The OTL breakthrough: not tubes but speakers?
 

I gave up the idea of direct tube drive of Quad ESL63 years ago as too
hot and too dangerous. But my recent experiences with Stax
electrostatic earphones driven by small OTL tube amps makes me wonder
if the next breakthrough in tube amps will not be a combination set of
a tube amp directly driving the associated fullsize electrostatic
fullrange speaker.

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

The OTL breakthrough: not tubes but speakers?
 

On Nov 21, 8:27 pm, Andre Jute wrote:
I gave up the idea of direct tube drive of Quad ESL63 years ago as too
hot and too dangerous. But my recent experiences with Stax
electrostatic earphones driven by small OTL tube amps makes me wonder
if the next breakthrough in tube amps will not be a combination set of
a tube amp directly driving the associated fullsize electrostatic
fullrange speaker.


Why speculate? Make one.

Maybe with a Lundahl power transformer? That would be one (1) of the
promised twelve, at the least. Better late than never.

Peter Wieck
Wyncote, PA

The OTL breakthrough: not tubes but speakers?
 

Andre Jute wrote:

I gave up the idea of direct tube drive of Quad ESL63 years ago as too
hot and too dangerous. But my recent experiences with Stax
electrostatic earphones driven by small OTL tube amps makes me wonder
if the next breakthrough in tube amps will not be a combination set of
a tube amp directly driving the associated fullsize electrostatic
fullrange speaker.

A full range ESL does require a large voltage swing.

Quad ESL57 have a 300:1 step up ratio tranny and so if 25Vrms is needed
at the low voltage input coil, then without the tranny
the voltage required is 7,500Vrms.

Where does one get a linear 7,500Vrms from?

ESL63 and later derivatives are slightly easier to drive, but basically
the drive voltage needed isn't ever going to become lower because of the
laws of physics
with regard to charged plates, the distance between them, membrane
travel requirements,
and high voltage Vdc charge biasing.

Its all OK in a pair of ESL headphones because the levels of Vac drive
and bias etc are
all far lower than the potentially lethal levels in a full size ESL
which has to fill a room with sound, not just an ear held close.


Morgan Jones explains how he uses a pair of PP 845 to get a high drive
Vac.
Many other tubes capable come to mind, 813 operating off several kV
perhaps but they will never ever
be adopted by anyone except the few clever types who enjoy doing things
the hard way.
Radio amateurs routinely ran their transmitters with several kV but it
takes
huge dedication to be a really capable radio amateur if one builds
each and every part of one's own radio station.
DIY with HV to get better audio than available by other means is also a
challenge.

I don't mind using magnetism as the means to get from tubes to ESL
panels.

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

The OTL breakthrough: not tubes but speakers?
 

Patrick Turner wrote:

Andre Jute wrote:

I gave up the idea of direct tube drive of Quad ESL63 years ago as too
hot and too dangerous. But my recent experiences with Stax
electrostatic earphones driven by small OTL tube amps makes me wonder
if the next breakthrough in tube amps will not be a combination set of
a tube amp directly driving the associated fullsize electrostatic
fullrange speaker.

A full range ESL does require a large voltage swing.

Quad ESL57 have a 300:1 step up ratio tranny and so if 25Vrms is needed
at the low voltage input coil, then without the tranny
the voltage required is 7,500Vrms.

Where does one get a linear 7,500Vrms from?

A few hundred 6BK4. They're good for 16000V before they turn into X-
ray machines. Unfortunately, they're only good for 1.6mA, so you'd
need to stack a *lot* of them. I seem to remember 344 tubes per PP
channel with over 85W of filament heating for stereo. (God, I love
understatement: not even an exclamation mark.) But, hey, Iain lives in
Finland where it freezes in winter; we *are* talking about a niche
market amp.

But parallel 450TH might have the oomph with 6000V at 75mA each, 150mA
the pair at around 6000V would probably give reasonable levels with
almost ESL63-sized panels. You could use stacked panels and stack up
the tubes another layer to boost the bass. (No idea what 450TH cost
these days...)

ESL63 and later derivatives are slightly easier to drive, but basically
the drive voltage needed isn't ever going to become lower because of the
laws of physics
with regard to charged plates, the distance between them, membrane
travel requirements,
and high voltage Vdc charge biasing.

***If there is demand, we could, as a theoretical exercise, work
through the math, see why in the end I and so many others bugged out
and decided a transformer is the way to go to drive ESL. Some might
find the theory interesting.

Its all OK in a pair of ESL headphones because the levels of Vac drive
and bias etc are
all far lower than the potentially lethal levels in a full size ESL
which has to fill a room with sound, not just an ear held close.

Yeah. But I've been there, done that, as you know. (Thanks again for
your help, Patrick, and all the others who looked over the designs of
my electrostatic headphone OTL amps.) I was just wondering semi-idly
if I didn't give up too quickly on direct drive.

Morgan Jones explains how he uses a pair of PP 845 to get a high drive
Vac.
Many other tubes capable come to mind, 813 operating off several kV
perhaps but they will never ever
be adopted by anyone except the few clever types who enjoy doing things
the hard way.

2200V from 813 and (from memory, a similar tube available back in
c1995) 8068 just isn't enough to play fullsize electrostats at
reasonable volumes.

Radio amateurs routinely ran their transmitters with several kV but it
takes
huge dedication to be a really capable radio amateur if one builds
each and every part of one's own radio station.
DIY with HV to get better audio than available by other means is also a
challenge.

Like I said, I do love understatement!

I don't mind using magnetism as the means to get from tubes to ESL
panels.

In the end it doesn't matter whether you mind the transformer or not,
the alternative of several (to hundreds in the ludicrous case) of
valves operating on many kV is too dangerous and too hot and too big
and too expensive to consider sanely. And, in truth, the ESL
transformer, at least in the Quad ESL I know, is blameless; you can't
ask more of iron than that. I wonder of Sander still lurks: what about
in your Maggies?

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

The OTL breakthrough: not tubes but speakers?
 

A full range ESL does require a large voltage swing.

Quad ESL57 have a 300:1 step up ratio tranny and so if 25Vrms is needed
at the low voltage input coil, then without the tranny
the voltage required is 7,500Vrms.

Where does one get a linear 7,500Vrms from?


There's triodes, intended for color TV high voltage regulation (like
25KV), like the 6BK4, with mu around 1000. The plate could feed into an
inductor similar to the inductance of the transformer normally used (to
step the audio up to the high voltage in the usual setup). The plate
would also connect to the electrostatic speaker.

The OTL breakthrough: not tubes but speakers?
 

On Nov 22, 11:02 pm, robert casey wrote:
A full range ESL does require a large voltage swing.

Quad ESL57 have a 300:1 step up ratio tranny and so if 25Vrms is needed
at the low voltage input coil, then without the tranny
the voltage required is 7,500Vrms.

Where does one get a linear 7,500Vrms from?

There's triodes, intended for color TV high voltage regulation (like
25KV), like the 6BK4, with mu around 1000. The plate could feed into an
inductor similar to the inductance of the transformer normally used (to
step the audio up to the high voltage in the usual setup). The plate
would also connect to the electrostatic speaker.

You could make a fine tweeter with the 6BK4 but for a fullrange
fullsize electrostat you'd need hundreds. That's a lot of filaments at
0.2A 6.3V...

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

The OTL breakthrough: not tubes but speakers?
 

You could make a fine tweeter with the 6BK4 but for a fullrange
fullsize electrostat you'd need hundreds. That's a lot of filaments at
0.2A 6.3V...


I'd then pick a radio transmitter tube, but the ones that I'm only at
best vaguely familiar with run at only a few thousand volts at most.
Depending on how loud you want your music, and how efficient
electrostatic speakers are, would determine how much power you'd need.
Someone mentioned 344 6BK4s each at 1.6ma at 16KV. That's about 550ma,
from a 16KV supply that's about 8KW of "input" power. That's an awful
lot of power, we must be off an order of magnitude or 2 here. Or are
electrostatic speakers really inefficient? But if I take a scientific
wild ass guess (SWAG) that 80W is more like it, then that would be 5ma
of plate current. 5 6BK4s (1ma each) could do that. But this must be
wrong as well...

The OTL breakthrough: not tubes but speakers?
 

robert casey wrote:


You could make a fine tweeter with the 6BK4 but for a fullrange
fullsize electrostat you'd need hundreds. That's a lot of filaments at
0.2A 6.3V...


I'd then pick a radio transmitter tube, but the ones that I'm only at
best vaguely familiar with run at only a few thousand volts at most.
Depending on how loud you want your music, and how efficient
electrostatic speakers are, would determine how much power you'd need.
Someone mentioned 344 6BK4s each at 1.6ma at 16KV. That's about 550ma,
from a 16KV supply that's about 8KW of "input" power. That's an awful
lot of power, we must be off an order of magnitude or 2 here. Or are
electrostatic speakers really inefficient? But if I take a scientific
wild ass guess (SWAG) that 80W is more like it, then that would be 5ma
of plate current. 5 6BK4s (1ma each) could do that. But this must be
wrong as well...

Possibly the 4-65A, which is spec-ed for 3kV DC on top.
A PP pair of those (feeding a centre-tapped choke, rather than a
transformer primary) could give 6kV swing each side, which is getting
into the right area.

The OTL breakthrough: not tubes but speakers?
 

Andre Jute wrote:

Patrick Turner wrote:

Andre Jute wrote:

I gave up the idea of direct tube drive of Quad ESL63 years ago as too
hot and too dangerous. But my recent experiences with Stax
electrostatic earphones driven by small OTL tube amps makes me wonder
if the next breakthrough in tube amps will not be a combination set of
a tube amp directly driving the associated fullsize electrostatic
fullrange speaker.

A full range ESL does require a large voltage swing.

Quad ESL57 have a 300:1 step up ratio tranny and so if 25Vrms is needed
at the low voltage input coil, then without the tranny
the voltage required is 7,500Vrms.

Where does one get a linear 7,500Vrms from?

A few hundred 6BK4. They're good for 16000V before they turn into X-
ray machines. Unfortunately, they're only good for 1.6mA, so you'd
need to stack a *lot* of them. I seem to remember 344 tubes per PP
channel with over 85W of filament heating for stereo. (God, I love
understatement: not even an exclamation mark.) But, hey, Iain lives in
Finland where it freezes in winter; we *are* talking about a niche
market amp.

But parallel 450TH might have the oomph with 6000V at 75mA each, 150mA
the pair at around 6000V would probably give reasonable levels with
almost ESL63-sized panels. You could use stacked panels and stack up
the tubes another layer to boost the bass. (No idea what 450TH cost
these days...)

ESL63 and later derivatives are slightly easier to drive, but basically
the drive voltage needed isn't ever going to become lower because of the
laws of physics
with regard to charged plates, the distance between them, membrane
travel requirements,
and high voltage Vdc charge biasing.

***If there is demand, we could, as a theoretical exercise, work
through the math, see why in the end I and so many others bugged out
and decided a transformer is the way to go to drive ESL. Some might
find the theory interesting.

Good question re demand.

I looked at ESL transformers very seriously earlier this year
and the problem is winding one with low self shunt capacitances.

If you have a typical 1:200 step up transformer with a miserly 100pF
across the HV sec,
it appears as 200 x 200 x 100 pF at the primary, ie, 6uF.

Quad's ESL57 had 35pF but their SUT was a wonder to behold, and full of
tricks to overcome the capacitance, mainly by means of using closer
gapped panels for
treble and with less drive voltage, so the capacitances were not
transformed upwards so badly.
The ESL57 had high bass panel C but with deliberately high leakage L it
forms a
low pass filter.

You need to read carefully what Peter Baxandal wrote about the
'equivalent circuit for the Quad ESL57'
It makes interesting reading, but I know of absolutely not one solitary
person who
has the feintest idea what Peter B was on about in his analysis.
The guys who have described their DIY ESL attempts on the NET display
their gross ignorance
by never daring to fully explain in Baxandal Style what is really
required.
Walker and Baxandal were REALLY CLEVER PEOPLE.
Alas nobody really cares to follow their understanding....
Or if they do, they stay out of here and stay private about it.


Its all OK in a pair of ESL headphones because the levels of Vac drive
and bias etc are
all far lower than the potentially lethal levels in a full size ESL
which has to fill a room with sound, not just an ear held close.

Yeah. But I've been there, done that, as you know. (Thanks again for
your help, Patrick, and all the others who looked over the designs of
my electrostatic headphone OTL amps.) I was just wondering semi-idly
if I didn't give up too quickly on direct drive.

Morgan Jones explains how he uses a pair of PP 845 to get a high drive
Vac.
Many other tubes capable come to mind, 813 operating off several kV
perhaps but they will never ever
be adopted by anyone except the few clever types who enjoy doing things
the hard way.

2200V from 813 and (from memory, a similar tube available back in
c1995) 8068 just isn't enough to play fullsize electrostats at
reasonable volumes.

Radio amateurs routinely ran their transmitters with several kV but it
takes
huge dedication to be a really capable radio amateur if one builds
each and every part of one's own radio station.
DIY with HV to get better audio than available by other means is also a
challenge.

Like I said, I do love understatement!

I don't mind using magnetism as the means to get from tubes to ESL
panels.

In the end it doesn't matter whether you mind the transformer or not,
the alternative of several (to hundreds in the ludicrous case) of
valves operating on many kV is too dangerous and too hot and too big
and too expensive to consider sanely. And, in truth, the ESL
transformer, at least in the Quad ESL I know, is blameless; you can't
ask more of iron than that. I wonder of Sander still lurks: what about
in your Maggies?


ESL57 tried to get one well made SUT to wear all the necessary hats at
once.
The ESL57 SUT has many MORE turns than found in any OPT for a pair of
tubes.

I know of nobody at all who winds anything what Quad did in 1957 like
this anymore.

Not even Quad. They went to using TWO SUTs, one for each phase applied
to each stator in ESL63.

This is a sensible option, because the size and turns per volt needed
all
favour reducing shunt C enormously. One can also use a separate
transformer for bass from 30Hz to 1kHz, which has a series resistance
feed to the panels,
and with the leakage inductance the capacitance never loads the amp.
Another transformer is used for the F above 1kHz which has far fewer
turns per volt
and has a smaller size and low self capacitance.
Its self capacitance will be less than the panel capacitance and the
total capacitance
seen by the amplifier can be kept below 2uF, and can be arranged so that
winding resistance and added trimming R will have 2ohms at least in
series with the 2uF.

Since there is such a tiny amount of audio power in signals above 5kHz,
2ohms plus 2uF is a load
tolerated by rather poor amps including Quad-II amps.

I sent an 8585 amp of mine to Melbourne where a guy has THREE stacked
pairs of ESL57.
The quad of KT90 have no trouble driving such a "horrendous load".

Its not as horrendous as it seems, because ESL have a rising Z as F
falls because capacitors have higher reactance
as F falls.

I am not sure what is being done inside the Quad 2805 ES speakers with
transformers.

But I could say the few people in the know at Quad are well in front of
me about
what works and what doesn't.

To avoid ppl like us copying their ideas, you NEVER see them explaining
it all in news groups.

Patrick Turner.



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

The OTL breakthrough: not tubes but speakers?
 

Patrick Turner wrote:

[snip]

I looked at ESL transformers very seriously earlier this year
and the problem is winding one with low self shunt capacitances.

If you have a typical 1:200 step up transformer with a miserly 100pF
across the HV sec,
it appears as 200 x 200 x 100 pF at the primary, ie, 6uF.

Quad's ESL57 had 35pF but their SUT was a wonder to behold, and full of
tricks to overcome the capacitance, mainly by means of using closer
gapped panels for
treble and with less drive voltage, so the capacitances were not
transformed upwards so badly.
The ESL57 had high bass panel C but with deliberately high leakage L it
forms a
low pass filter.

200 x 200 x 100 pF = 4uF, not 6. Never mind :-)

More to the point, isn't a large part of the problem (with ESL
transformers) ours to remove? Meaning that most ESL transformers are
designed to be driven from a 4/8 ohm source. Logical enough, if you're
making transformers to plug on to an existing amplifier.
For example, the ones on Sowter's website, with ratios around 100:1.

But if we have a tube output stage, the B+ is (say) 400V, and we can
have a 600V swing quite easily. Stepping that up to 3kV is only 5:1,
which should be quite manageable.
The drawback is that this is definitely a special transformer. (While
we're winding it, let's put a reduced-power 3rd winding on, to run that
bass speaker as well.)