Well I finally got around to trying some Svetlana 6KG6s for single ended
and heres the design I'm using:

http://68.162.128.102:808/~tesla/6KG6_SE.gif

The datasheet says 50Vp-p of drive and the 417A supplys it well. In
triode mode I get a clean 3 watts with a B+ of 300 at 75mA and 2v input.

Pentode mode is an entirely different matter. No matter what plate
voltage or how the tube is biased the output is always high in
distortion and the sinewave looks like garbage. The sweet spot was a
plate voltage of 250 at 65mA. Power output was 6w and distortion pretty
high at 7%.

Any ideas on how to improve pentode operation? I'm having a tough time
finding the right operating points. Thanks.


Adam

How about regulating the screen at a lower voltage?

Jon



in article , Adam Stouffer at
wrote on 7/19/04 10:33 PM:

Well I finally got around to trying some Svetlana 6KG6s for single ended
and heres the design I'm using:

http://68.162.128.102:808/~tesla/6KG6_SE.gif

The datasheet says 50Vp-p of drive and the 417A supplys it well. In
triode mode I get a clean 3 watts with a B+ of 300 at 75mA and 2v input.

Pentode mode is an entirely different matter. No matter what plate
voltage or how the tube is biased the output is always high in
distortion and the sinewave looks like garbage. The sweet spot was a
plate voltage of 250 at 65mA. Power output was 6w and distortion pretty
high at 7%.

Any ideas on how to improve pentode operation? I'm having a tough time
finding the right operating points. Thanks.


Adam

Just wondered if you'd seen this. I'm just awaiting transformers for this
to build it myself!

http://www.svetlana.com/docs/products/glass/EL509.html

also, on the authors own home page

http://www.geocities.com/bobdanielak/technoteNo33.html

Has anyone else attempted this amp? For my novelty input, I'll be using
EY500 rectifiers!

Si.




"Adam Stouffer" wrote in message
...
Well I finally got around to trying some Svetlana 6KG6s for single ended
and heres the design I'm using:

http://68.162.128.102:808/~tesla/6KG6_SE.gif

The datasheet says 50Vp-p of drive and the 417A supplys it well. In
triode mode I get a clean 3 watts with a B+ of 300 at 75mA and 2v input.

Pentode mode is an entirely different matter. No matter what plate
voltage or how the tube is biased the output is always high in
distortion and the sinewave looks like garbage. The sweet spot was a
plate voltage of 250 at 65mA. Power output was 6w and distortion pretty
high at 7%.

Any ideas on how to improve pentode operation? I'm having a tough time
finding the right operating points. Thanks.


Adam

Vs down around 40V or Rl down around 1k. It's a thick tube and requires
thick current, 300V 100mA isn't the way to go.

Tim

--
"I've got more trophies than Wayne Gretsky and the Pope combined!"
- Homer Simpson
Website @ http://webpages.charter.net/dawill/tmoranwms

"Adam Stouffer" wrote in message
...
Well I finally got around to trying some Svetlana 6KG6s for single ended
and heres the design I'm using:

http://68.162.128.102:808/~tesla/6KG6_SE.gif

The datasheet says 50Vp-p of drive and the 417A supplys it well. In
triode mode I get a clean 3 watts with a B+ of 300 at 75mA and 2v input.

Pentode mode is an entirely different matter. No matter what plate
voltage or how the tube is biased the output is always high in
distortion and the sinewave looks like garbage. The sweet spot was a
plate voltage of 250 at 65mA. Power output was 6w and distortion pretty
high at 7%.

Any ideas on how to improve pentode operation? I'm having a tough time
finding the right operating points. Thanks.


Adam

Adam Stouffer wrote:

Well I finally got around to trying some Svetlana 6KG6s for single ended
and heres the design I'm using:

http://68.162.128.102:808/~tesla/6KG6_SE.gif

The datasheet says 50Vp-p of drive and the 417A supplys it well. In
triode mode I get a clean 3 watts with a B+ of 300 at 75mA and 2v input.

In triode, with the 22.5 watts of dissipation, you should get 30%
efficiency,
which is 6.75 watts at around 7% thd.

I don't like this tube much at all unless one uses it in PP
with lots of class A and with cathode FB windings on the OPT,
and then its starts to clean up its act; see below about SECFB.



Pentode mode is an entirely different matter. No matter what plate
voltage or how the tube is biased the output is always high in
distortion and the sinewave looks like garbage. The sweet spot was a
plate voltage of 250 at 65mA. Power output was 6w and distortion pretty
high at 7%.

The anode dissipation you have is only 16.3 watts, so try 300v at 90 mA,
and a lower value RL, and try running the screem at some point between +150v
and +200v.
You should get 10.8 watts max, at 40% efficiency.



Any ideas on how to improve pentode operation? I'm having a tough time
finding the right operating points. Thanks.

Traditional SE Pentode operation of EL509 is always going to be a pita.

For good power, and best linearity, use screen drive, with G1 taken tom 0V,
and have around 27 watts of anode dissipation, 300v x 90 mA will do.
The screen needs a direct coupled cathode follower with a preceding voltage
amp
capable of around 25vrms , or 75vrms for low driver thd.
See enhanced triode operation the svetlana site.

Better still is the use of a distributed loading on the primary of the OPT,
so that you have 80% of the P turns between B+ and anode,
and 20% between cathode and 0V, which is an application of 20%
of the voltage across the tube being cathode feedback.
Ea = +300 v, Eg2 = +150v( a fixed voltage, Ia = 90 mA,
and po should be about 9 watts.
This should give Ro slightly lower than triode, and around 2% max thd,
and need around the same drive voltage as triodes.

The driver could be a normal triode resistance loaded stage using 12BH7
paralleled, which should give around 1% thd at the max drive voltage,
which will cancel half the 509 thd at most load values, especially the lower

load values, since the main thd is 2H.

Patrick Turner.







Adam

The major THD component is 2H

Jon Yaeger wrote:

How about regulating the screen at a lower voltage?

Jon

Indeed G2 supply should be much lower than Ea, and if so, IG2
is at its lowest value, and there is no need to regulated for
class A SE; just use 100uF from a resistance dropper from the B+,
and she will be as good as it gets with music operation.
But when tested with a sine wave, there will be a drop in EG2,
which will increase the thd, so shunt regulate for tests,
and so have say 3 x 65v x 5W zeners across a 100 Uf fed by R from the B+,
and this allows the G2 voltage top still sag if there is a tube problem and
IG2
increases to more than twice the idle value.
Don't use series pass element regulators for G2.

Patrick Turner.


in article , Adam Stouffer at
wrote on 7/19/04 10:33 PM:

Well I finally got around to trying some Svetlana 6KG6s for single ended
and heres the design I'm using:

http://68.162.128.102:808/~tesla/6KG6_SE.gif

The datasheet says 50Vp-p of drive and the 417A supplys it well. In
triode mode I get a clean 3 watts with a B+ of 300 at 75mA and 2v input.

Pentode mode is an entirely different matter. No matter what plate
voltage or how the tube is biased the output is always high in
distortion and the sinewave looks like garbage. The sweet spot was a
plate voltage of 250 at 65mA. Power output was 6w and distortion pretty
high at 7%.

Any ideas on how to improve pentode operation? I'm having a tough time
finding the right operating points. Thanks.


Adam

si wrote:

Just wondered if you'd seen this. I'm just awaiting transformers for this
to build it myself!

http://www.svetlana.com/docs/products/glass/EL509.html

also, on the authors own home page

http://www.geocities.com/bobdanielak/technoteNo33.html

Has anyone else attempted this amp? For my novelty input, I'll be using
EY500 rectifiers!

Si.

The circuit at http://www.geocities.com/bobdanielak/technoteNo33.html
looks fine, but I would recommend
using a stand alone EL84 as the cathode follower buffer between the
EL509 screen and input triode driver, which I think should be a 12AX7
paralleled,
since it has to make quite a few volts of drive.

NFB is essential in this circuit, since Ra in enhanced triode mode is around 4
k ohms.
About 12 dB of NFB would be the minimum amount of NFB.

Patrick Turner.



"Adam Stouffer" wrote in message
...
Well I finally got around to trying some Svetlana 6KG6s for single ended
and heres the design I'm using:

http://68.162.128.102:808/~tesla/6KG6_SE.gif

The datasheet says 50Vp-p of drive and the 417A supplys it well. In
triode mode I get a clean 3 watts with a B+ of 300 at 75mA and 2v input.

Pentode mode is an entirely different matter. No matter what plate
voltage or how the tube is biased the output is always high in
distortion and the sinewave looks like garbage. The sweet spot was a
plate voltage of 250 at 65mA. Power output was 6w and distortion pretty
high at 7%.

Any ideas on how to improve pentode operation? I'm having a tough time
finding the right operating points. Thanks.


Adam

Hi,

I've build an integrated amp with this amp as the power amp. I liked the
design with no coupling caps and I had a few PL519's around. I used the
design before this revised design, with no connection between the grid and
ground, and maybe I should connect these, but then again: if it is not
broken, don't fix it.
I like the sound of it, I liked it even better then my previous SE EL84 amp.
I used the PL519 instead of the EL509 and a PCL82 instead of the ECL82 (So
you do not need to raise the V of the ECL82 heaters ( max V Cathode Heater
PCL82=200V)). I was also thinking of using PY500A rectifiers, but ended up
with 2 plain old EZ81 rectifiers.

With kind regards,

Simon

"si" ' schreef in bericht
...
Just wondered if you'd seen this. I'm just awaiting transformers for this
to build it myself!

http://www.svetlana.com/docs/products/glass/EL509.html

also, on the authors own home page

http://www.geocities.com/bobdanielak/technoteNo33.html

Has anyone else attempted this amp? For my novelty input, I'll be using
EY500 rectifiers!

Si.




"Adam Stouffer" wrote in message
...
Well I finally got around to trying some Svetlana 6KG6s for single ended
and heres the design I'm using:

http://68.162.128.102:808/~tesla/6KG6_SE.gif

The datasheet says 50Vp-p of drive and the 417A supplys it well. In
triode mode I get a clean 3 watts with a B+ of 300 at 75mA and 2v input.

Pentode mode is an entirely different matter. No matter what plate
voltage or how the tube is biased the output is always high in
distortion and the sinewave looks like garbage. The sweet spot was a
plate voltage of 250 at 65mA. Power output was 6w and distortion pretty
high at 7%.

Any ideas on how to improve pentode operation? I'm having a tough time
finding the right operating points. Thanks.


Adam

Bruce Rosenblitz has a nice EL509 OTL design in his book "Audio Reality".
I've got one layed out on a chassis but have got a ways to go before I hear
anything.

Jon



in article , Simon at
wrote on 7/20/04 2:57 PM:

Hi,

I've build an integrated amp with this amp as the power amp. I liked the
design with no coupling caps and I had a few PL519's around. I used the
design before this revised design, with no connection between the grid and
ground, and maybe I should connect these, but then again: if it is not
broken, don't fix it.
I like the sound of it, I liked it even better then my previous SE EL84 amp.
I used the PL519 instead of the EL509 and a PCL82 instead of the ECL82 (So
you do not need to raise the V of the ECL82 heaters ( max V Cathode Heater
PCL82=200V)). I was also thinking of using PY500A rectifiers, but ended up
with 2 plain old EZ81 rectifiers.

With kind regards,

Simon

"si" ' schreef in bericht
...
Just wondered if you'd seen this. I'm just awaiting transformers for this
to build it myself!

http://www.svetlana.com/docs/products/glass/EL509.html

also, on the authors own home page

http://www.geocities.com/bobdanielak/technoteNo33.html

Has anyone else attempted this amp? For my novelty input, I'll be using
EY500 rectifiers!

Si.




"Adam Stouffer" wrote in message
...
Well I finally got around to trying some Svetlana 6KG6s for single ended
and heres the design I'm using:

http://68.162.128.102:808/~tesla/6KG6_SE.gif

The datasheet says 50Vp-p of drive and the 417A supplys it well. In
triode mode I get a clean 3 watts with a B+ of 300 at 75mA and 2v input.

Pentode mode is an entirely different matter. No matter what plate
voltage or how the tube is biased the output is always high in
distortion and the sinewave looks like garbage. The sweet spot was a
plate voltage of 250 at 65mA. Power output was 6w and distortion pretty
high at 7%.

Any ideas on how to improve pentode operation? I'm having a tough time
finding the right operating points. Thanks.


Adam

Hello "myself"!
Thanks for your input! I am still considering the preamp to go with this. As
I wish to include phono (MM) I will keep it seperate from the power amp,
most probably fed by good ol' umbilical. Several pairs of cascodes and mabe
a few CF's are amongst my favorite ideas, but keeping phase of both output
and tape out, including tone controls, and keeping number of valves 'down'
are also major considerations.
PL-series valves are very much undervalued. I keep all mine should I ever
go 'hybrid' and decide on things like DC heaters!
I'm surprised you got away with some earlier design with the grid
disconnected (I did not get to see this). My friend and I were looking at
the (newer) design overall and decided that the whole lot would go up if
this was not connected, and to be very diligent about this. I now suspect
the stability of the bias suffers with the connection not made?
Do you know much power out you can get on your amp? How does it sound?

Si


"Simon" wrote in message
l...
Hi,

I've build an integrated amp with this amp as the power amp. I liked the
design with no coupling caps and I had a few PL519's around. I used the
design before this revised design, with no connection between the grid and
ground, and maybe I should connect these, but then again: if it is not
broken, don't fix it.
I like the sound of it, I liked it even better then my previous SE EL84
amp.
I used the PL519 instead of the EL509 and a PCL82 instead of the ECL82 (So
you do not need to raise the V of the ECL82 heaters ( max V Cathode Heater
PCL82=200V)). I was also thinking of using PY500A rectifiers, but ended up
with 2 plain old EZ81 rectifiers.

With kind regards,

Simon

"si" ' schreef in bericht
...
Just wondered if you'd seen this. I'm just awaiting transformers for
this
to build it myself!

http://www.svetlana.com/docs/products/glass/EL509.html

also, on the authors own home page

http://www.geocities.com/bobdanielak/technoteNo33.html

Has anyone else attempted this amp? For my novelty input, I'll be using
EY500 rectifiers!

Si.




"Adam Stouffer" wrote in message
...
Well I finally got around to trying some Svetlana 6KG6s for single
ended
and heres the design I'm using:

http://68.162.128.102:808/~tesla/6KG6_SE.gif

The datasheet says 50Vp-p of drive and the 417A supplys it well. In
triode mode I get a clean 3 watts with a B+ of 300 at 75mA and 2v
input.

Pentode mode is an entirely different matter. No matter what plate
voltage or how the tube is biased the output is always high in
distortion and the sinewave looks like garbage. The sweet spot was a
plate voltage of 250 at 65mA. Power output was 6w and distortion
pretty
high at 7%.

Any ideas on how to improve pentode operation? I'm having a tough time
finding the right operating points. Thanks.


Adam

Hi,

The grid was not disconnected, but connected to the secundairy of the OPT.
The secundairy of the OPT is not connected to the ground. I guess that it
has more NFB now. That also explaines why this version is not very
sensistive, and I had to make an extra driver stage to drive it up (I tried
also a pcl86 instead of the pcl82) But then I got minor problems with the
operating points of the amp, so I changed it back. I saw the revised edition
when I already finished the amp... And I am just to lazy to make the
adjustment. Maybe I should change it because( if I remember well) , like you
wrote, the V of the kathode varied a bit.
It should give about 8 watts, I need to look up the measurements, or measure
them again. It has more than enough power for my 91db speakers.

How does it sounds? I'm not good in telling how it sounds, I have not much
to compare it with, but I found that it sounds better to me then my SE
penthode EL84 amp. I also liked the sound better with the OPT primairy as
4k8 ohm instead of 2k4 ohm.

Kind regards,


Simon


"si" ' schreef in bericht
...
Hello "myself"!
Thanks for your input! I am still considering the preamp to go with this.
As
I wish to include phono (MM) I will keep it seperate from the power amp,
most probably fed by good ol' umbilical. Several pairs of cascodes and
mabe
a few CF's are amongst my favorite ideas, but keeping phase of both output
and tape out, including tone controls, and keeping number of valves 'down'
are also major considerations.
PL-series valves are very much undervalued. I keep all mine should I ever
go 'hybrid' and decide on things like DC heaters!
I'm surprised you got away with some earlier design with the grid
disconnected (I did not get to see this). My friend and I were looking at
the (newer) design overall and decided that the whole lot would go up if
this was not connected, and to be very diligent about this. I now suspect
the stability of the bias suffers with the connection not made?
Do you know much power out you can get on your amp? How does it sound?

Si


"Simon" wrote in message
l...
Hi,

I've build an integrated amp with this amp as the power amp. I liked the
design with no coupling caps and I had a few PL519's around. I used the
design before this revised design, with no connection between the grid
and
ground, and maybe I should connect these, but then again: if it is not
broken, don't fix it.
I like the sound of it, I liked it even better then my previous SE EL84
amp.
I used the PL519 instead of the EL509 and a PCL82 instead of the ECL82
(So
you do not need to raise the V of the ECL82 heaters ( max V Cathode
Heater
PCL82=200V)). I was also thinking of using PY500A rectifiers, but ended
up
with 2 plain old EZ81 rectifiers.

With kind regards,

Simon

"si" ' schreef in bericht
...
Just wondered if you'd seen this. I'm just awaiting transformers for
this
to build it myself!

http://www.svetlana.com/docs/products/glass/EL509.html

also, on the authors own home page

http://www.geocities.com/bobdanielak/technoteNo33.html

Has anyone else attempted this amp? For my novelty input, I'll be
using
EY500 rectifiers!

Si.




"Adam Stouffer" wrote in message
...
Well I finally got around to trying some Svetlana 6KG6s for single
ended
and heres the design I'm using:

http://68.162.128.102:808/~tesla/6KG6_SE.gif

The datasheet says 50Vp-p of drive and the 417A supplys it well. In
triode mode I get a clean 3 watts with a B+ of 300 at 75mA and 2v
input.

Pentode mode is an entirely different matter. No matter what plate
voltage or how the tube is biased the output is always high in
distortion and the sinewave looks like garbage. The sweet spot was a
plate voltage of 250 at 65mA. Power output was 6w and distortion
pretty
high at 7%.

Any ideas on how to improve pentode operation? I'm having a tough
time
finding the right operating points. Thanks.


Adam

Patrick Turner wrote:

Indeed G2 supply should be much lower than Ea, and if so, IG2
is at its lowest value, and there is no need to regulated for
class A SE; just use 100uF from a resistance dropper from the B+,
and she will be as good as it gets with music operation.
But when tested with a sine wave, there will be a drop in EG2,
which will increase the thd, so shunt regulate for tests,
and so have say 3 x 65v x 5W zeners across a 100 Uf fed by R from the B+,
and this allows the G2 voltage top still sag if there is a tube problem and
IG2
increases to more than twice the idle value.

Thanks for the info everyone. I'll do some more tests with a lower G2
voltage and post the results.

Don't use series pass element regulators for G2.

Because it could start supplying too much current? How about an 0A2?


Adam

Adam Stouffer wrote:

Patrick Turner wrote:

Indeed G2 supply should be much lower than Ea, and if so, IG2
is at its lowest value, and there is no need to regulated for
class A SE; just use 100uF from a resistance dropper from the B+,
and she will be as good as it gets with music operation.
But when tested with a sine wave, there will be a drop in EG2,
which will increase the thd, so shunt regulate for tests,
and so have say 3 x 65v x 5W zeners across a 100 Uf fed by R from the B+,
and this allows the G2 voltage top still sag if there is a tube problem and
IG2
increases to more than twice the idle value.

Thanks for the info everyone. I'll do some more tests with a lower G2
voltage and post the results.

Pentode with lower Eg2 only marginally reduces thd, but it does reduce IG2,
and marginally increase plate efficiency, and this tube like many of that time
were
designed for low EG2 voltages.



Don't use series pass element regulators for G2.

Because it could start supplying too much current?

Indeed.


How about an 0A2?

It may be OK, but it turns into a relaxation oscillator if cap bypassed.
I'd prefer zeners, which I have often bypassed to ensure their silence.

Pentode operation nearly always needs the use of 20 dB of NFB
to reduce the Ro to 0.5 ohms.
One can have 10 dB in the output stage from a 12.5% CFB winding,
and 10 dB of loop FB, so the total FB is 20 dB.
But with more output stage, the driver should be fairly linear.
Its not critcal though with CFB because the driver 2H cancels the 2H in the
output stage
to some extent.

Ther total amount of applied FB is very high when RL is high, and very low when
RL is low, since pentode open loop gain is directly proportional to load.
So don't forget a zobel RC network across the OPT primary
to give the amp a load at HF, and to damp the resonant effects of leakage L and
shunt C in the OPT.

Patrick Turner.




Adam

I've officially given up on using these EL509s. They are too much of a
pain and normal "audio" tubes work much better. I tried the screen
driven version and it was horrible as well. The first time I fired it up
my scope showed a nice sinewave, at 41Mhz!

I've had these OPTs from an old Magnavox hifi sitting around for a
while. They used a quad of 6V6s so a pair of 6L6s run conservatively
should be a good match.

Again thanks for the help everyone.


Adam

I was wondering about something that may be related to this. I once saw in a
circuit from who-knows-where, a resistor (about magnitude of 100R) in the
anode circuit of the valve presumably to stop this kind of activity. If this
were mounted as close as possible to the anode cap (with obviously some very
good insulation covering) I would expect the oscillation to be considerably
reduced. However, Bob, who originally did the Svetlana screen coupled
circuit (see further down this post) did not appear to have this problem....

Si
"Adam Stouffer" wrote in message
...
I've officially given up on using these EL509s. They are too much of a
pain and normal "audio" tubes work much better. I tried the screen
driven version and it was horrible as well. The first time I fired it up
my scope showed a nice sinewave, at 41Mhz!

I've had these OPTs from an old Magnavox hifi sitting around for a
while. They used a quad of 6V6s so a pair of 6L6s run conservatively
should be a good match.

Again thanks for the help everyone.


Adam

si wrote:

I was wondering about something that may be related to this. I once saw in a
circuit from who-knows-where, a resistor (about magnitude of 100R) in the
anode circuit of the valve presumably to stop this kind of activity.

Very common in RF power amp circuits. Sometimes the resister has
several turns of small diameter wire (say # 30) wound around its body
& connected to each end. The whole assembly forms a parasitic
oscillation suppressor. Good Luck, John Stewart

If this
were mounted as close as possible to the anode cap (with obviously some very
good insulation covering) I would expect the oscillation to be considerably
reduced. However, Bob, who originally did the Svetlana screen coupled
circuit (see further down this post) did not appear to have this problem....

Si
"Adam Stouffer" wrote in message
...
I've officially given up on using these EL509s. They are too much of a
pain and normal "audio" tubes work much better. I tried the screen
driven version and it was horrible as well. The first time I fired it up
my scope showed a nice sinewave, at 41Mhz!

I've had these OPTs from an old Magnavox hifi sitting around for a
while. They used a quad of 6V6s so a pair of 6L6s run conservatively
should be a good match.

Again thanks for the help everyone.


Adam

"si" ' said:

I was wondering about something that may be related to this. I once saw in a
circuit from who-knows-where, a resistor (about magnitude of 100R) in the
anode circuit of the valve presumably to stop this kind of activity. If this
were mounted as close as possible to the anode cap (with obviously some very
good insulation covering) I would expect the oscillation to be considerably
reduced. However, Bob, who originally did the Svetlana screen coupled
circuit (see further down this post) did not appear to have this problem....

This is common practice with output tubes with top anodes.
Look at it as a oscillation stopper. I've seen this done on some 807
amplifiers. It replaces the G2 resistor.

--
Sander deWaal
"SOA of a KT88? Sufficient."

I just suppose that the 509 is a line-scan TV tube, not designed for
linearity but for power and efficiency, and this should be the way to use
it: PP pentode, adequate A-A load, very high anode B+ and tons of feedback
to get the distortion back to a sensible figure. AFAIK there was a US made
electrostatic LS manufacturer (Acoustat) who made a giant unit with a
built-in tubed power amp (some 150-200W per unit); the power tubes they used
were US horizontal deflection tetrodes, 36LW6s or some stuff like that,
connected as PP tetrodes, and the whole stuff sounded great, as far as I can
remember (I just heard 'em at a HiFi show in Milan some 20+ yrs ago..).
Well, they had a price tag around 6'000$ in 1980, so they HAD to sound at
least decent.....
If anybody deserves good sound from a simple circuit, without or with little
NFB, there's a helluva "plain audio" tubes from 7W ECL82 to 125W GM-70,
crossing by the way the rust-free 6L6GC and 6550; after all, who'd choose a
truck for racing at 200 mph or a Ferrari for loading 20 tons of concrete?

Ciao

Fabio


"Patrick Turner" ha scritto nel messaggio
...


Adam Stouffer wrote:

Well I finally got around to trying some Svetlana 6KG6s for single ended
and heres the design I'm using:

http://68.162.128.102:808/~tesla/6KG6_SE.gif

The datasheet says 50Vp-p of drive and the 417A supplys it well. In
triode mode I get a clean 3 watts with a B+ of 300 at 75mA and 2v input.

In triode, with the 22.5 watts of dissipation, you should get 30%
efficiency,
which is 6.75 watts at around 7% thd.

I don't like this tube much at all unless one uses it in PP
with lots of class A and with cathode FB windings on the OPT,
and then its starts to clean up its act; see below about SECFB.



Pentode mode is an entirely different matter. No matter what plate
voltage or how the tube is biased the output is always high in
distortion and the sinewave looks like garbage. The sweet spot was a
plate voltage of 250 at 65mA. Power output was 6w and distortion pretty
high at 7%.

The anode dissipation you have is only 16.3 watts, so try 300v at 90 mA,
and a lower value RL, and try running the screem at some point between
+150v
and +200v.
You should get 10.8 watts max, at 40% efficiency.



Any ideas on how to improve pentode operation? I'm having a tough time
finding the right operating points. Thanks.

Traditional SE Pentode operation of EL509 is always going to be a pita.

For good power, and best linearity, use screen drive, with G1 taken tom
0V,
and have around 27 watts of anode dissipation, 300v x 90 mA will do.
The screen needs a direct coupled cathode follower with a preceding
voltage
amp
capable of around 25vrms , or 75vrms for low driver thd.
See enhanced triode operation the svetlana site.

Better still is the use of a distributed loading on the primary of the
OPT,
so that you have 80% of the P turns between B+ and anode,
and 20% between cathode and 0V, which is an application of 20%
of the voltage across the tube being cathode feedback.
Ea = +300 v, Eg2 = +150v( a fixed voltage, Ia = 90 mA,
and po should be about 9 watts.
This should give Ro slightly lower than triode, and around 2% max thd,
and need around the same drive voltage as triodes.

The driver could be a normal triode resistance loaded stage using 12BH7
paralleled, which should give around 1% thd at the max drive voltage,
which will cancel half the 509 thd at most load values, especially the
lower

load values, since the main thd is 2H.

Patrick Turner.







Adam

The major THD component is 2H

To me the the 6KG6/EL509 is fine as a ham/cb linear output tube...not an
audio
output tube.....George


"Fabio Berutti" wrote in message
...
I just suppose that the 509 is a line-scan TV tube, not designed for
linearity but for power and efficiency, and this should be the way to use
it: PP pentode, adequate A-A load, very high anode B+ and tons of feedback
to get the distortion back to a sensible figure. AFAIK there was a US
made
electrostatic LS manufacturer (Acoustat) who made a giant unit with a
built-in tubed power amp (some 150-200W per unit); the power tubes they
used
were US horizontal deflection tetrodes, 36LW6s or some stuff like that,
connected as PP tetrodes, and the whole stuff sounded great, as far as I
can
remember (I just heard 'em at a HiFi show in Milan some 20+ yrs ago..).
Well, they had a price tag around 6'000$ in 1980, so they HAD to sound at
least decent.....
If anybody deserves good sound from a simple circuit, without or with
little
NFB, there's a helluva "plain audio" tubes from 7W ECL82 to 125W GM-70,
crossing by the way the rust-free 6L6GC and 6550; after all, who'd choose
a
truck for racing at 200 mph or a Ferrari for loading 20 tons of concrete?

Ciao

Fabio


"Patrick Turner" ha scritto nel messaggio
...


Adam Stouffer wrote:

Well I finally got around to trying some Svetlana 6KG6s for single
ended
and heres the design I'm using:

http://68.162.128.102:808/~tesla/6KG6_SE.gif

The datasheet says 50Vp-p of drive and the 417A supplys it well. In
triode mode I get a clean 3 watts with a B+ of 300 at 75mA and 2v
input.

In triode, with the 22.5 watts of dissipation, you should get 30%
efficiency,
which is 6.75 watts at around 7% thd.

I don't like this tube much at all unless one uses it in PP
with lots of class A and with cathode FB windings on the OPT,
and then its starts to clean up its act; see below about SECFB.



Pentode mode is an entirely different matter. No matter what plate
voltage or how the tube is biased the output is always high in
distortion and the sinewave looks like garbage. The sweet spot was a
plate voltage of 250 at 65mA. Power output was 6w and distortion
pretty
high at 7%.

The anode dissipation you have is only 16.3 watts, so try 300v at 90 mA,
and a lower value RL, and try running the screem at some point between
+150v
and +200v.
You should get 10.8 watts max, at 40% efficiency.



Any ideas on how to improve pentode operation? I'm having a tough time
finding the right operating points. Thanks.

Traditional SE Pentode operation of EL509 is always going to be a pita.

For good power, and best linearity, use screen drive, with G1 taken tom
0V,
and have around 27 watts of anode dissipation, 300v x 90 mA will do.
The screen needs a direct coupled cathode follower with a preceding
voltage
amp
capable of around 25vrms , or 75vrms for low driver thd.
See enhanced triode operation the svetlana site.

Better still is the use of a distributed loading on the primary of the
OPT,
so that you have 80% of the P turns between B+ and anode,
and 20% between cathode and 0V, which is an application of 20%
of the voltage across the tube being cathode feedback.
Ea = +300 v, Eg2 = +150v( a fixed voltage, Ia = 90 mA,
and po should be about 9 watts.
This should give Ro slightly lower than triode, and around 2% max thd,
and need around the same drive voltage as triodes.

The driver could be a normal triode resistance loaded stage using 12BH7
paralleled, which should give around 1% thd at the max drive voltage,
which will cancel half the 509 thd at most load values, especially the
lower

load values, since the main thd is 2H.

Patrick Turner.







Adam

The major THD component is 2H

Adam Stouffer wrote:
I've officially given up on using these EL509s. They are too much of a
pain and normal "audio" tubes work much better. I tried the screen
driven version and it was horrible as well. The first time I fired it up
my scope showed a nice sinewave, at 41Mhz!



Try grid stoppers. Those are just resistors of about 2K wired right
at the tube socket G1 pin. In series with the coupling cap and
biasing resistor circuit node. Another suggestion is the use of a
very small cap from G1 to the cathode, something like 50pF. This
should kill the tube's ability to do anything at supersonic or RF
frequencies.

george craig wrote:
To me the the 6KG6/EL509 is fine as a ham/cb linear output tube...not an
audio
output tube.....George




Any sharp cutoff triode or pentode should work fine for audio. Assuming
proper biasing, B+ and load. High Gm tubes do need special care to
avoid supersonic or RF oscillations. Grid stopper resistors usually
take care of this.

Robert Casey wrote:

george craig wrote:
To me the the 6KG6/EL509 is fine as a ham/cb linear output tube...not an
audio
output tube.....George



Any sharp cutoff triode or pentode should work fine for audio. Assuming
proper biasing, B+ and load. High Gm tubes do need special care to
avoid supersonic or RF oscillations. Grid stopper resistors usually
take care of this.

Tim DeParavicini used a pair of PL509 for his 100 watt
monos with a circuit like the McIntosh, running the tube in near
class B conditions, PP.
I saw no sign of parasitic oscillation.
The amount of NFB used totalled 44 dB into the rated load,
in 3 applied loops.

Someone here said screen drive is prone to RF oscillations, in which case,
take what steps are necessary to stop them, and that isn't too hard.

But this tube is OK if used with CFB windings, like in a Quad II output stage,

and where the applied G2 voltage can be kept low.

For PP, class A working with CFB is very good, and a search
of thew archives will reveal a lot about EL509 and CFB over the last few
years.

But for SE, the CFB option is still OK.

Patrick Turner.