[Dave]

Is there any quick "in a nutshell" explanation why there aren't more pentode
amp designs? I'm guessing it has to do with extreme non-linearity. I am
building a 10W EL84-based ultralinear amp. The triode conversion is a snap
but it drops me down to 5W which is not enough oomph for my 88dB speakers.
Pentode mode would give me 15W which I am inclined to try if it's not too
difficult.

Can anybody point me to a resource where I could figure out the proper
screen voltage required for pentode mode EL84's with an 8K output tranny and
a B+ of 300V? does this voltage have to be regulated, or could I just drop
the B+ with a resistor?

Thx

Dave

[Phil Allison]

"Dave"

Is there any quick "in a nutshell" explanation why there aren't more
pentode amp designs?


** There are in fact thousands of them.


I'm guessing it has to do with extreme non-linearity.


** False assumption.


I am building a 10W EL84-based ultralinear amp. The triode conversion is
a snap but it drops me down to 5W which is not enough oomph for my 88dB
speakers. Pentode mode would give me 15W which I am inclined to try if
it's not too difficult.


** You want "ultralinear " or "pentode" ??

Make up your mind.


Can anybody point me to a resource where I could figure out the proper
screen voltage required for pentode mode EL84's with an 8K output tranny
and a B+ of 300V?


** Errrr - how about any tube data book ??

Is that too way out for you ??



does this voltage have to be regulated, or could I just drop the B+ with a
resistor?


** The screen of an EL84 is happiest at 300 volts - the plate voltage can
go a bit higher.

10% regulation of the screen supply voltage is OK, must be free of ripple
but.




....... Phil

[Eeyore]

Dave wrote:

Is there any quick "in a nutshell" explanation why there aren't more pentode
amp designs?

Eh ? Why do you think there's a lack of them ?

Graham

[Ian Iveson]

Dave wrote:

Is there any quick "in a nutshell" explanation why there
aren't more pentode amp designs? I'm guessing it has to
do with extreme non-linearity. I am building a 10W
EL84-based ultralinear amp. The triode conversion is a
snap but it drops me down to 5W which is not enough oomph
for my 88dB speakers. Pentode mode would give me 15W which
I am inclined to try if it's not too difficult.

Harder to get right for hi-fi. Not fashionable. If you're
going to go the route of increasing gain and feedback you
may as well use transistors. That kind of thing. Not a
nutshell...the devil is in the detail as far as getting it
right is concerned.

Can anybody point me to a resource where I could figure
out the proper screen voltage required for pentode mode
EL84's with an 8K output tranny and a B+ of 300V? does
this voltage have to be regulated, or could I just drop
the B+ with a resistor?

http://www.duncanamps.com/

For datasheets and to download TDSL and the power supply
designer while you're there.

I would keep the screen as near as possible to 300V. It is
common to use a choke and cap to filter the screen supply.
You may wish to regulate coz it's cheap (but less reliable
maybe) and quiet. IIRC there is at least one regulated
screen supply detailed in Duncan's guitar amp designs. There
must be loads of designs kicking around.

You should check that the pentode amp is reliably stable.

It would be interesting to know whether you get the oomph
you are looking for, so please report back when you've done
it.

The big differences are that the pentode has more gain and a
much higher output resistance. Given the same transformer,
the open-loop bandwidth of the output stage is reduced,
particularly at the bottom end. You use the extra gain to
provide more feedback, which restores the bandwidth and
reduces the consequent distortion. It would be nice if the
result could end up the same, but with more power. At the
frequency extremes, however, so much gain is lost by the
coupling of the pentodes to the output transformer that
there is not much feedback to reduce the distortion. If
that's where your oomph is, you may not be satisfied with
the result. Same for the top end at full power, maybe.

You need more Watts, probably. For proper slam you also need
a lower OPT turns ratio, requiring valves in parallel.

cheers, Ian

[Engineer[_2_]]

On Oct 31, 10:41 am, "Dave" wrote:
Is there any quick "in a nutshell" explanation why there aren't more pentode
amp designs? I'm guessing it has to do with extreme non-linearity. I am
building a 10W EL84-based ultralinear amp. The triode conversion is a snap
but it drops me down to 5W which is not enough oomph for my 88dB speakers.
Pentode mode would give me 15W which I am inclined to try if it's not too
difficult.

Can anybody point me to a resource where I could figure out the proper
screen voltage required for pentode mode EL84's with an 8K output tranny and
a B+ of 300V? does this voltage have to be regulated, or could I just drop
the B+ with a resistor?

Thx

Dave

There are many...
Try this:
http://www.bonavolta.ch/hobby/en/audio/el84.htm
I like the "Mullard 5-10". I built two of them decades ago....
Cheers,
Roger

[Andre Jute]

On Oct 31, 2:41 pm, "Dave" wrote:
Is there any quick "in a nutshell" explanation why there aren't more pentode
amp designs? I'm guessing it has to do with extreme non-linearity. I am
building a 10W EL84-based ultralinear amp. The triode conversion is a snap
but it drops me down to 5W which is not enough oomph for my 88dB speakers.
Pentode mode would give me 15W which I am inclined to try if it's not too
difficult.

Can anybody point me to a resource where I could figure out the proper
screen voltage required for pentode mode EL84's with an 8K output tranny and
a B+ of 300V? does this voltage have to be regulated, or could I just drop
the B+ with a resistor?

Thx

Dave

I wasn't aware of a shortage of pentode amps. But a pentode amp has
more gain than those with ultralinear connection, and thus can have
more global feedback applied, and thus lower THD.

However, I put a pentode/ultralinear/triode switch and a variable/off
negative feedback switch in the best amp I ever designed, my T113
"Triple Threat", Class A1 PP EL34 of pretty straightforward,
conservative design. Then I tested it with chamber musicians
listerning to their own recordings. They universally hated the pentode
sound, and loved the triode sound, on which the amp is usually played
with zero NFB.

The explanation is simple. Loop NFB causes artifacts of ever lower
magnitude but higher order. These high order distortions, even at 60dB
below conscious perception are very, very disturbing, whereas second
harmonic up to three-quarter per cent cannot even be distinguished by
professional musicians. It is a subliminal effect, and of course in
pentode mode it is exaggerated. People look at the total harmonic
distortion but in fact the higher harmonics must be weighted much more
heavily than the second harmonic to account for its extraordinary
subliminal effect, often described as "edgineess" by professional
musicians.

You listen to your speakers, not your amp; I would really advise you,
unless you are married to these insensitive speakers, first to get
good sensitive speakers you can live with forever, then build an amp
to suit them. If you have the space, an inexpensive sensitive speaker
you can build easily is The Impresario on my netsite, url under my
sig. EL84 are particularly sweet in triode but of course pitifully
underpowered for insensitive speakers.

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]

Dave wrote:

Is there any quick "in a nutshell" explanation why there aren't more pentode
amp designs? I'm guessing it has to do with extreme non-linearity. I am
building a 10W EL84-based ultralinear amp. The triode conversion is a snap
but it drops me down to 5W which is not enough oomph for my 88dB speakers.
Pentode mode would give me 15W which I am inclined to try if it's not too
difficult.

Despite the low efficiency of triodes, and reduction of power,
pentodes are not seen as better sounding.

But the vast majority of amplifiers made commercially have been
in either pentode or low% tap UL, with cathode bias and where the UL AB
power is close to the pentode PO.

That there are mostly triode amps out there is a complete myth.


Can anybody point me to a resource where I could figure out the proper
screen voltage required for pentode mode EL84's with an 8K output tranny and
a B+ of 300V? does this voltage have to be regulated, or could I just drop
the B+ with a resistor?

Screen voltage supplies are not critical with Ea 350V for EL84.

So for Ea = 300V, Eg2 can be anywhere between +270 and +300V.

The use of +270V probably is easier to find by means of a resistance
from the B+, and the Eg2 should be shunted with 47uF to 0V.

For best operation for full PO, fixed bias is used, and the anode supply
is low impedance,
and the Eg2 is regulated.

Patrick Turner.




Thx

Dave

[Dave]

"Andre Jute" wrote in message
ps.com...

However, I put a pentode/ultralinear/triode switch and a variable/off
negative feedback switch in the best amp I ever designed, my T113
"Triple Threat", Class A1 PP EL34 of pretty straightforward,
conservative design. Then I tested it with chamber musicians
listerning to their own recordings. They universally hated the pentode
sound, and loved the triode sound, on which the amp is usually played
with zero NFB.

The explanation is simple. Loop NFB causes artifacts of ever lower
magnitude but higher order. These high order distortions, even at 60dB
below conscious perception are very, very disturbing, whereas second
harmonic up to three-quarter per cent cannot even be distinguished by
professional musicians. It is a subliminal effect, and of course in
pentode mode it is exaggerated. People look at the total harmonic
distortion but in fact the higher harmonics must be weighted much more
heavily than the second harmonic to account for its extraordinary
subliminal effect, often described as "edgineess" by professional
musicians.


This was the explanation I was looking for. thank you.

You listen to your speakers, not your amp; I would really advise you,
unless you are married to these insensitive speakers, first to get
good sensitive speakers you can live with forever, then build an amp
to suit them. If you have the space, an inexpensive sensitive speaker
you can build easily is The Impresario on my netsite, url under my
sig. EL84 are particularly sweet in triode but of course pitifully
underpowered for insensitive speakers.


Yes, I know, efficient speakers are on my long-term shopping list. My
current speakers will run nicely on 15W (or even 12W). I have somewhat
limited financial means at my disposal for projects of this sort, but I do
like the look of the EL34 just because of its' high power output.

What's an acceptable value for distortion? Right now I'm running a solid
state amp which is rated at 50W RMS with THD of 0.08%. I guess this isn't a
reasonable target for a tube amp unless I run multiple parallel UL PP pairs
per channel...

I guess it depends on what you consider inexpensive. The Lowther drivers
run $600USD for the rock-bottom lowest-priced pair to $2K for the "A"
series. Your cabinet and tweeter placement look an awful lot like the
Hammer Dynamics Super-12, whose drivers run only $650USD for the pair and
include tweeters. Is there a recommended 6 kHz crossover schematic for your
Impresario?

Thanks for the info. I like the look of your amp, it doesn't get much
simpler than 3 tubes. I've already got the vast majority of what I'd need
to build it (3K output transformers and 230V isolation x-former).
Hmmmm.....

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

[Dave]

"Dave" wrote in message
news:MznWi.54019$GO5.44974@edtnps90...


I guess it depends on what you consider inexpensive. The Lowther drivers
run $600USD for the rock-bottom lowest-priced pair to $2K for the "A"
series. Your cabinet and tweeter placement look an awful lot like the
Hammer Dynamics Super-12, whose drivers run only $650USD for the pair and
include tweeters. Is there a recommended 6 kHz crossover schematic for
your Impresario?

Aaaahhhhh, I see, you are not specifying the PM6A for this speaker, the 12LT
is only $60! I have ordered one.

[Andre Jute]

Dave wrote:

"Andre Jute" wrote in message
ps.com...

However, I put a pentode/ultralinear/triode switch and a variable/off
negative feedback switch in the best amp I ever designed, my T113
"Triple Threat", Class A1 PP EL34 of pretty straightforward,
conservative design. Then I tested it with chamber musicians
listerning to their own recordings. They universally hated the pentode
sound, and loved the triode sound, on which the amp is usually played
with zero NFB.

The explanation is simple. Loop NFB causes artifacts of ever lower
magnitude but higher order. These high order distortions, even at 60dB
below conscious perception are very, very disturbing, whereas second
harmonic up to three-quarter per cent cannot even be distinguished by
professional musicians. It is a subliminal effect, and of course in
pentode mode it is exaggerated. People look at the total harmonic
distortion but in fact the higher harmonics must be weighted much more
heavily than the second harmonic to account for its extraordinary
subliminal effect, often described as "edgineess" by professional
musicians.


This was the explanation I was looking for. thank you.

You listen to your speakers, not your amp; I would really advise you,
unless you are married to these insensitive speakers, first to get
good sensitive speakers you can live with forever, then build an amp
to suit them. If you have the space, an inexpensive sensitive speaker
you can build easily is The Impresario on my netsite, url under my
sig. EL84 are particularly sweet in triode but of course pitifully
underpowered for insensitive speakers.


Yes, I know, efficient speakers are on my long-term shopping list. My
current speakers will run nicely on 15W (or even 12W). I have somewhat
limited financial means at my disposal for projects of this sort, but I do
like the look of the EL34 just because of its' high power output.

What's an acceptable value for distortion?

Doesn't work like that. All my amps are in fact very silent by
ultrafidelista standards; that essentially means zero negative
feedback or very little, that is, NFB is not thrown at the amp in big
gobs to get the numbers down willynilly. What you really want to know
after the amp is reasonably silent is what the proportion of third and
higher residual harmonics is to the residual second harmonic. You're
trying to reduce this to a low proportion.

Right now I'm running a solid
state amp which is rated at 50W RMS with THD of 0.08%. I guess this isn't a
reasonable target for a tube amp unless I run multiple parallel UL PP pairs
per channel...

And NFB up the kazoo. 0.1 per cent is a figure commonly bandied about
but I'm not overly impressed. I know for a fact that people cannot
hear as much as 3/4 per cent second harmonic but can hear that much
odd harmonic very clearly, and many people can hear or are made
uncomfortable by 0.3 per cent odd harmonics. What you want to do is
not to measure at full power but at some lower power where you will
actually use the amp; 1W or 2.83V into 8ohm is a common level with
even semi-sensitive speakers.

I guess it depends on what you consider inexpensive. The Lowther drivers
run $600USD for the rock-bottom lowest-priced pair to $2K for the "A"
series.

The cheap speaker I refer to is not the Lowther PM6A (that's pretty
pricey) fitted to my bicor horns but the Eminence guitar driver in
The Impresario. You have to be a little careful on my netsite; the
projects are not grouped by cost, and insanely expensive projects for
plutocratic audiophiles rub shoulders with projects I designed to fit
the budgets of students...

Your cabinet and tweeter placement look an awful lot like the
Hammer Dynamics Super-12, whose drivers run only $650USD for the pair and
include tweeters. Is there a recommended 6 kHz crossover schematic for your
Impresario?

The Eminence Beta 12 LT was such a strikingly suitable speaker for
this sort of work, several people designed around it when it came out.
My design was first, then the Afterburner by Thorsten Loesch, then
John Wyckoff's Hammer Dynamics design. My Impresario and Thorsten's
Afterburner are designed to use the standard driver; the Hammer
Dynamics design has a specially stiffened cast basket and also comes
with the tweeter and maybe components for a crossover.

It is not surprising at all that the three competent designs (there
are also some shoddy designs around this driver out there) all look
very much alike; such a speaker is designed around its Thiele-Small
parameters and to a particular purpose (audio range) and there is thus
little scope for individuality: enclosures for that driver by
competent designers will all appear to be near-clones of each other.
You can build any of the three designs with pretty much the same
results. My design is distinguished by having the box internal
relationships in the Golden Mean of classical architecture, a formula
often observed in nature too, the mathematical relationship Phi
(0.618:1.00:1.618); you may view it as little more than a fancy of
mine or as a fundamental stroke of genius. Thorsten said once, IIRC,
that his speaker was designed by the width of some boards he had which
were approximately right. I imagine the Wyckoff design is nudged a
little for sawing economy from one sheet of ply. Such are the audio-
imperatives of great and famous designs! I didn't have the privilege
of knowing Wyckoff (whose design is now marketed by his widow) but I
have no hesitation in asserting that he shared with Thorsten and me an
essential characteristic of a good audio designer (in fact a good
designer of anything): wu, or the luck of an artist. We just do things
right, often not for reasons that rational people will believe if told
the truth.

It is years since any of The Impresario I built had the tweeter
connected. Most people cannot hear that the tweeter is disconnected,
and the sound does get a certain greater clarity, deeper apparent
bass, better balance, and all-round greater calm authority with only
the actual point source 12LT "fullranger" in use. If you're under
thirty you may want the tweeter connected, over thirty (as all the
other owners are) forget it. The tweeter should be chosen to cross
over naturally or with just a cap.

****
In a later post Dave wrote:

Aaaahhhhh, I see, you are not specifying the PM6A for this speaker, the 12LT
is only $60! I have ordered one.

****

Careful that you get the 12LT and not the newer 12LTA which appears to
be slightly different.

Thanks for the info. I like the look of your amp, it doesn't get much
simpler than 3 tubes. I've already got the vast majority of what I'd need
to build it (3K output transformers and 230V isolation x-former).
Hmmmm.....

I imagine you're talking about my Type 201 EL34 SEntry, which is an
amp specifically designed for a student living in digs, and to be
build in a Hammond 17x4x10in ali box, with a 6x2x4 ali box inserted to
which the valves bases are attached along one short edge, with the
tubes inside. The box is intended to stand on one short edge on a
mantelpiece. The thing was designed to be built entirely out of my
junkbox and happily turned out to be another inspired design, sounding
so much better than it deserved if its origins, intentions (it had to
play loud rock as well) and components were any indication. Quite a
few have been built by people who write to me that solid state just
didn't prepare them for this explosion of music. Note that you need
single-ended transformers, not push-pull. A source of cheap SE
transformers is Thordarson Vertical Transformers out of tube-era TV
sets, sometimes available from surplus stores.

Happy to be of service to a hands-on audiophile. Good luck.

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

[Eeyore]

Dave wrote:

What's an acceptable value for distortion? Right now I'm running a solid
state amp which is rated at 50W RMS with THD of 0.08%.

That's pretty high by modern standards. Quite routine, everyday non-exotic
amplification can easily better that by a factor of ten to one and still be
quite inexpensive.

Graham

[Phil Allison]

"Andre Jute"

I wasn't aware of a shortage of pentode amps. But a pentode amp has
more gain than those with ultralinear connection, and thus can have
more global feedback applied, and thus lower THD.

** Not true at all.

http://www.tube-amps.net/images/Data...llard_1964.pdf

The drive levels for pentode PP and UL are almost he same, UL having the
edge in sensitivity.


The explanation is simple. Loop NFB causes artifacts of ever lower
magnitude but higher order.


** Not generally true.

Only true when the OP stage has high levels of gain non-linearity and lotsa
2nd harmonic as a result.

When the stage has good gain linearity ( say 2% or better ) there is no
such effect.


These high order distortions, even at 60dB
below conscious perception are very, very disturbing, whereas second
harmonic up to three-quarter per cent cannot even be distinguished by
professional musicians.


** Totally misleading.

You CANNOT have 2nd harmonic distortion without the accompanying
intermodulation distortion.

It is the IM that adds audible " mud " to the sound of music, not the
production of musical sounding 2nd harmonics.


NB:

Here is a recent design for an EL84 UL amplifier.

http://www.tube-amps.net/EA_Hashimoto_EL84_PP_01.htm

http://www.tube-amps.net/images/EL84_PP_01/EL84_HD.jpg


THD is well below 0.1 % for most of the power range AND the residual is
predominantly 2nd harmonic.

UL works even with humble EL84s, as advertised.



....... Phil

[Ian Iveson]

Dave wrote:

....

This was the explanation I was looking for. thank you.

LOL.

Ian

[Dave]

"Eeyore" wrote in message
...


Dave wrote:

What's an acceptable value for distortion? Right now I'm running a solid
state amp which is rated at 50W RMS with THD of 0.08%.

That's pretty high by modern standards. Quite routine, everyday non-exotic
amplification can easily better that by a factor of ten to one and still
be
quite inexpensive.

It's a decidedly non-modern amp.

[Dave]

"Andre Jute" wrote in message
ups.com...

And NFB up the kazoo. 0.1 per cent is a figure commonly bandied about
but I'm not overly impressed. I know for a fact that people cannot
hear as much as 3/4 per cent second harmonic but can hear that much
odd harmonic very clearly, and many people can hear or are made
uncomfortable by 0.3 per cent odd harmonics. What you want to do is
not to measure at full power but at some lower power where you will
actually use the amp; 1W or 2.83V into 8ohm is a common level with
even semi-sensitive speakers.


So... is ALL negative feedback bad? I am 40 with some background in
electrical engineering and learned nothing but transistor theory at
school... thermal runaway pretty much mandates feedback in all transistor
amps. A lot of what I read about 'audiophile' grade amps is pro-znfb. Why?


The cheap speaker I refer to is not the Lowther PM6A (that's pretty
pricey) fitted to my bicor horns but the Eminence guitar driver in
The Impresario. You have to be a little careful on my netsite; the
projects are not grouped by cost, and insanely expensive projects for
plutocratic audiophiles rub shoulders with projects I designed to fit
the budgets of students...


I'm closer to the student end of things as I have kids entering their
"expensive years" but don't mind slumming the thrift stores and surplus
stores for parts.


Happy to be of service to a hands-on audiophile. Good luck.


Thanks, I may just need some luck.

Dave

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

[Andre Jute]

On Nov 2, 2:12 pm, "Dave" wrote:
"Andre Jute" wrote in message

ups.com...



And NFB up the kazoo. 0.1 per cent is a figure commonly bandied about
but I'm not overly impressed. I know for a fact that people cannot
hear as much as 3/4 per cent second harmonic but can hear that much
odd harmonic very clearly, and many people can hear or are made
uncomfortable by 0.3 per cent odd harmonics. What you want to do is
not to measure at full power but at some lower power where you will
actually use the amp; 1W or 2.83V into 8ohm is a common level with
even semi-sensitive speakers.

So... is ALL negative feedback bad?

There are some who think that NFB is the devil's work, there are some
who claim a little NFB can sometimes be beneficial without doing too
much harm, there are some who order another 10x NFB as a reflex,
though in silicon rather than tube design. I think that the best sound
comes from triodes or triode-tied pentodes operating at a conservative
dissipation in Class A1 with ZNFB. (This is what people generally call
"Single Ended Sound" but what it is, in fact, is Class A1 triode sound
with very little or zero NFB.) There is a tongue-in-cheek article
about NFB here that gives the ultrafidelista view:
http://members.lycos.co.uk/fiultra/K...dre%20Jute.htm
Note however that the horror displayed by the trendies whenever
negative feedback is mentioned is entirely misinformed. There is no
such thing as an amplifieer without negative feedback. Many of the
topologies beloved of the tubie elite are nothing but feedback
mechanisms: SRPP (a push pull totem pole, and not a CCS either),
cathode followers (they work *because* of their NFB), etc. The correct
way to look at NFB is that loop or global feedback is evil but local
feedback in the hands of an expert can do wonders. Of course, the
greatest experts make designs so elegant that they require "zero added
NFB". None of my published amps require any (added) NFB at all for
correct operation.

I am 40

Your age isn't connected to NFB but to your hearing bandwidth. Only a
tiny fraction of people can ever hear up to the so-called "audio
bandwidth" of 20 kiloHertz. As you grow older, even if you don't abuse
your ears with loud music or steam rivetters, your hearing band
declines until at forty you'll be lucky to hear anything over 14kHz.
The reason Eminence's Beta LT works so well is that it is "fullrange"
for the intended audience; I designed The Impresario first for a
middle-aged impresario, and quite a few have gone to similar people,
who don't care overly much for accuracy on a meter but insist on the
nearest possible to a natural sound, as if the singer is in the room
with them. The Beta LT further helps this illusion along, at least in
the enclosure I designed, by naturally balancing the top and bottom
end against each other -- this is without the tweeter now. With the
tweeter, given that the tweeter is selected to cross over as indicated
on the plan and not in the common 3500Hz region, it sounds more
"accurate" but also more uninvolving, more artificial, more silicon-
like.

with some background in
electrical engineering and learned nothing but transistor theory at
school... thermal runaway pretty much mandates feedback in all transistor
amps. A lot of what I read about 'audiophile' grade amps is pro-znfb. Why?

As explained above, a lot of that is caused by street corner gossip
repeated by people quite ignorant of electronics. Loop feedback is
definitely evil, local feedback can be useful to novice designers,
some topologies are nothing but negative feedback devices -- and are
found in the amps of the loudest opponents of NFB!

The cheap speaker I refer to is not the Lowther PM6A (that's pretty
pricey) fitted to my bicor horns but the Eminence guitar driver in
The Impresario. You have to be a little careful on my netsite; the
projects are not grouped by cost, and insanely expensive projects for
plutocratic audiophiles rub shoulders with projects I designed to fit
the budgets of students...

I'm closer to the student end of things as I have kids entering their
"expensive years" but don't mind slumming the thrift stores and surplus
stores for parts.

You'd be smart to start cheap because it is very easy to waste
boutique parts on an amp that doesn't satisfy but is too pretty to
carve up for another try. The few guys on RAT who've actually built
amps from scratch are like your doctor: we're only showing you the
successful cures and you have to find out for yourself about the
failures we buried.

Happy to be of service to a hands-on audiophile. Good luck.

Thanks, I may just need some luck.

Dave

Andre Jute
Our legislators managed to criminalize fox-hunting and smoking; when
will they get off their collective fat backside and criminalize
negative feedback? It is clearly consumed only by thickoes.

[Patrick Turner]

Phil Allison wrote:

"Andre Jute"

I wasn't aware of a shortage of pentode amps. But a pentode amp has
more gain than those with ultralinear connection, and thus can have
more global feedback applied, and thus lower THD.

** Not true at all.


This is news to me.

For a given load, say 5k for a pair of UL EL34, the drive voltage
required
in pentode mode is less than 1/2 compared to triode for a givenm output
power, say 10 watts,
which all modes will produce without clipping or grid current. UL is
somehwere in between
depending on the % of the UL tapping for the screen.


http://www.tube-amps.net/images/Data...llard_1964.pdf

The drive levels for pentode PP and UL are almost he same, UL having the
edge in sensitivity.

I have never seen this, ever. Pentodes are always much easier to drive.



The explanation is simple. Loop NFB causes artifacts of ever lower
magnitude but higher order.

** Not generally true.



Only true when the OP stage has high levels of gain non-linearity and lotsa
2nd harmonic as a result.

Or lotsa any harmonic. If 3H is high ( 10% ), and GNFB amount is low,
say 12dB added higher harmonics appear
where none existed before GFB is connected.

Usually its not an issue, because at listening levels the open loop THD
is low, less than 1%,
and 12dB around a typical PP triode amp will not create serious
secondary products.


When the stage has good gain linearity ( say 2% or better ) there is no
such effect.

Not one that is easily noticable.

These high order distortions, even at 60dB
below conscious perception are very, very disturbing, whereas second
harmonic up to three-quarter per cent cannot even be distinguished by
professional musicians.

** Totally misleading.

You CANNOT have 2nd harmonic distortion without the accompanying
intermodulation distortion.

It is the IM that adds audible " mud " to the sound of music, not the
production of musical sounding 2nd harmonics.

THD is fine, 2% is quite OK, but the resulting IMD is the bad one.
The IMD products are not harmonically related to the music.



NB:

Here is a recent design for an EL84 UL amplifier.

http://www.tube-amps.net/EA_Hashimoto_EL84_PP_01.htm

http://www.tube-amps.net/images/EL84_PP_01/EL84_HD.jpg

THD is well below 0.1 % for most of the power range AND the residual is
predominantly 2nd harmonic.

UL works even with humble EL84s, as advertised.

...... Phil

UL is good, especially class A, with high value loads,
which means only 10 watts from a pair of EL84.

Cathode FB from a tertiary winding and fixed screen bias works better,
like in the Acoustical Quad-II connection.

10% of the primary voltage fed back to the cathodes is enough
to make a pair of output pentodes/beam tetrodes display the Ra of a
triode,
and not be any harder to drive than triode, but retain very near the
full
power of the pentode/beam config.
The CFB gives a greater reduction to effective Ra than plain 43% UL.

Good PP amps don't have as much 2H as the 3H shown in the above URLs
for THD graphs. If the driver stage is poorly designed, and the graphs
above
strongly indicate an accountant designed the amp, they will have used a
poorly
set up input triode and following concertina phase inverter which makes
1% of 2H, more than the 3H apparently is made in the output stage.

Its all reduced 10 times by the 21dB GNFB, so open loop 1% becomes 0.1%
closed loop.

I have been there done all this dozens of times, except that all my PP
amps
have only mainly 3H, and very low levels of it.

All the better known amps with an SE triode input stage and with
decently
set up LTP all have much less 2H than the Japanese thinges above.

Mullard 510, 520, Williamson, McIntosh, leak, Radford all measured
very well, with low 2H.


Patrick Turner.

[Patrick Turner]

Andre Jute wrote:

On Oct 31, 2:41 pm, "Dave" wrote:
Is there any quick "in a nutshell" explanation why there aren't more pentode
amp designs? I'm guessing it has to do with extreme non-linearity. I am
building a 10W EL84-based ultralinear amp. The triode conversion is a snap
but it drops me down to 5W which is not enough oomph for my 88dB speakers.
Pentode mode would give me 15W which I am inclined to try if it's not too
difficult.

Can anybody point me to a resource where I could figure out the proper
screen voltage required for pentode mode EL84's with an 8K output tranny and
a B+ of 300V? does this voltage have to be regulated, or could I just drop
the B+ with a resistor?

Thx

Dave

I wasn't aware of a shortage of pentode amps. But a pentode amp has
more gain than those with ultralinear connection, and thus can have
more global feedback applied, and thus lower THD.

However, I put a pentode/ultralinear/triode switch and a variable/off
negative feedback switch in the best amp I ever designed, my T113
"Triple Threat", Class A1 PP EL34 of pretty straightforward,
conservative design. Then I tested it with chamber musicians
listerning to their own recordings. They universally hated the pentode
sound, and loved the triode sound, on which the amp is usually played
with zero NFB.

The explanation is simple. Loop NFB causes artifacts of ever lower
magnitude but higher order. These high order distortions, even at 60dB
below conscious perception are very, very disturbing, whereas second
harmonic up to three-quarter per cent cannot even be distinguished by
professional musicians. It is a subliminal effect, and of course in
pentode mode it is exaggerated. People look at the total harmonic
distortion but in fact the higher harmonics must be weighted much more
heavily than the second harmonic to account for its extraordinary
subliminal effect, often described as "edgineess" by professional
musicians.

-60dB, at average listeining levels????

Say a volt is average, then -60dB artifcats = 1mV.

1mV of grunge isn't exactly loud in most speakers......


You listen to your speakers, not your amp; I would really advise you,
unless you are married to these insensitive speakers, first to get
good sensitive speakers you can live with forever, then build an amp
to suit them. If you have the space, an inexpensive sensitive speaker
you can build easily is The Impresario on my netsite, url under my
sig. EL84 are particularly sweet in triode but of course pitifully
underpowered for insensitive speakers.

I quite like UL EL84 good for 12 watts AB. But you do need
sensitive speakers over 92dB/W/M.

I'd prefer an SE 300B amp though....

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 CfGWi.1359$8S5.242@edtnps82, "Dave"
wrote:

"Andre Jute" wrote in message
ups.com...

And NFB up the kazoo. 0.1 per cent is a figure commonly bandied about
but I'm not overly impressed. I know for a fact that people cannot
hear as much as 3/4 per cent second harmonic but can hear that much
odd harmonic very clearly, and many people can hear or are made
uncomfortable by 0.3 per cent odd harmonics. What you want to do is
not to measure at full power but at some lower power where you will
actually use the amp; 1W or 2.83V into 8ohm is a common level with
even semi-sensitive speakers.


So... is ALL negative feedback bad? I am 40 with some background in
electrical engineering and learned nothing but transistor theory at
school... thermal runaway pretty much mandates feedback in all transistor
amps.

There are typically two different feedback mechanisms at work in a
transistor amplifier. The first, which is the one you are referring to
stabilizes the DC bias, or operating point, of the transistors to
prevent thermal runaway as you describe, and also to insure that the
"crossover" region of the transfer characteristic has the desired degree
of overlap. The second is the overall loop NFB, which is what Andre is
talking about, and is used to reduce the audio frequency distortion and
noise produced in the amplifier, as well as to stabilize the gain of the
amplifier, as the gain of discrete transistors is notoriously
inconsistent from sample to sample.

A lot of what I read about 'audiophile' grade amps is pro-znfb. Why?

It's because of a lack of understanding in the audiophile community.
NFB can work wonders when correctly applied by someone who knows what
they are doing, but it can also do serious harm when incompetently
applied. Bad experiences with amplifiers using NFB of the latter type
lead to the typical audiophile' negative feelings towards NFB.


Regards,

John Byrns

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

[Arny Krueger]

"Dave" wrote in message
news:CfGWi.1359$8S5.242@edtnps82...

"Andre Jute" wrote in message
ups.com...

And NFB up the kazoo. 0.1 per cent is a figure commonly bandied about
but I'm not overly impressed.

At 0.1% THD it is not hard to set up a listening test where the difference
is audible.

I know for a fact that people cannot
hear as much as 3/4 per cent second harmonic but can hear that much
odd harmonic very clearly, and many people can hear or are made
uncomfortable by 0.3 per cent odd harmonics.

This order of harmonics issue is way overblown. Doesn't matter whether it is
odd or even order nonlinearity, it all makes IM. IM is generally aharmonic
and it all sounds bad.

What you want to do is
not to measure at full power but at some lower power where you will
actually use the amp; 1W or 2.83V into 8ohm is a common level with
even semi-sensitive speakers.

Agreed.

So... is ALL negative feedback bad?

Of course not. If negative feedback was as bad as some make out, the modern
world pretty much wouldn't work - inverse feedback is that pervasive. NFB
makes both the space shuttle and your car go. Your home heating system no
doubt implements it. Thing is, you have to understand calculus to really do
NFB right unless you're cook-booking someone else's design. There really
arne;t that many people who passed 4 semesters of calculus, and far fewer
who remember it well enough when they get old enough to want to waste their
time building power amps.

I am 40 with some background in electrical engineering and learned
nothing but transistor theory at school... thermal runaway pretty much
mandates feedback in all transistor amps.

That's DC inverse feedback, not AC inverse feedback which is mostly what we
talk about in audio. There have been designers who were light on the topic
of bias stability (synonum for DC inverse feedback), and there were even a
few house fires as a result.

A lot of what I read about 'audiophile' grade amps is pro-znfb. Why?

ZNFB paranoia sells amps to people who are technically ignorant but think
they are sophisticated. That describes most of the high end audiophile
market.

[Patrick Turner]

Dave wrote:

"Andre Jute" wrote in message
ups.com...

And NFB up the kazoo. 0.1 per cent is a figure commonly bandied about
but I'm not overly impressed. I know for a fact that people cannot
hear as much as 3/4 per cent second harmonic but can hear that much
odd harmonic very clearly, and many people can hear or are made
uncomfortable by 0.3 per cent odd harmonics. What you want to do is
not to measure at full power but at some lower power where you will
actually use the amp; 1W or 2.83V into 8ohm is a common level with
even semi-sensitive speakers.


So... is ALL negative feedback bad? I am 40 with some background in
electrical engineering and learned nothing but transistor theory at
school... thermal runaway pretty much mandates feedback in all transistor
amps. A lot of what I read about 'audiophile' grade amps is pro-znfb. Why?

Maybe you should listen to a few amps without GNFB loops to find out
why,
if that is at all possible.

Audiophiles have this quaint idea that engineers should
be able to just hook up devices without global NFB loops between output
and input,
and N&D will be low, BW will remain wide, Rout will be low, and channel
gain will be the same.

Audiophiles think one never should have to connect
correctional circuitry which is sorcery and trickery used to cover up
mistakes. A bandaid solution,
a fudge, wedge, trick, con, etc.
They think the NFB cannot "ever act quick enough" to fix mistakes
so the circuits must be designed to give hi-fi without GNFB.


But this limits them to triode amps. Triodes have lots of internal NFB
anyway.
But this triode NFB is local, internal, and electrostatical,
and natural, ( an unavoidable ), so its OK.


All other devices are entirely unusable without a considerable amount of
correctional FB in each stage, or around two or more stages.

Audiophiles are often technically illiterate and one eyed obsesive
personalities.
And inconsistent. They'll use a technical argument to back up some claim
about
"the sound" but often have not the slightest idea about building even
the most
simple amplifier.
The won't sit down for serious AB testing, oh no, eg, they don't like it
when they can't
tell the difference between a good SS amp and a tube amp.
Or tell the difference between capacitors, resistors, cables and lotsa
other things better than random
chance would indicate, ie, 50% right most of the time.

While we would be foolish to dismiss what they hear, we should always
keep an open mind
about what they say, and when you get a lot of them together,
they will have opposite ideas about "the sound", so who do you believe?
If they are all more or less happy with something you've made,
is certainly a success.

Patrick Turner.




The cheap speaker I refer to is not the Lowther PM6A (that's pretty
pricey) fitted to my bicor horns but the Eminence guitar driver in
The Impresario. You have to be a little careful on my netsite; the
projects are not grouped by cost, and insanely expensive projects for
plutocratic audiophiles rub shoulders with projects I designed to fit
the budgets of students...


I'm closer to the student end of things as I have kids entering their
"expensive years" but don't mind slumming the thrift stores and surplus
stores for parts.


Happy to be of service to a hands-on audiophile. Good luck.


Thanks, I may just need some luck.

Dave

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 ,
"Arny Krueger" wrote:

"Dave" wrote in message
news:CfGWi.1359$8S5.242@edtnps82...

"Andre Jute" wrote in message
ups.com...

I know for a fact that people cannot
hear as much as 3/4 per cent second harmonic but can hear that much
odd harmonic very clearly, and many people can hear or are made
uncomfortable by 0.3 per cent odd harmonics.

This order of harmonics issue is way overblown. Doesn't matter whether it is
odd or even order nonlinearity, it all makes IM. IM is generally aharmonic
and it all sounds bad.

But even order IM is much more obvious and annoying in listening tests,
although I suppose others may find odd order IM more annoying, its
probably partly a matter of personal preference.


Regards,

John Byrns

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

[Andre Jute]

On Nov 2, 4:28 pm, Patrick Turner wrote:
Andre Jute wrote:

On Oct 31, 2:41 pm, "Dave" wrote:
Is there any quick "in a nutshell" explanation why there aren't more pentode
amp designs? I'm guessing it has to do with extreme non-linearity. I am
building a 10W EL84-based ultralinear amp. The triode conversion is a snap
but it drops me down to 5W which is not enough oomph for my 88dB speakers.
Pentode mode would give me 15W which I am inclined to try if it's not too
difficult.

Can anybody point me to a resource where I could figure out the proper
screen voltage required for pentode mode EL84's with an 8K output tranny and
a B+ of 300V? does this voltage have to be regulated, or could I just drop
the B+ with a resistor?

Thx

Dave

I wasn't aware of a shortage of pentode amps. But a pentode amp has
more gain than those with ultralinear connection, and thus can have
more global feedback applied, and thus lower THD.

However, I put a pentode/ultralinear/triode switch and a variable/off
negative feedback switch in the best amp I ever designed, my T113
"Triple Threat", Class A1 PP EL34 of pretty straightforward,
conservative design. Then I tested it with chamber musicians
listerning to their own recordings. They universally hated the pentode
sound, and loved the triode sound, on which the amp is usually played
with zero NFB.

The explanation is simple. Loop NFB causes artifacts of ever lower
magnitude but higher order. These high order distortions, even at 60dB
below conscious perception are very, very disturbing, whereas second
harmonic up to three-quarter per cent cannot even be distinguished by
professional musicians. It is a subliminal effect, and of course in
pentode mode it is exaggerated. People look at the total harmonic
distortion but in fact the higher harmonics must be weighted much more
heavily than the second harmonic to account for its extraordinary
subliminal effect, often described as "edgineess" by professional
musicians.

-60dB, at average listeining levels????

Say a volt is average, then -60dB artifcats = 1mV.

1mV of grunge isn't exactly loud in most speakers......

Try Quad ESL, though of course not at 1W. Or try horns at 1W. You'd be
amazed what you can hear. About half of professional musicians can are
still disturbed by the artifacts at 60dB below; that's more than
enough for me to believe in subliminal perception of IM artifacts. But
anyone can hear them when they're only 30dB below perception. I'll
remind you that half a century ago -30dB was considered near enough
perfection... Arny Krueger says in another post that it is easy to set
up a test of whether people can hear 1mV IM; actually I found it very
frustrating getting qualified listeners, which is why in the end I
concentrated on testing professional musicians.

You listen to your speakers, not your amp; I would really advise you,
unless you are married to these insensitive speakers, first to get
good sensitive speakers you can live with forever, then build an amp
to suit them. If you have the space, an inexpensive sensitive speaker
you can build easily is The Impresario on my netsite, url under my
sig. EL84 are particularly sweet in triode but of course pitifully
underpowered for insensitive speakers.

I quite like UL EL84 good for 12 watts AB. But you do need
sensitive speakers over 92dB/W/M.

Of the common tubes, EL84 and EL34 are the greatest ever made, in my
humble opinion.

I'd prefer an SE 300B amp though....

Nah, not for a first amp; too expensive for messing around. I think
Dave has already made the choice in the SEntry.

Patrick Turner.

Andre Jute
Visit Jute on Amps athttp://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

[Andre Jute]

On Nov 2, 4:31 pm, John Byrns wrote:
In article CfGWi.1359$8S5.242@edtnps82, "Dave"
wrote:



"Andre Jute" wrote in message
oups.com...

And NFB up the kazoo. 0.1 per cent is a figure commonly bandied about
but I'm not overly impressed. I know for a fact that people cannot
hear as much as 3/4 per cent second harmonic but can hear that much
odd harmonic very clearly, and many people can hear or are made
uncomfortable by 0.3 per cent odd harmonics. What you want to do is
not to measure at full power but at some lower power where you will
actually use the amp; 1W or 2.83V into 8ohm is a common level with
even semi-sensitive speakers.

So... is ALL negative feedback bad? I am 40 with some background in
electrical engineering and learned nothing but transistor theory at
school... thermal runaway pretty much mandates feedback in all transistor
amps.

There are typically two different feedback mechanisms at work in a
transistor amplifier. The first, which is the one you are referring to
stabilizes the DC bias, or operating point, of the transistors to
prevent thermal runaway as you describe, and also to insure that the
"crossover" region of the transfer characteristic has the desired degree
of overlap. The second is the overall loop NFB, which is what Andre is
talking about, and is used to reduce the audio frequency distortion and
noise produced in the amplifier, as well as to stabilize the gain of the
amplifier, as the gain of discrete transistors is notoriously
inconsistent from sample to sample.

A lot of what I read about 'audiophile' grade amps is pro-znfb. Why?

It's because of a lack of understanding in the audiophile community.
NFB can work wonders when correctly applied by someone who knows what
they are doing, but it can also do serious harm when incompetently
applied.

That's true, of course, but I think some, who were educated in silicon
only as Dave was, come to tube amps with a silicon mindset that NFB is
the cure-all, that if a little is good, more must be better, and lots
will be superb. You can't really blame them: it is as natural for them
to reach for NFB as it is for you to breathe. Half our problems with
the likes of Poopie and Pinko arise from the fact that from hard
experience we scorn concepts that they revere, that to them are
perfectly axiomatic, beyond question in their firmament. To them we
are horrid heretics, questioning the very fundament of their faith
when we attack NFB.

Bad experiences with amplifiers using NFB of the latter type
lead to the typical audiophile' negative feelings towards NFB.

Regards,

John Byrns

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

Andre Jute
"I was at a board meeting for the LA Chapter of the Audio Engineering
Society last night on XM Satellite radio audio and data transmission.
Sadly, we missed you there, and at the SMPTE and Acoustical Society
recent meetings as well. Everyone was asking, 'Where is that wonderful
Andre Jute? The world just doesn't rotate without him...'" -- John
Mayberry, Emmaco

[Arny Krueger]

"John Byrns" wrote in message
...
In article ,
"Arny Krueger" wrote:

"Dave" wrote in message
news:CfGWi.1359$8S5.242@edtnps82...

"Andre Jute" wrote in message
ups.com...

I know for a fact that people cannot
hear as much as 3/4 per cent second harmonic but can hear that much
odd harmonic very clearly, and many people can hear or are made
uncomfortable by 0.3 per cent odd harmonics.

This order of harmonics issue is way overblown. Doesn't matter whether it
is
odd or even order nonlinearity, it all makes IM. IM is generally
aharmonic
and it all sounds bad.

But even order IM is much more obvious and annoying in listening tests,
although I suppose others may find odd order IM more annoying, its
probably partly a matter of personal preference.

I've never seen that. Got a reference?

Here's what my analysis says:

Fundamentals: F1 and F2

Second Order IM: F1±F2, F2-F1

Third Order IM: 2F1±F2, 2F2±F1

Fourth Order IM: 2F1±2F2, 2F2±2F1

Fifth Order IM: 3F1±2F2, 3F2±2F1

Almost all sums and differences are inharmonic and therefore likely to be
objectionable.

[Arny Krueger]

"Andre Jute" wrote in message
ups.com...

That's true, of course, but I think some, who were educated in silicon
only as Dave was, come to tube amps with a silicon mindset that NFB is
the cure-all, that if a little is good, more must be better, and lots
will be superb.

Spoken like someone who lacks experience with how coursed in feedback is
taught.

Courses in feedback are rarely if ever couched in either terms of either SS
or tubes. They are almost always taught in more abstract terms. Gain is
gain, frequency is frequency, phase is phase, poles are poles, and zeroes
are zeroes. Doesn't matter what technology is used to implement them.

Feedback has been around since no latter than the 1930s when tubes were all
we had.

I've seen dozens of feedback systems implemented in 100% tubes. Back in the
days when tubes were all we had, the intelligent use of loop feedback was
one of the hallmarks of a good amplfier. Virtually every Marantz, Dyna, and
McIntosh tubed preamp and amp incorporated loop feedback. There were tubed
amps with loop feedback that were bad designs and had stability problems.

[John Byrns]

In article ,
"Arny Krueger" wrote:

"John Byrns" wrote in message
...
In article ,
"Arny Krueger" wrote:

"Dave" wrote in message
news:CfGWi.1359$8S5.242@edtnps82...

"Andre Jute" wrote in message
ups.com...

I know for a fact that people cannot
hear as much as 3/4 per cent second harmonic but can hear that much
odd harmonic very clearly, and many people can hear or are made
uncomfortable by 0.3 per cent odd harmonics.

This order of harmonics issue is way overblown. Doesn't matter whether it
is
odd or even order nonlinearity, it all makes IM. IM is generally
aharmonic
and it all sounds bad.

But even order IM is much more obvious and annoying in listening tests,
although I suppose others may find odd order IM more annoying, its
probably partly a matter of personal preference.

I've never seen that. Got a reference?

No, it's my own result from listening tests, YMMV of course.

Here's what my analysis says:

Fundamentals: F1 and F2

Second Order IM: F1±F2, F2-F1

Third Order IM: 2F1±F2, 2F2±F1

Fourth Order IM: 2F1±2F2, 2F2±2F1

Fifth Order IM: 3F1±2F2, 3F2±2F1

Almost all sums and differences are inharmonic and therefore likely to be
objectionable.

Yes, but the even IM products tend to be removed in frequency from the
tones that produce them, so they are not masked to the extent that the
odd IM products are, which are closer in frequency to the tones that
produce them.


Regards,

John Byrns

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

[Dave]

"Andre Jute" wrote in message
ups.com...
Note that you need
single-ended transformers, not push-pull.

Would a push-pull present only half the impedence to the tube? I.e. a pp is
sort of like two transformers in one?

[Dave]

"Andre Jute" wrote in message
ps.com...
And

I note that on your SEntry amp schematic, you appear to have everything
grounded to earth (AC mains) ground, not to the CT of the power transformer.
Can this be correct?

I'm thinking maybe it's just an abbreviated schematic assuming one has built
power supplies before (which I have, hence my confusion at your notation).

Many PS designs which have center-tapped heater windings show them grounded
as well, is it standard to tie all unused center taps to (signal) ground?

Dave

[Andre Jute]

On Nov 2, 6:28 pm, "Arny Krueger" wrote:
"Andre Jute" wrote in message

ups.com...

That's true, of course, but I think some, who were educated in silicon
only as Dave was, come to tube amps with a silicon mindset that NFB is
the cure-all, that if a little is good, more must be better, and lots
will be superb.

Spoken like someone who lacks experience with how coursed in feedback is
taught.

Courses in feedback are rarely if ever couched in either terms of either SS
or tubes.

You're wanking, Krueger. By the time Dave was educated, there were no
more tubes. -- AJ

They are almost always taught in more abstract terms. Gain is
gain, frequency is frequency, phase is phase, poles are poles, and zeroes
are zeroes. Doesn't matter what technology is used to implement them.

Feedback has been around since no latter than the 1930s when tubes were all
we had.

I've seen dozens of feedback systems implemented in 100% tubes. Back in the
days when tubes were all we had, the intelligent use of loop feedback was
one of the hallmarks of a good amplfier. Virtually every Marantz, Dyna, and
McIntosh tubed preamp and amp incorporated loop feedback. There were tubed
amps with loop feedback that were bad designs and had stability problems.

[Andre Jute]

On Nov 2, 8:20 pm, "Dave" wrote:
"Andre Jute" wrote in message

ups.com...

Note that you need
single-ended transformers, not push-pull.

Would a push-pull present only half the impedence to the tube? I.e. a pp is
sort of like two transformers in one?

If you mean can you use one half of a PP output transformer for SE,
the answer is most definitely no. An SE transfomer is distinguished
from a PP one by an air gap. The SEntry works with any old SE output;
there are good inexpensive ones available in the States called, I
think, One Electron. (Where's Al Marcy? I think he has a pair.)

There is a way to use a transformer without an airgap for SE; it is
called parafeed, to indicate that the B+ is shunted in parallel to the
transformer primary, so that the output doesn't carry the current and
can therefore be designed to a more relaxed specificaiton. You also
need a big blocking cap, which some say is out of the signal and some
say can be heard; you make your choice and take your chances.

Andre Jute
Impedance is futile, you will be simulated into the triode of the
Borg. -- Robert Casey

[Eeyore]

Andre Jute wrote:

"Arny Krueger" wrote:
"Andre Jute" wrote in message

That's true, of course, but I think some, who were educated in silicon
only as Dave was, come to tube amps with a silicon mindset that NFB is
the cure-all, that if a little is good, more must be better, and lots
will be superb.

Spoken like someone who lacks experience with how coursed in feedback is
taught.

Courses in feedback are rarely if ever couched in either terms of either SS
or tubes.

You're wanking, Krueger. By the time Dave was educated, there were no
more tubes. -- AJ

Irrelevant to feedback theory. Feedback works in the same way, whatever active
parts are in use. It is, after all, merely maths at the end of the day.

Graham

[Eeyore]

Andre Jute wrote:

"Dave" wrote:
"Andre Jute" wrote in message

Note that you need
single-ended transformers, not push-pull.

Would a push-pull present only half the impedence to the tube? I.e. a pp is
sort of like two transformers in one?

If you mean can you use one half of a PP output transformer for SE,
the answer is most definitely no. An SE transfomer is distinguished
from a PP one by an air gap.

Actually, it's distinguished by the adility to carry a significant DC current. An
air gap is merely one way to achieve that.

Graham

[Andre Jute]

On Nov 2, 11:19 pm, Eeyore
wrote:
Andre Jute wrote:
"Dave" wrote:
"Andre Jute" wrote in message

Note that you need
single-ended transformers, not push-pull.

Would a push-pull present only half the impedence to the tube? I.e. a pp is
sort of like two transformers in one?

If you mean can you use one half of a PP output transformer for SE,
the answer is most definitely no. An SE transfomer is distinguished
from a PP one by an air gap.

Actually, it's distinguished by the adility to carry a significant DC current. An
air gap is merely one way to achieve that.

That too (1). So refreshing to see you being constructive and on-
topic, Graham. Can we look forward to more of the same behaviour?

Graham
Andre Jute
"I was at a board meeting for the LA Chapter of the Audio Engineering
Society last night on XM Satellite radio audio and data transmission.
Sadly, we missed you there, and at the SMPTE and Acoustical Society
recent meetings as well. Everyone was asking, 'Where is that wonderful
Andre Jute? The world just doesn't rotate without him...'" -- John
Mayberry, Emmaco

(1) When I was young and cheerful, before I took up engineering and
became grimly determined and dull, I wrote a prizewinning broadcast
serial about a secret agent even clumsier than Stevenson; the hero,
for whom things always accidentally turned out right, had a tagline.
When people told him something he should have thought of himself, he
would say, "That too!"

[Andre Jute]

On Nov 2, 11:17 pm, Eeyore
wrote:
Andre Jute wrote:
"Arny Krueger" wrote:
"Andre Jute" wrote in message

That's true, of course, but I think some, who were educated in silicon
only as Dave was, come to tube amps with a silicon mindset that NFB is
the cure-all, that if a little is good, more must be better, and lots
will be superb.

Spoken like someone who lacks experience with how coursed in feedback is
taught.

Courses in feedback are rarely if ever couched in either terms of either SS
or tubes.

You're wanking, Krueger. By the time Dave was educated, there were no
more tubes. -- AJ

Irrelevant to feedback theory. Feedback works in the same way, whatever active
parts are in use. It is, after all, merely maths at the end of the day.

Graham

You two soundbite mentalities have missed the point altogether. Again.

The principles of negative feedback are just math, sure. But
electronics engineers who are taught to apply negative feedback are
now differently instructed to the time when tubes were dominant. For
instance, Julian L. Bernstein, Associate Dean of the RCA Day Schools,
in his book Audio Systems makes the point that NFB must be used with
restraint. By contrast, modern EEs are taught that mo' neg' is
betta neg' or what whatever the current street-talk-derived
vernacular is on campus this week. You two superannuates are just too
old and disconnected and unhip to discover this for yourselves. I made
exactly two phone calls to find out all of that.

Andre Jute
Mr Street Cred

[Phil Allison]

"Dave"

So... is ALL negative feedback bad? I am 40 with some background in
electrical engineering and learned nothing but transistor theory at
school... thermal runaway pretty much mandates feedback in all transistor
amps.


** What utter nonsense.

The NFB loop has no effect on "thermal runaway".



........ Phil

[Eeyore]

Andre Jute wrote:

Eeyore wrote:
Andre Jute wrote:
"Dave" wrote:
"Andre Jute" wrote in message

Note that you need
single-ended transformers, not push-pull.

Would a push-pull present only half the impedence to the tube? I.e. a pp is
sort of like two transformers in one?

If you mean can you use one half of a PP output transformer for SE,
the answer is most definitely no. An SE transfomer is distinguished
from a PP one by an air gap.

Actually, it's distinguished by the adility to carry a significant DC current. An
air gap is merely one way to achieve that.

That too (1). So refreshing to see you being constructive and on-
topic, Graham. Can we look forward to more of the same behaviour?

I am always constructive and on-topic unless provoked otherwise.

Can I perhaps advise that your posts would benefit from having less bile ?

Graham

* note 'constructive' does not mean being 'pro tube' regardless of application.
Suggestions that they are the only 'magically pure' way to listen to music will meet
with severe disdain.

[Eeyore]

Phil Allison wrote:

"Dave"

So... is ALL negative feedback bad? I am 40 with some background in
electrical engineering and learned nothing but transistor theory at
school... thermal runaway pretty much mandates feedback in all transistor
amps.

** What utter nonsense.

The NFB loop has no effect on "thermal runaway".

Oh, I missed that one. What a corker !

You are of course absolutely right. Good solid state amplifier design embodies
'thermal feedback' for stability of the quiescent current which has NOTHING to
do with the audio signal path and merely affects the static 'DC conditions'.

Apples and oranges.

Graham

[Andre Jute]

On Nov 2, 8:54 pm, "Dave" wrote:
"Andre Jute" wrote in message

ps.com...
And

I note that on your SEntry amp schematic, you appear to have everything
grounded to earth (AC mains) ground, not to the CT of the power transformer.
Can this be correct?

Some isolation transformers have no centre tap.

Standard valve amp practice is to ground the signal section at the
input and the power supplyg at the negative end of the first cap. The
centretap if present and in use (not when the windings are seriesed
for higher voltage!) is also grounded. All these grounds are then led
to the mains ground. The case is also grounded to the mains. So much
for standard practice.

When an amp is as simple as many of mine are, and the T201 SEntry
http://members.lycos.co.uk/fiultra/Jute-EL34-SEntry.jpg
certainly is, a much more ideal grounding scheme becomes possible. It
is called "Star Grounding". The Star Ground is the point in the middle
of the SEntry circuit schematic where all the 0V lines meet. This Star
Ground point is then grounded to mains earth. Notice that an
additional ground, from metal chassis to mains earth is not shown but
is essential.

In practice where you place the star ground depends on your layout.
For instance, in another of my amps, T68bis Minus Zero
http://members.lycos.co.uk/fiultra/K...0T68MZ417A.jpg
http://members.lycos.co.uk/fiultra/t...17acircuit.jpg
http://members.lycos.co.uk/fiultra/t68bismzlayout.jpg
the star ground is at the bottom end of the bleed resistor, quite
literally built onto a piece of stiff wire soldered into the lead hole
of an ali-cased power resister; it's where those green line go in the
middle of the layout suggestion. From there a wire is led to the
mains earth (the green line that ends at the square power socket
cutout bottom right). Another wire from mains earth is bolted to the
chassis very nearby with toothed washers backed by flat washers both
sideas to ensure that ali corrosion doesn't break the most crucial
contact in the amp!

I'm thinking maybe it's just an abbreviated schematic assuming one has built
power supplies before (which I have, hence my confusion at your notation).

Sure. I don't know how you will build it. I have to assume you know
enough to ground your chassis! I also assume you will use a three-
point (IEC) grounded plug, not those dangerous American two-point
things.

Many PS designs which have center-tapped heater windings show them grounded
as well, is it standard to tie all unused center taps to (signal) ground?

No. It depends on what you're trying to achieve. In particular, I want
to warn you against making more than one ground connection to the
case, as ground loops can be infuriating to track down and fix. That
vintage practice of using the chassis as a bus bar is entirely
unacceptable today.

Dave

In general, I and everyone else publish a circuit on the understanding
that only those with the correct level of experience will attempt to
build it. I cannot tell you the hundreds of details you will only
learn through doing it. I withdrew my kilovolt amps, 845 and 211 and a
particularly attractive 80W SE SV572-3 amp from publication years ago
after receiving too many letters that started, "After reading your
wonderful netsite I have just bought my first soldering iron and DMM
and ordered 845 tubes to starting building a 1250V tube amp just like
yours...." But even a "low" voltage amp like the SEntry will kill you
stone dead if you are careless, and your pets and family too. That is
why to people with less appropriate backgrounds than yours we normally
suggest that they buy a kit with good, complete instructions rather
than try building from scratch.

It is important that you must not switch on your amp once built, or
poke around in it, unless someone is standing by who knows where to
pull the plug. I got an experienced ham to supervise my first efforts,
and he brought several buddies as a vote of confidence. Once we
switched the amp on and watched it for a while for the appearance of
*smoke*, we had a good time listening to it, comparing it with my
bought amps and the DIY amps the other guys brought. Dawn was breaking
when that party broke up. Those guys also each brought me something
from their junkboxes towards my next amp, which was a nice touch.

HTH. You have only to ask. If I'm not here, there are others with
experience who will help you. I particularly recommend Patrick Turner
and John Byrns. Phil Allison also knows what he talks of but getting a
full explanation from him is like pulling teeth.

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

[robert casey]

If you mean can you use one half of a PP output transformer for SE,
the answer is most definitely no. An SE transfomer is distinguished
from a PP one by an air gap.

There is a way to use a transformer without an airgap for SE; it is
called parafeed, to indicate that the B+ is shunted in parallel to the
transformer primary, so that the output doesn't carry the current and
can therefore be designed to a more relaxed specificaiton. You also
need a big blocking cap, which some say is out of the signal and some
say can be heard; you make your choice and take your chances.


One could use a push-pull output transformer om a single-ended design,
if you also send a steady DC current (equal to the quiescent class A
current of the tube) thru the other half of the primary from the
centertap (connected to B+) to the unused side, via most likely a power
resistor to ground. But that's rather wasteful of the power you went to
the effort to build a power supply to supply that and the tube's side.
So nobody ever does that, except maybe only if someone has on hand some
P-P output transformers but insists on single endedness. But it would
be a kludge... :-)

Or regap that transformer, but you could ruin the sound quality that was
designed into it (I'm not familiar with transformer design, but I've
since learned not to waste time fooling with power transformers in audio
output work :-) ). You could use the old centertap as a kind of UL
tap, but the tap position is not optimal for UL work. But you'd be best
off saving the P-P transformers for another project, and get single
ended transformers that were designed for that service instead.