s. LEDs aren't made of silicon so there's no stigma ! ;-)

Graham

Adam Stouffer wrote:
I ask this question because earlier in the evening I wired up a single
triode gain stage based on a 5610.

http://www.nj7p.org/Tube1.php?tube=5610

With a 5.6k plate resistor and 130 B+ the plate sees 78v at 10ma. The
cathode resistor was 220r and grid leak 470k. A 1v input gave 5v out
with a nice sinewave. Adding a 220uf or 1000uf cap on the cathode
resistor caused a noticable change on the bottom peaks, of course the
gain was higher at 7v out.

Using a generic red LED on the cathode the output was again 7v and plate
current was now 12ma. The output looked much cleaner and had little
change from 20hz to over 100kz.

Other than the stigma of using silicon in the signal path why isn't this
used in more designs?


Adam

Essentially because a LED doesn't control the current but fixes the
voltage bias. Fixed bias, though it gives better bass, to the golden
ears unbalances the sound spectrum. So the ultrafidelista don't use
LEDs for bias much. You might say it is hypocritical because battery
bias, which is "acceptable", is also fixed bias.

LEDs are sometimes used elsewhere, though not by the hardcore faithful.
Designs by Eric Barbour and Steve Bench and others of the same
practical stamp, often have LEDs, and designs by those acceptable to
the untrafidelista DIYers because they are obsessively cheap, like Dan
Schmalle of Bottlehead/Valve/Electronic Tonalities fame, sometimes have
LEDs.

There's a lot of hypocrisy in high end tubes, but you can make yourself
unpopular by pointing out that an SRPP is not a constant current
device, that a cathode follower adds negative feedback, that a DHT
works so well because it has negative feedback built in, that trioded
pentodes sound brilliant, that DHT in PP can sound better than in SE,
and so on, a whole testament full of articles of faith contrary to
known electronic theory and the experience of the open-minded.

Of course, I could argue that LEDs should be acceptable because they
are, specifically, not silicon...

Another reason for the men, mostly middeaged, who build tube amps is
that LED's are just too twee. A film cap is a big fistful of component,
easily soldered in, costs plenty and looks like it. A Kiwame resistor
is a fat green slug looking purposeful. A battery used for bias ditto.
A LED shines like cheap jewellery. If we wanted to be in cheap
itty-bitty parts, we'd build silicon amps.

Andre Jute

Andre Jute wrote:

Adam Stouffer wrote:
I ask this question because earlier in the evening I wired up a single
triode gain stage based on a 5610.

http://www.nj7p.org/Tube1.php?tube=5610

With a 5.6k plate resistor and 130 B+ the plate sees 78v at 10ma. The
cathode resistor was 220r and grid leak 470k. A 1v input gave 5v out
with a nice sinewave. Adding a 220uf or 1000uf cap on the cathode
resistor caused a noticable change on the bottom peaks, of course the
gain was higher at 7v out.

Using a generic red LED on the cathode the output was again 7v and plate
current was now 12ma. The output looked much cleaner and had little
change from 20hz to over 100kz.

Other than the stigma of using silicon in the signal path why isn't this
used in more designs?


Adam

Essentially because a LED doesn't control the current but fixes the
voltage bias. Fixed bias, though it gives better bass, to the golden
ears unbalances the sound spectrum. So the ultrafidelista don't use
LEDs for bias much. You might say it is hypocritical because battery
bias, which is "acceptable", is also fixed bias.

LEDs are sometimes used elsewhere, though not by the hardcore faithful.
Designs by Eric Barbour and Steve Bench and others of the same
practical stamp, often have LEDs, and designs by those acceptable to
the untrafidelista DIYers because they are obsessively cheap, like Dan
Schmalle of Bottlehead/Valve/Electronic Tonalities fame, sometimes have
LEDs.

There's a lot of hypocrisy in high end tubes, but you can make yourself
unpopular by pointing out that an SRPP is not a constant current
device, that a cathode follower adds negative feedback, that a DHT
works so well because it has negative feedback built in, that trioded
pentodes sound brilliant, that DHT in PP can sound better than in SE,
and so on, a whole testament full of articles of faith contrary to
known electronic theory and the experience of the open-minded.

I don't see what the fuss is about with diodes or batteries;
maybe they don't spoil music, but surely neither
improves it.

Fixed bias belongs in output stages in PP amps, because the
tube currents become full of 2H beyond the class A working
threshold to class AB.
In totally class A circuits, battery, diode, RC cathode bias, or plain fixed
grid bias
all should sound IDENTICAL if its done right.

But RC cathode bias is the only one giving regulation of the
bias current because at DC there is some negative current FB.


SRPP sure doesn't include any CCS loading for any part of the two
triodes used; SRPP is where the Rk of the bottom tube = the Rk
of the top tube, and without any load connected to the top tube cathode,
the load seen by the bottom is merely ( µ +1 ) x Rk, so that if µ = 20,
and Rk = 1.5k like an SRPP with 6SN7, then the bottom tube sees
a load of only 31.5k, which will give more 2h than with a 47k resistor
to deliver the DC to the bottom tube.

But due to the load, if connected, and the 2H of the top tube, some
cancelations of 2H occur,, and lowish thd measures result, although 3H is
slightly more.
Ro is usually only about 1/4 of Ra for one of the triodes.

Better of course is to have a µ follower, with the upper Rk about
10k, and then the bottom tube sees a **much ** higher load
approaching CCS, and the top tube works more like a follower,
with lower Ro, and the whole thing has more natural low thd of the
triode untroubled by too low a load used in a gain stage.
Thus 0.2% at 10Vrms is attainable, instead of 2%.

And in a preamp at 0.1Vrms out, thd = 2H = 0.01% is possible,
and it sounds well.....

Cathode followers are very misunderstood.
Ask anyone who complains about their sonic signature to tell you
how they work.
Expect a range of dumb answers.
I find that most followers preserve dynamics and clarity
and the fidelity of the gain stages preceeding them.



Of course, I could argue that LEDs should be acceptable because they
are, specifically, not silicon...

Another reason for the men, mostly middeaged, who build tube amps is
that LED's are just too twee. A film cap is a big fistful of component,
easily soldered in, costs plenty and looks like it. A Kiwame resistor
is a fat green slug looking purposeful. A battery used for bias ditto.
A LED shines like cheap jewellery. If we wanted to be in cheap
itty-bitty parts, we'd build silicon amps.

I don't use leds except to tell me when an amp is turned on,
or if there is a fault, but never in any tube audio circuit, and rarely in any

SS circuit.
I don't like having to use light bulbs to do what leds do far more easily and
reliably.

Patrick Turner.





Andre Jute

Patrick Turner wrote:
Andre Jute wrote:
There's a lot of hypocrisy in high end tubes, but you can make yourself
unpopular by pointing out that an SRPP is not a constant current
device, that a cathode follower adds negative feedback, that a DHT
works so well because it has negative feedback built in, that trioded
pentodes sound brilliant, that DHT in PP can sound better than in SE,
and so on, a whole testament full of articles of faith contrary to
known electronic theory and the experience of the open-minded.

I don't see what the fuss is about with diodes or batteries;
maybe they don't spoil music, but surely neither
improves it.

Fixed bias belongs in output stages in PP amps, because the
tube currents become full of 2H beyond the class A working
threshold to class AB.
In totally class A circuits, battery, diode, RC cathode bias, or plain fixed
grid bias
all should sound IDENTICAL if its done right.

The ultrafidelista fancy they can hear better bass with fixed bias,
including battery bias in Class A circuits. Certainly, before you use
battery bias, you'd better see to the top end extension in any DHT amp,
or the sound balance will be odd.

But RC cathode bias is the only one giving regulation of the
bias current because at DC there is some negative current FB.

Another NFB I missed out in my list above... (When the ultrafidelista
say "negative feedback" with a sneer, they mean only user-added
feedback. Feedback intrinsic to favourite tubes and topologies is like
the virginity of Mary Magdalen, very doubtful but never mentioned.)

SRPP sure doesn't include any CCS loading for any part of the two
triodes used; SRPP is where the Rk of the bottom tube = the Rk
of the top tube, and without any load connected to the top tube cathode,
the load seen by the bottom is merely ( µ +1 ) x Rk, so that if µ =20,
and Rk = 1.5k like an SRPP with 6SN7, then the bottom tube sees
a load of only 31.5k, which will give more 2h than with a 47k resistor
to deliver the DC to the bottom tube.

SRPP is more flexible than just Rkbottom equal Rktop; you can vary them
for various effects but it makes calculation more complex. See the MIT
Radiation Laboratory Series, Vol 18: Vacuum Tube Amplifiers; they call
SRPP a "two tube series arrangement"; cf p456, and check also pp 464
and 438 to be able to put together a coherent formula. I did this in a
set of posts a long time ago and don't now have time to do it again.

SRPP's purpose for the general neddy public is lowering Zo and
maximizing gain, but within that you can tune for max gain or max
current. I put all this in a spreadsheet about ten or twelve years ago,
so I no longer have the formula at my fingertips, which was of course
the point of building a model and a spreadsheet.

But due to the load, if connected, and the 2H of the top tube, some
cancelations of 2H occur,, and lowish thd measures result, although 3H is
slightly more.
Ro is usually only about 1/4 of Ra for one of the triodes.

Better of course is to have a µ follower, with the upper Rk about
10k, and then the bottom tube sees a **much ** higher load
approaching CCS, and the top tube works more like a follower,
with lower Ro, and the whole thing has more natural low thd of the
triode untroubled by too low a load used in a gain stage.
Thus 0.2% at 10Vrms is attainable, instead of 2%.

And you need a lot of parts. Japanese ultrafidelista like the SRPP for
its superb, lively sound. A mu follower is definitely cleaner but if
you dimension the parts wrongly, it can sound clinical, and trial and
error can take a long time. Alan Kimmel published dimensioning formulae
that should be stuck to rather closely.

And in a preamp at 0.1Vrms out, thd = 2H = 0.01% is possible,
and it sounds well.....

One of the finest pre-amps of all time, the Guls design, was SRPP
throughout.

Cathode followers are very misunderstood.
Ask anyone who complains about their sonic signature to tell you
how they work.
Expect a range of dumb answers.
I find that most followers preserve dynamics and clarity
and the fidelity of the gain stages preceeding them.

That's what SRPP do, precisely because it is a follower topology.

Of course, I could argue that LEDs should be acceptable because they
are, specifically, not silicon...

Another reason for the men, mostly middeaged, who build tube amps is
that LED's are just too twee. A film cap is a big fistful of component,
easily soldered in, costs plenty and looks like it. A Kiwame resistor
is a fat green slug looking purposeful. A battery used for bias ditto.
A LED shines like cheap jewellery. If we wanted to be in cheap
itty-bitty parts, we'd build silicon amps.

I don't use leds except to tell me when an amp is turned on,
or if there is a fault, but never in any tube audio circuit, and rarely in any

SS circuit.
I don't like having to use light bulbs to do what leds do far more easilyand
reliably.

Huh? A tube amp lights up its tubes when you switch it on. If the
switch is down and the tubes don't light up, something is broken. I
don't need a LED or a bulb to duplicate the effect.

Andre Jute

"Patrick" said:

An infrared LED will not shine cheap. And most of these take higher currents
than the visible ones.


A quote from the instruction manual:
"The bias of the power tubes is derived from infrared laser leds.
For musical ecstasy, just look into the lens and adjust until totally
blind."

--

"Audio as a serious hobby is going down the tubes."
- Howard Ferstler, 25/4/2005

Adam Stouffer wrote:

I ask this question because earlier in the evening I wired up a single
triode gain stage based on a 5610.

http://www.nj7p.org/Tube1.php?tube=5610

With a 5.6k plate resistor and 130 B+ the plate sees 78v at 10ma. The
cathode resistor was 220r and grid leak 470k. A 1v input gave 5v out
with a nice sinewave. Adding a 220uf or 1000uf cap on the cathode
resistor caused a noticable change on the bottom peaks, of course the
gain was higher at 7v out.

Using a generic red LED on the cathode the output was again 7v and plate
current was now 12ma. The output looked much cleaner and had little
change from 20hz to over 100kz.

Other than the stigma of using silicon in the signal path why isn't this
used in more designs?

Adam

Did anyone mention that LEDs, like most diodes are a non-linear device? Not
a good choice in a circuit where the signal level could be a significant
part of the resultng bias. I still like R's, with & without a paralleled C!
But the diode might cancel some distortion at certain levels. Not something
I'd want to depend on!

My 2 cents. And I sold millions of those LEDs & Schottkys, so I'm BIASED, I
Guess!!!!!!!!!!!!!

HO, HO, John Stewart

Andre Jute wrote:

Patrick Turner wrote:
Andre Jute wrote:
There's a lot of hypocrisy in high end tubes, but you can make yourself
unpopular by pointing out that an SRPP is not a constant current
device, that a cathode follower adds negative feedback, that a DHT
works so well because it has negative feedback built in, that trioded
pentodes sound brilliant, that DHT in PP can sound better than in SE,
and so on, a whole testament full of articles of faith contrary to
known electronic theory and the experience of the open-minded.

I don't see what the fuss is about with diodes or batteries;
maybe they don't spoil music, but surely neither
improves it.

Fixed bias belongs in output stages in PP amps, because the
tube currents become full of 2H beyond the class A working
threshold to class AB.
In totally class A circuits, battery, diode, RC cathode bias, or plain fixed
grid bias
all should sound IDENTICAL if its done right.

The ultrafidelista fancy they can hear better bass with fixed bias,
including battery bias in Class A circuits. Certainly, before you use
battery bias, you'd better see to the top end extension in any DHT amp,
or the sound balance will be odd.

But RC cathode bias is the only one giving regulation of the
bias current because at DC there is some negative current FB.

Another NFB I missed out in my list above... (When the ultrafidelista
say "negative feedback" with a sneer, they mean only user-added
feedback. Feedback intrinsic to favourite tubes and topologies is like
the virginity of Mary Magdalen, very doubtful but never mentioned.)

SRPP sure doesn't include any CCS loading for any part of the two
triodes used; SRPP is where the Rk of the bottom tube = the Rk
of the top tube, and without any load connected to the top tube cathode,
the load seen by the bottom is merely ( µ +1 ) x Rk, so that if µ = 20,
and Rk = 1.5k like an SRPP with 6SN7, then the bottom tube sees
a load of only 31.5k, which will give more 2h than with a 47k resistor
to deliver the DC to the bottom tube.

SRPP is more flexible than just Rkbottom equal Rktop; you can vary them
for various effects but it makes calculation more complex. See the MIT
Radiation Laboratory Series, Vol 18: Vacuum Tube Amplifiers; they call
SRPP a "two tube series arrangement"; cf p456, and check also pp 464
and 438 to be able to put together a coherent formula. I did this in a
set of posts a long time ago and don't now have time to do it again.

SRPP's purpose for the general neddy public is lowering Zo and
maximizing gain, but within that you can tune for max gain or max
current. I put all this in a spreadsheet about ten or twelve years ago,
so I no longer have the formula at my fingertips, which was of course
the point of building a model and a spreadsheet.

But due to the load, if connected, and the 2H of the top tube, some
cancelations of 2H occur,, and lowish thd measures result, although 3H is
slightly more.
Ro is usually only about 1/4 of Ra for one of the triodes.

Better of course is to have a µ follower, with the upper Rk about
10k, and then the bottom tube sees a **much ** higher load
approaching CCS, and the top tube works more like a follower,
with lower Ro, and the whole thing has more natural low thd of the
triode untroubled by too low a load used in a gain stage.
Thus 0.2% at 10Vrms is attainable, instead of 2%.

And you need a lot of parts. Japanese ultrafidelista like the SRPP for
its superb, lively sound. A mu follower is definitely cleaner but if
you dimension the parts wrongly, it can sound clinical, and trial and
error can take a long time. Alan Kimmel published dimensioning formulae
that should be stuck to rather closely.

And in a preamp at 0.1Vrms out, thd = 2H = 0.01% is possible,
and it sounds well.....

One of the finest pre-amps of all time, the Guls design, was SRPP
throughout.

Cathode followers are very misunderstood.
Ask anyone who complains about their sonic signature to tell you
how they work.
Expect a range of dumb answers.
I find that most followers preserve dynamics and clarity
and the fidelity of the gain stages preceeding them.

That's what SRPP do, precisely because it is a follower topology.

But it isn't really much of a follower where Rk top = Rk bottom.
So the Ra is only reduced to about 1/4 Ra, so that buffering is minimised,
ie, the load chacteristics affect the gain stage considerably.
The load connection to the cathode of the top triode has quite an effect on the gain
of the bottom triode.
One can alter the value of the cap coupled RL to make sure nearly equal signal
currents
flow in upper and lower triodes. ( the signal currents cannot ever become exactly
equal.)
When closest to equal, the maximum push pull effect is occuring, and is the function

of current regulation.
Usually the value of load to give maximum PP action is quite low, and each tube is
loaded
down to have a load that is way too low, so the SRPP is not normally used to
get PP action, since loads actually used like a 100k or 50k volume pot
cause hardly any PP action and almost no 2H reduction due to 2H current cancelation
between the bottom and top triodes.
But it really doesn't matter much in preamps what amount of PP action is present
because thd is usually well below any level that will ruin the sound.
Although topology is important, so that if you have a CD signal feeding a
SRPP stage, ( or any other gain stage ) and the volume control after the
this gain stage then thd will be much higher than if the VC is before the gain
stage.
0.5V input becomes 7Vrms output where gain is 14, then that is reduced with the VC,
so imho, stages like this should never have gain 4, so I use a shunt FB network
applying about 10 dB of NFB around a medium µ triode stage which may be
SRPP, µfoll, whatever.

Recently I revised a guy's preamp which had been
volume pot, µ follower, output, to
gain triode, SS CCS load for DC, shunt FB loop with switchable hi-cut shelving
in 3 x 2dB steps, then volume pot, then CF output triode.
He said the sound improved.

Another guy will be forced to change his kit he bought on the Internet
which uses 6N1P tubes to have gain tube, shunt FB, volume pot and CF
because with the gain at 16 times at it is now, he can hardly use it,
there is far too much gain since he only needs about
0.05Vrms average level to his VTL 100 watt power amps and sensitive speakers.
He really needs no gain at all, so taking the input signal to a volume pot,
buffering with a CF would be all he ever would need, and its what i use most days,
and methinks that sounds better than a straight 50k volume pot with
shielded cable off the wiper to the power amps.
The CF allows the preamp to be alngside the listening seat with a CD player,
with a long interconnect cable to the power amps over near the speakers,
so that the sources are well away from speakers and magnetic fields of amps etc,
and a remote control isn't required.
Usually divorces are caused by this arrangement, but it saves someone
getting up out of the chair and walking to change levels 100 times a night.
There are negligible cable losses after a CF buffer.

CF are examples of lots of series voltage NFB, so what,
to my ears they are quite blameless; the prejudice against them
is usually quite irrational.

I lent my own preamp with mainly µ follower stages within to a guy while I changed
his
preamp to similar. During a visit after a week I found he'd had the tone controls of
my amp
switched in all week. Switched in or out did not alter the sound one iota, unless
a boost/cut position was selected.
Pots used were cheap generic carbon track types.
The topology is µ follower plus a Baxandal shunt FB network that gives
an effective anode follower with gain just less than 1.
Its instances like this where you **cannot** tell if the extra stage is switched in
or not
that notions that " tone control stages always ruin the sound"
or that "FB is always detrimental" and so on are found to be notions that
definately just not universally true at all.

In the case of the µ follower, the gain tube is operating at
maximum natural fidelity due to the CCS loading which
maximises gain and reduces thd to a mimimum without
any reliance on cancelling 2H currents.
The CF top tube buffer is acting more like a real cathode follower,
or direct coupled cathode follower, so it can better buffer the gain stage.

In the preamp I modded last week the gain triode has a transistor
CCS, so no thd is generated by a resistor bringing DC to the triode,
and the only load the triode sees is the cap coupled 100k DACT gain attenuator
plus FB R.
Gain of the 1/2 6CG7 used is reduced with a shunt FB network with
47k input R and 270k FB R, so the load the triode sees is 100k and 270k in parallel,

which is a load many times Ra, so thd is low even without the FB reduction
of about 10 dB. Ro is about 1k with such a shunt FB arrangement,
and the load is stable, just the pot and FB R.
The other 1/2 of the 6CG7 is athe cathode follower CF buffer to
isolate the highish Ro of the pot from cable and power amp capacitances.
thd is less than 0.02% in normal use, and talk about thd affecting sound quality
is silly, because the thd is too low to be heard, ( just about all 2H ) thus the
mysterious sound
changes between brands of 6CG7 can be enjoyed or argued about
when tube changes are made.
Some sound better than others of course, but then it depends which sound best
with what power amps and speakers and TT one uses, so the
list of variables grows, and they all interact, so there are NO universal
lessons about what is the best sounding tube or topology to be learned.

The only truth that exists is the truth you establish with thine own ears and
application of thine own knowledge.
Trials of whatever tubes one buys can indeed be tedious and time consuming,
but for the last 80 years its been one of many ways audiophiles optimise their
music quality.

There are no strictly correct adices to be given, but there are mountains of
anecdotal evidence to suggest tube brand A is better than tube brand B,
or that topology A is better than B, but alas tube A in topology B
may sound different to tube B in topology A, and which is preferable
depends on listener X preferences which may be different to listener Y preferences.

On and on the arguments swing, maybe I am damned whatever I choose to do,
but if i took all the arguments all seriously I could make only nothing,
my hands would be paralysed by a brain in conflict!

I just build 'em.


Of course, I could argue that LEDs should be acceptable because they
are, specifically, not silicon...

Another reason for the men, mostly middeaged, who build tube amps is
that LED's are just too twee. A film cap is a big fistful of component,
easily soldered in, costs plenty and looks like it. A Kiwame resistor
is a fat green slug looking purposeful. A battery used for bias ditto.
A LED shines like cheap jewellery. If we wanted to be in cheap
itty-bitty parts, we'd build silicon amps.

I don't use leds except to tell me when an amp is turned on,
or if there is a fault, but never in any tube audio circuit, and rarely in any

SS circuit.
I don't like having to use light bulbs to do what leds do far more easily and
reliably.

Huh? A tube amp lights up its tubes when you switch it on. If the
switch is down and the tubes don't light up, something is broken. I
don't need a LED or a bulb to duplicate the effect.

True, but I have customers who forget to turn off their amps; there isn't
any blue eye or red eye saying "i'm still on" late at night.
And these are sober ppl, the drunks always forget....
And customers sometimes *insist* they need that blue led.....

OK, I just keep 'em happy.

Patrick Turner.







Andre Jute