[TerryG]

Hi all,

I am new to this forum but have known several members here for quite some time. I know not many of you will recognise me from this handle, but you will I know.

Anyway I have been enjoying SEP amplifiers this last year an 1/2. Don't know why they have the reputation they do compared to SET's, but that is another issue.

I have designed and built several pentode single ended amplifiers at this point. And the one point that still alludes me is: How from what the datasheets say can you figure out what resistor to start with on g2?

I understand the DC voltage value and that is not difficult to figure out, but I like pentodes best when there is an unbypass resistor on g2.

I have found this does several things, it moves the top bias lines down, and the bottom bias lines up at the same time raising ( or lowering depending on how you look at it) the cut off bias voltage. Usually by putting a unbypassed g2 resistor on g2 your current drops compared to bypassing, and a little more wattage is possible for the same operating point. It seems most g2's have a certain impedance, EL84s you can start with 1.5K and go to above 10K with the resistor have more effect on the signal traces. This also lowers distortion it appears, but there is always a point of diminished returns for each tube.

It sure would be nice to be able to do a few math problem when looking at datasheets and know where to start with a tube.

I will introduce a controversial subject with g2 later, but for now I just want to know what information on the datasheet will tell me what I want to know.

Terry

[Roger Jones]

On May 16, 7:18*pm, TerryG wrote:
Hi all,

I am new to this forum but have known several members here for quite
some time. *I know not many of you will recognise me from this handle,
but you will I know.

Anyway *I have been enjoying SEP amplifiers this last year an 1/2.
Don't know why they have the reputation they do compared to SET's, but
that is another issue.

I have designed and built several pentode single ended amplifiers at
this point. *And the one point that still alludes me is: *How from what
the datasheets say can you figure out what resistor to start with on g2?

I understand the DC voltage value and that is not difficult to figure
out, but I like pentodes best when there is an unbypass resistor on g2.

I have found this does several things, it moves the top bias lines down,
and the bottom bias lines up at the same time raising ( or lowering
depending on how you look at it) the cut off bias voltage. *Usually by
putting a unbypassed g2 resistor on g2 your current drops compared to
bypassing, and a little more wattage is possible for the same operating
point. *It seems most g2's have a certain impedance, EL84s you can start
with 1.5K and go to above 10K with the resistor have more effect on the
signal traces. *This also lowers distortion it appears, but there is
always a point of diminished returns for each tube.

It sure would be nice to be able to do a few math problem when looking
at datasheets and know where to start with a tube.

I will introduce a controversial subject with g2 later, but for now I
just want to know what information on the datasheet will tell me what I
want to know.

Terry

--
TerryG

Terry, I've played with quite a few Single-Ended Pentode (SEP) amps
over the years...in my student days it was for economic reasons! I
still do them now in vintage radio restoration. I sometimes use a g2
resistor to ensure the g2 voltage does not exceed the plate voltage
(OPT primary resistance drops the latter a bit.) I do the same in
pentode/beam tetrode P-P amps for the same reason (if they're not UL)
- just done it with an Eico HF12 (EL84's); actually I just moved the
g2 tap further down the B+ filter and adjusted the R values to get the
right voltages back. My reasoning is to reduce screen power and
heating. However, in both, I always decouple the g2 resistor (above B+
filter does it anyway.) I would expect this to change the dynamic
operating point, tube impedance and max. power wrt a naked R...
obviously, it can't change the DC operation. But I've never studied
the effect... mea culpa! I trust the tube modellers will help us.
Cheers,
Roger

[Roger Jones]

On May 18, 5:26*pm, Roger Jones wrote:
On May 16, 7:18*pm, TerryG wrote:









Hi all,

I am new to this forum but have known several members here for quite
some time. *I know not many of you will recognise me from this handle,
but you will I know.

Anyway *I have been enjoying SEP amplifiers this last year an 1/2.
Don't know why they have the reputation they do compared to SET's, but
that is another issue.

I have designed and built several pentode single ended amplifiers at
this point. *And the one point that still alludes me is: *How from what
the datasheets say can you figure out what resistor to start with on g2?

I understand the DC voltage value and that is not difficult to figure
out, but I like pentodes best when there is an unbypass resistor on g2.

I have found this does several things, it moves the top bias lines down,
and the bottom bias lines up at the same time raising ( or lowering
depending on how you look at it) the cut off bias voltage. *Usually by
putting a unbypassed g2 resistor on g2 your current drops compared to
bypassing, and a little more wattage is possible for the same operating
point. *It seems most g2's have a certain impedance, EL84s you can start
with 1.5K and go to above 10K with the resistor have more effect on the
signal traces. *This also lowers distortion it appears, but there is
always a point of diminished returns for each tube.

It sure would be nice to be able to do a few math problem when looking
at datasheets and know where to start with a tube.

I will introduce a controversial subject with g2 later, but for now I
just want to know what information on the datasheet will tell me what I
want to know.

Terry

--
TerryG

Terry, I've played with quite a few Single-Ended Pentode (SEP) amps
over the years...in my student days it was for economic reasons! *I
still do them now in vintage radio restoration. I sometimes use a g2
resistor to ensure the g2 voltage does not exceed the plate voltage
(OPT primary resistance drops the latter a bit.) *I do the same in
pentode/beam tetrode P-P amps for the same reason (if they're not UL)
- just done it with an Eico HF12 (EL84's); actually I just moved the
g2 tap further down the B+ filter and adjusted the R values to get the
right voltages back. My reasoning is to reduce screen power and
heating. However, in both, I always decouple the g2 resistor (above B+
filter does it anyway.) *I would expect this to change the dynamic
operating point, tube impedance and max. power wrt a naked R...
obviously, it can't change the DC operation. But I've never studied
the effect... mea culpa! *I trust the tube modellers will help us.
Cheers,
Roger

Good replies... many thanks. I conclude that a lower g2 voltage (than
plate), fully decoupled, is advantageous. FWIW, in vintage radio
restoration (keeping same 6K6, 41, etc), I invariably add NFB from OPT
secondary to the cathode of the triode AF tube and mostly replace the
detector diode with a 1N34A point contact s/s diode. Audio gain goes
down, of course, but since I just listen to local AM radio on these
sets, no problem. Actually, it makes the range on the volume control
better... it's further "off the end".
Cheers,
Roger

[Patrick Turner]

On May 21, 11:02*pm, Roger Jones wrote:
On May 18, 5:26*pm, Roger Jones wrote:





On May 16, 7:18*pm, TerryG wrote:

Hi all,

I am new to this forum but have known several members here for quite
some time. *I know not many of you will recognise me from this handle,
but you will I know.

Anyway *I have been enjoying SEP amplifiers this last year an 1/2.
Don't know why they have the reputation they do compared to SET's, but
that is another issue.

I have designed and built several pentode single ended amplifiers at
this point. *And the one point that still alludes me is: *How from what
the datasheets say can you figure out what resistor to start with on g2?

I understand the DC voltage value and that is not difficult to figure
out, but I like pentodes best when there is an unbypass resistor on g2.

I have found this does several things, it moves the top bias lines down,
and the bottom bias lines up at the same time raising ( or lowering
depending on how you look at it) the cut off bias voltage. *Usually by
putting a unbypassed g2 resistor on g2 your current drops compared to
bypassing, and a little more wattage is possible for the same operating
point. *It seems most g2's have a certain impedance, EL84s you can start
with 1.5K and go to above 10K with the resistor have more effect on the
signal traces. *This also lowers distortion it appears, but there is
always a point of diminished returns for each tube.

It sure would be nice to be able to do a few math problem when looking
at datasheets and know where to start with a tube.

I will introduce a controversial subject with g2 later, but for now I
just want to know what information on the datasheet will tell me what I
want to know.

Terry

--
TerryG

Terry, I've played with quite a few Single-Ended Pentode (SEP) amps
over the years...in my student days it was for economic reasons! *I
still do them now in vintage radio restoration. I sometimes use a g2
resistor to ensure the g2 voltage does not exceed the plate voltage
(OPT primary resistance drops the latter a bit.) *I do the same in
pentode/beam tetrode P-P amps for the same reason (if they're not UL)
- just done it with an Eico HF12 (EL84's); actually I just moved the
g2 tap further down the B+ filter and adjusted the R values to get the
right voltages back. My reasoning is to reduce screen power and
heating. However, in both, I always decouple the g2 resistor (above B+
filter does it anyway.) *I would expect this to change the dynamic
operating point, tube impedance and max. power wrt a naked R...
obviously, it can't change the DC operation. But I've never studied
the effect... mea culpa! *I trust the tube modellers will help us.
Cheers,
Roger

Good replies... many thanks. *I conclude that a lower g2 voltage (than
plate), fully decoupled, is advantageous. FWIW, in vintage radio
restoration (keeping same 6K6, 41, etc), I invariably add NFB from OPT
secondary to the cathode of the triode AF tube and mostly replace the
detector diode with a 1N34A point contact s/s diode. *Audio gain goes
down, of course, but since I just listen to local AM radio on these
sets, no problem. *Actually, it makes the range on the volume control
better... it's further "off the end".

Yikes, WTF are you discussing. Please post a schematic somewhere so a
picture says a thousands words you can't seem to type.

As one reduces Eg2, the G2 biasing Vdc, then the set of Ra curves for
Ia vs Ea for given Eg1 values changes. Most ppl won't use AB2 and the
limit of operation and maximum PO is determined by the Ra curve for
Eg1 = 0.0V. This curve rises steeply at first from 0V, 0A, often on a
slope = 200 ohms, then with Ea at about 50V to 100V the line rolls
over to the right and assumes a slope of perhaps 38k, for an EL84. The
KNEE of the curve and the height of the flat part of the curve above
Ea = 100V appear lower on the graph as Eg2 is reduced.

For class AB and highest possible PO, Eg2 is kept as high as possible
to get maximal Ia swing into minimal RLa which is 1/4 x RLa-a for AB
operation. But for class A operation the RLa = 1/2 RLa-a and Ia change
for this higher RLa value is LESS than for class AB, so you **don't
need** to have Eg2 as close to Ea as possible. This is especially true
for SE pentodes.

Tube data from old books etc often show 3 sets of Ra curves for
differing values of fixed Eg2, and for an EL34 in SE mode, Ea might be
420V, but Eg2 at only 300V. This means Ig2 is much lower than if Eg2 =
Ea = +420V. One will find that the class A performance is no worse for
the lower Eg2, ie, Ra and THD is not significantly higher, see my
pages on SE35 to find out more.
http://www.turneraudio.com.au/se35cfb-monobloc.htm

I use CFB windings wherever I can for maximum linearity, and to get Ra
effectively lower than triode before adding a little GNFB.

The lower Eg2 means that Eg1 can be lower, so that for cathode biasing
there is less power wasted in Rk because Ek is low.

For high PP AB PO, Eg2 is high as possible, and this causes Ia dc to
increase, so Eg1 must be increased to counter the effect of high Eg2,
but of course this is OK because one wants a high Vg1 swing without
getting any grid current.

In guitar amps, you often see 6L6GC with Ea = +450V, and Eg2 a fixed
+440V, and grid stopper series R to G2 are often 470 ohms, 1 Watt.
Some Rg2 are special fusable resistors and they look like they have a
high Watt rating but actually don't. When they fuse, there are no
flames and smoke. Normal G2 power input at idle might be 4mA x 440V =
1.76 Watts, and so Pd in 470 ohms is 0.0075 Watt only, so the 470 ohms
could have a rating of 0.25 Watts, just a small metal film R. If the R
heat went to 1W, R would fail, and to cause this the Ig2 would be
46mA,
and I have seen such G2 resistors go open circuit often where an OP
tube has gone into thermal runaway, red hot anode with Ia dc = 300mA
perhaps, and hence Ig2 increases hugely.
Once Rg2 fuses open, Eg2 collapses to 0V, and Ia is cut off. I've used
low power rated Rg2 as an extra layer of protection against bias
failure. It isn't a good protection measure because after Rg2 fuses
open someone has to solder in a new Rg2, and the best form of
protection is via active monitoring of idle bias current and a circuit
which automatically turns the amp off when Ia dc exceeds 1.5 x idle
value for longer than 4 seconds. My website has details.

Patrick Turner.

[Phil Allison[_3_]]

"TerryG"

I am new to this forum but have known several members here for quite
some time. I know not many of you will recognise me from this handle,
but you will I know.

Anyway I have been enjoying SEP amplifiers this last year an 1/2.
Don't know why they have the reputation they do compared to SET's, but
that is another issue.


** Just curious - but do you use any NFB with your SEP designs ?

BTW:

the first valve amp I built used a 6V6GT and a decent size SE tranny from
A&R of Melbourne.

The amp also used a war surplus 6AC7 as a preamp and a 5Y3 rectifier.

http://www.r-type.org/exhib/aaa0053.htm



..... Phil

[Patrick Turner]

On May 19, 5:29*pm, "Phil Allison" wrote:
"TerryG"



I am new to this forum but have known several members here for quite
some time. *I know not many of you will recognise me from this handle,
but you will I know.

Anyway *I have been enjoying SEP amplifiers this last year an 1/2.
Don't know why they have the reputation they do compared to SET's, but
that is another issue.

** Just curious - but do you use any NFB with your SEP designs ?

BTW:

the first valve amp I built used a 6V6GT and a decent size SE tranny from
A&R of Melbourne.

The amp also used a war surplus 6AC7 as a preamp and a 5Y3 rectifier.

http://www.r-type.org/exhib/aaa0053.htm

.... * Phil

The screen of a pentode or beam tetrode, ie, G2, acts like a
electrostatic screen to prevent the change of voltage causing much
effect on the electron flow. If G2 is held at an unchanging potential
Vdc between cathode and G2 then the tube exhibits the high Ra, and
high amplification factor µ.

If you consider that G2 is bypassed to cathode, and cathode is
bypassed to 0V, then while the tube is operated by G1 grid, there is a
signal flow from B+ to G2.
If there is an added resistance between G2 and B+, and this resistance
is not bypassed to cathode, then a signal voltage is developed across
the R which has an effect on the electron flow to reduce the gain of
the tube below the fully bypassed state.

Let us say you have a 6550 with Ia at 55mA, Ea = 450V and Eg2 = 350V,
you would find that G1 gm, transconductance, = 5.5mA/V approx. You
would also find that if you allow the signal voltage between G2 and
cathode, VG2-k, to vary, then the G2 screen has gm = about 0.83mA/V,
considerably less than the G1, but very significant in audio amps
especially where Ultralinear operation or triode operation s
considered. To really understand the what is happening inside a
pentode I suggest you try to undertsand the *equivalent models* of the
tube.
I have a number of web pages which may allow a better understanding,
see http://www.turneraudio.com.au/basic-tube-1.html
This page plus many others following might be of use.

One of the simplest models of a pentode for its signal ac operation
without being
confused by the DC operation is to **imagine** it this way :-
Ra is a resistance between anode terminal and cathode terminal, about
32,000 ohms
with the 6550 sample above.
Then you have a current generator connected between anode terminal and
cathode, This CS is an imaginery device with high grid input impedance
but with infinitely high output impedance which produces current at
the rate of G1 Vin x gm, ie, 5.5mA ac change for 1Vac Vin. The Screen
is also a CS strapped across anode terminal and cathode and also
shunting the 32k.

While ever VG2 is without any Vac relative to cathode, the Ra measured
is always going to be just 32k, because the two CS I've mentioned have
infinite R. Hence tube gain = Va / Vg = G1 gm x ( Ra in parallel with
RL ).

But if you connect the anode to the screen, then whatever Vac is
applied to the G2 changes tube current and current in Ra according to
screen gm x anode Vac.
When you work out what is happening simultaneously, you should find
the combined effect of Vac applied to G1 and G1 gives triode
operation, and the measured Ra and µ are both very much reduced, along
with distortion spectra, so triodes are favoured for hi-fi.

With a fairly large value of unbypassed R between B+ and G2, say more
than 1k5, then you may find the distortion and Ra are not much reduced
and less maximum PO is available and the operation is
dissapointing.

Pentodes tend to give much more circuit gain than triodes, especially
with high gm input pentods like the 6CA7 to which Phil refers, and
which has a very high "figure of merit" so it could be used in many RF
amps and AF apps. I have one used as a gain tube ahead of a 6CM5
cathode follower buffer which together form the second stage after a
12AT7 differential gain stage in a 6 band Wien Bridge oscillator I
made about 10 years ago. After 6CA7 which was around in WW2, many
other pentodes with high gm became available. Some, depending on the
use, could be unstable because of high gain, and then the series R to
G2 was employed to help prevent oscillations at HF, because there is
significant C between G2 and G1, and with some added R you have a low
pass shunt FB filter which tends to stop oscillations above F where
the tube is intended to operate.

I'm not much of a 6V6 fan, so when I do restore an old radio, I might
junk the 5Y3, install Si diodes, then put in an EL34 in triode with
same Ia but higher available B+, and the OPT will usually work well
with the EL34 in triode. The EL34 goes into the 5Y3 socket, and a
6SH7, 6SJ7, 6CA7, 6SL7 can be used where the 6V6 was, and a useful
amount of global NFB applied, about 12dB is plenty, and the you have a
really nice sounding SE amp where before no NFB at all with 6V6, you
had 10 times the THD/IMD as the EL34 in triode. I would much rather
listen with 4Watts from an EL34 in triode with GNFB than a poor 3Watts
from 6V6 in pure beam tetrode. One cannot usually avoid the common 20%
widing losses in old radio OPTs, but often their speakers are exremely
sensitive and 4 Watts for a 12" Rola DeLux hi-fi speaker made in Oz in
1953 will fill a kitchen with plate rattling bass tones and music to
make the glass of shiraz taste all the better.

Patrick Turner.

[flipper]

On Wed, 16 May 2012 23:18:38 +0000, TerryG
wrote:


Hi all,

I am new to this forum but have known several members here for quite
some time. I know not many of you will recognise me from this handle,
but you will I know.

Anyway I have been enjoying SEP amplifiers this last year an 1/2.
Don't know why they have the reputation they do compared to SET's, but
that is another issue.

Because, all else being equal (which it seldom is), a SE pentode will
have higher output impedance (worse damping factor) and more
distortion than a SET. The saving graces are it takes less input to
drive and produces more output power.

I have designed and built several pentode single ended amplifiers at
this point. And the one point that still alludes me is: How from what
the datasheets say can you figure out what resistor to start with on g2?

I understand the DC voltage value and that is not difficult to figure
out, but I like pentodes best when there is an unbypass resistor on g2.

I think you answered your own question. The common purposes of a G2
resistor are to bias, which you note is easy enough to determine from
the existing data, and to kill parasitic oscillations, which are not
'static' characteristics and, so, not something for a static graph.

An unpypassed G2, of the nature you're describing, is not 'typical'
operation.

I have found this does several things, it moves the top bias lines down,
and the bottom bias lines up at the same time raising ( or lowering
depending on how you look at it) the cut off bias voltage. Usually by
putting a unbypassed g2 resistor on g2 your current drops compared to
bypassing, and a little more wattage is possible for the same operating
point.

G2 current will be less because you are, in effect, applying negative
feedback to the screen with an unbypassed G2 resistor. Plate current
will also be less because you are, in effect, applying negative
feedback to the screen with an unbypassed G2 resistor.

I don't know how you arrive at the 'more wattage' conclusion.

As a side note, the 'opposite polarity' Rg2 derived screen signal was
used for the, so called, "self split" (Class A) PP amplifiers. It
drives the second grid so you don't need 'another tube'.

Negative feedback is also why the SET has lower impedance, lower
distortion, less power and also why it takes more signal to drive it.
Ia of a triode depends on *BOTH* G1 and plate voltages so if you stick
an impedance between plate and B+ the generated signal is negative
feedback to the plate. That is precisely what the 'fixed' screen
voltage in a pentode prevents, plate voltage feedback, hence it's
higher plate impedance, gain, and distortion. Putting an unbypassed
screen resistor on G2 means you're applying 'a little' of the feedback
a triode would see, plus lowering screen voltage.

It seems most g2's have a certain impedance,

Clearly it has impedance or else screen voltage would be 0 with any
resistance between G2 and B+.

EL84s you can start
with 1.5K and go to above 10K with the resistor have more effect on the
signal traces. This also lowers distortion it appears,

Negative feedback.

but there is
always a point of diminished returns for each tube.

Sure, because the larger Rg2 the lower screen voltage, lower Ia, and
less power.

The 'better' way of doing it is UL. You still get less power but at
least screen voltage isn't also reduced so you only 'sacrifice' that
which goes directly to reducing distortion. (Allegedly there's a
'sweet spot' ratio where 'maximum benefit' of both is achieved).

It sure would be nice to be able to do a few math problem when looking
at datasheets and know where to start with a tube.

The rest is usually contained in the 'typical' operating
characteristics or example circuits.

I will introduce a controversial subject with g2 later, but for now I
just want to know what information on the datasheet will tell me what I
want to know.

I imagine that, with enough perseverance, one could derive screen
current magnitude, the resulting Rg2 voltage, and the degree of
feedback but it's not 'straight on the graphs' because they don't
expect them to be operated that way.

Terry

Something that might be interesting to try would be a sort of 'poor
man's UL' with a tube like the 6CW5 that can use 250V B+ but is screen
limited to 200V. There a 'significant size' unbypassed Rg2 of 10k
would put screen right at the spec'd 200 V.

[Patrick Turner]

On May 20, 2:05*pm, flipper wrote:
On Wed, 16 May 2012 23:18:38 +0000, TerryG

wrote:

Hi all,

I am new to this forum but have known several members here for quite
some time. *I know not many of you will recognise me from this handle,
but you will I know.

Anyway *I have been enjoying SEP amplifiers this last year an 1/2.
Don't know why they have the reputation they do compared to SET's, but
that is another issue.

Because, all else being equal (which it seldom is), a SE pentode will
have higher output impedance (worse damping factor) and more
distortion than a SET. The saving graces are it takes less input to
drive and produces more output power.

I have designed and built several pentode single ended amplifiers at
this point. *And the one point that still alludes me is: *How from what
the datasheets say can you figure out what resistor to start with on g2?

I understand the DC voltage value and that is not difficult to figure
out, but I like pentodes best when there is an unbypass resistor on g2.

I think you answered your own question. The common purposes of a G2
resistor are to bias, which you note is easy enough to determine from
the existing data, and to kill parasitic oscillations, which are not
'static' characteristics and, so, not something for a static graph.

An unpypassed G2, of the nature you're describing, is not 'typical'
operation.

I have found this does several things, it moves the top bias lines down,
and the bottom bias lines up at the same time raising ( or lowering
depending on how you look at it) the cut off bias voltage. *Usually by
putting a unbypassed g2 resistor on g2 your current drops compared to
bypassing, and a little more wattage is possible for the same operating
point.

G2 current will be less because you are, in effect, applying negative
feedback to the screen with an unbypassed G2 resistor. Plate current
will also be less because you are, in effect, applying negative
feedback to the screen with an unbypassed G2 resistor.

I don't know how you arrive at the 'more wattage' conclusion.

As a side note, the 'opposite polarity' Rg2 derived screen signal was
used for the, so called, "self split" (Class A) PP amplifiers. It
drives the second grid so you don't need 'another tube'.

Negative feedback is also why the SET has lower impedance, lower
distortion, less power and also why it takes more signal to drive it.
Ia of a triode depends on *BOTH* G1 and plate voltages so if you stick
an impedance between plate and B+ the generated signal is negative
feedback to the plate. That is precisely what the 'fixed' screen
voltage in a pentode prevents, plate voltage feedback, hence it's
higher plate impedance, gain, and distortion. Putting an unbypassed
screen resistor on G2 means you're applying 'a little' of the feedback
a triode would see, plus lowering screen voltage.

*It seems most g2's have a certain impedance,

Clearly it has impedance or else screen voltage would be 0 with any
resistance between G2 and B+.

EL84s you can start
with 1.5K and go to above 10K with the resistor have more effect on the
signal traces. *This also lowers distortion it appears,

Negative feedback.

I've always wondered how much good is the "negative feedback" that is
obtained in this very simple manner.

First of all, you ought to define what sort of NFB, and where there is
a series R to G2 then to me the NFB must be negative current FB
because the more Ia you get, the more screen current flows so the Vg2
- k signal voltage increases, so the screen resistor sort of works
like an unbypassed Rk, except that the screen current which generates
the Vg2-k signal voltage is fairly non linear.

Nobody has ever provided test results to show the resulting Ra if you
apply say 1Vrms to G1, and have say 5k0 primary load on an OPT and
then vary RG2 from zero to say 10k0 or say 1 x EL34 in SEP. Everyone
so far has found the use of large value RG2 to give worse performance
than can be gathered by other conventional well known circuit
arrangements.

Having unbypassed Rk gives negative current FB and Ra rises,
distortion reduces according to the change in closed loop gain. The
open loop pentode gain will always be highest where G2 is well
bypassed to cathode, but sometimes one sees a pentode deliberately set
up with well bypassed G2 to k, and with fairly high Rk, and such a
thing has very high effecxtive Ra and the Rout from the pentode anode
is mainly determined by the value of dc carrying anode RL and
following cap coupled RL in parallel.

Take an EL84 in pentode for a preamp. One might have Ia at 20mA, RL dc
= 20k0, cap coupled to a volume pot of 50k0. So RL total = 8.33k. Open
loop gain with G2 well bypassed to k is going to be about µ x RL /
( RL + Ra ) = roughly 400 x 8k33 / ( 8k33 + 70k0 ) = 42.4,
approximately. One might find that this could produce about 50Vrms
easily at anode, but THD spectra looks quite ****house, with
2,3,4,5,6,7,8,9H all present at all levels and too high to sound very
well, although tolerable at very low levels when gain is turned up.

For a line level preamp, open loop gain would better be at about say
4.2x, ie, about 1/10 of the open loop gain of the pentode. So, for say
4.15Vrms ouput into 8.3k anode load, we would need need just 0.1Vrms
applied between g1 and cathode, so to reduce closed loop gain to 4.2
the Grid input to 0V signal has to be 4.15 / 4.2 = 0.988Vrms, and the
current FB voltage at k must be 0.988V - 0.1V = 0.888V. The cathode
current = anode current + screen current, and we don't know what the
lesser screen current is, so lets just neglect if for now. Cathode
current might be about 4.15V / 8.33k = 0.5mA, so Rk MUST be
approximately 0.888V / 0.5mA = 1.776k, so let's say 1k8.

In the real world, using Rk = 1k8 with Ik dc at say about 25mA
( including both Ia dc and IG2 dc ) means Ek = +45Vdc, and biasing for
G1 still needs to be maybe -12Vdc, so G1 must be fed a bias voltage of
around +33Vdc, and of course B+ and Eg2 must be raised by +45V to
accomodate the high Ek. Its all very well me talking about doing
simple things in circuits, but always the biasing and idle conditions
must all be changed in response to anything that is changed with Rk.
Don't ever expect total perfect spoon feeding from me because I expect
anyone reading the posts here to have a pair of ears exuding steam
because of high power thinking about the Overall Picture, and Aim For
The Mission, and all curly awkward details.

The THD/IMD will be reduced around about -17dB, not just by a factor
of -20dB ie, by 1/10, which is the gain reduction factor we have aimed
for. I am simplifying the analysis a bit, and the only REAL WAY to
findoutabout is to get away from the PC and go solder up a circuit and
measure it all.

So, if THD at 50Vrms output with no NCFB was a typical 7%, then with
NCFB it would be about 0.08% at 50Vrms and at 0.5Vrms we might expect
0.008% because THD reduces about proportionately to Va. So, if you
have closed loop gain of 4.2 instead of 42.0, and you have a CD player
producing an average level of 0.2Vrms, not uncommon, then average Va =
0.84, and average THD might be about 0.01%, and regardless of the
volume control setting the THD will always be 0.01%, and I'd say most
ppl would find the sound to be OK.

And probably a huge amount better than using a weak feeble EF86 with
Ia at 1mA or less.

I say all this because anyone asking the original poster's question
needs to begin to ask 100 MORE QUESTIONS, and get out to his workshop
where he should become at least +20dB MORE ACTIVE with soldering iron,
voltmeters, THD measuring gear, and oscilloscope!

Only then will all answers be revealed about what likely outcomes will
be, and whether any outcomes are better or worse than doing it some
other way, and whether the criteria for citing "better way" is valid
or not.

So after several busy days I'd say any moderately intelligent person
should fill an exercise book up with carefully recorded results which
he/she can apply later when building any amp.

The conventional approach to using an EL84 as a preamp, ( just for
this crazy example ) is to strap it in triode with low value Rg2 grid
stopper of say 220 ohms between anode and G2.

Then you can still have total anode load of 8k3, and get OLG = 15
approx, and THD at 50Vrms might be 3%, with mainly 2H, and much lower
relative levels of other H compared to the "wild" natural untamed
pentode gain. NCFB could still be used to reduce CLG to the wanted
4.2, and gain reduction = 1/ 3.57, and THD at 50Vrms = approx 1.2%,
and at 0.5Vrms output its going to be around 0.012%, and still mainly
all 2H, and I suggest maybe you get better sound. I would suggest that
THD will rise if the anode to screen R is increased. One result with
increased R in series with G2 is much lower Eg2 relative to Ea. Most
pentodes "like" Eg2 slightly lower than Ea, maybe 20% lower for normal
class A operation especially in SE power stages or pentode signal
driver/preamp stages. Pentode IG2 is mostly a higher % than IG2 for
beam tetrodes, and for signal pentodes, Ig2 can be 30% of Ia. I can't
remember EL84 Ig2 figures, but if Ia was 20mA, expect Ig2 at 5mAdc if
Eg2 = Ea, and if series R to G2 was 10k0, then the drop across the R
G2 = 50Vdc, and one would find the EL84 would be a bit sick in triode
if the R between a and G2 = 10k0, so please try to use 220 ohms, my
humble request, so you might get better music. One could use a choke
from screen to a regulated B+ just for the screen and cap couple the
screen to anode OR TO THE CATHODE, and this will keep Eg2 high and at
the same level as Ea because a choke has fairly low wire resistance.
The one can instantly compare the difference between triode signal op
and pentode op. Chokes for such things need to be above 50H for low
Idc, and they don't grow on trees, and are a PIA to wind. So that's
another reason to use 220 ohms between a and G2 for triode, and for
pentode, you'll need a resistance of about 3 times the value of the
anode RL dc, and if that's 25k, then R supplying G2 from the same B+
of say +350Vdc might be 82k.
The bypass cap from G2 to k would be say 10uF, and one is free to add
in some extra series R between top of bypass cap and G2. Then you can
also try using G2 bypassed to 0V instead of to k, so that whille you
have NCFB applied in series with the VG1-0V signal, there is also a
Vg2-k signal voltage present. Don't ask me to explain further what
exact effects you might find, or what's the best.

I like shunt negative FB much more than any form of current FB,
because the tube Ra is lowered as well as the THD. The down side is
that Rin is lower, but the R1 arm of shunt FB can be kept at say 47k
and no lower, and R2 arm is typically 270k, thus not adding to the ac
loading on anode, and to findoutabout more, read my website on
preamps.

I've never used much shunt FB in power amp OP stages with pentodes,
because I favour cathode FB windings on the OPT.

but there is
always a point of diminished returns for each tube.

Sure, because the larger Rg2 the lower screen voltage, lower Ia, and
less power.

The 'better' way of doing it is UL. You still get less power but at
least screen voltage isn't also reduced so you only 'sacrifice' that
which goes directly to reducing distortion. (Allegedly there's a
'sweet spot' ratio where 'maximum benefit' of both is achieved).

It sure would be nice to be able to do a few math problem when looking
at datasheets and know where to start with a tube.

The rest is usually contained in the 'typical' operating
characteristics or example circuits.

I will introduce a controversial subject with g2 later, but for now I
just want to know what information on the datasheet will tell me what I
want to know.

I imagine that, with enough perseverance, one could derive screen
current magnitude, the resulting Rg2 voltage, and the degree of
feedback but it's not 'straight on the graphs' because they don't
expect them to be operated that way.

Terry

Something that might be interesting to try would be a sort of 'poor
man's UL' with a tube like the 6CW5 that can use 250V B+ but is screen
limited to 200V. There a 'significant size' unbypassed Rg2 of 10k
would put screen right at the spec'd 200 V.

Ah, the poor man's UL! aka Free Lunch UL. It doesn't work too well
because screen input resistance is fairly low, and where you have RG2
high, then VG2 becomes high, but screen FB applied this way is a non-
linear way of applying FB. The conventional UL OPT supplies the screen
input power, ie, Vg2 x Ig2 input from a relatively low source
resistance.
Efforts have been made to make a voltage divider between anode and 0V,
and then use a direct coupled cathode follower ( or a damn mosfet ) as
a low impedance driver of the screen. This sort of UL operation is
that of the un-poor man, or an Un-free Lunch, aka a bodge, or fudge,
because it just does something in a complex way that should and can be
so much simpler, ie, taps on anode windings. But the divider+CF buffer
drive to G2 can provide an excellent low impedance drive where it is
wanted, ie, to the screen, so it can be made to work OK but its more
to go wrong, so there ain't one commercially built amp with that
included.

But where you have say a damn lousy PP OPT with UL taps at only 20%,
and a lousy pri to sec ratio and where you want the RLa-a to be
higher, for the same sec speaker load, then the use of a circlotron
for the two OP tubes can be used, with EG2 being a fixed Vdc. The OPT B
+ tap is to 0V! The two B+ supplies aof say 350Vdc each are between
anode and primary, and cathodes are connected to UL taps. Fixed bias
might be used, and I leave you all to work out the possibilities and
outcomes with such an arrangement. Its FAR BETTER than any poor man's
UL, or the lunch that poor man might afford. In fact to avoid the
horrible poor man's free lunch, you are going to have to spend a bit
more on a different PT for circlotron operation.

I never go to restaurants, and always cook at home, as I know I get
better nutrition that way, and sure, its more expensive than fast food
from McMuckkers, and for amplifiers, I apply the same idea because
what I end up with for music consumption seems to work better than if
I shopped for amps.

My 3 dracma's worth.

Patrick Turner.

[flipper]

On Sun, 20 May 2012 04:28:58 -0700 (PDT), Patrick Turner
wrote:

On May 20, 2:05*pm, flipper wrote:
On Wed, 16 May 2012 23:18:38 +0000, TerryG

wrote:

Hi all,

I am new to this forum but have known several members here for quite
some time. *I know not many of you will recognise me from this handle,
but you will I know.

Anyway *I have been enjoying SEP amplifiers this last year an 1/2.
Don't know why they have the reputation they do compared to SET's, but
that is another issue.

Because, all else being equal (which it seldom is), a SE pentode will
have higher output impedance (worse damping factor) and more
distortion than a SET. The saving graces are it takes less input to
drive and produces more output power.

I have designed and built several pentode single ended amplifiers at
this point. *And the one point that still alludes me is: *How from what
the datasheets say can you figure out what resistor to start with on g2?

I understand the DC voltage value and that is not difficult to figure
out, but I like pentodes best when there is an unbypass resistor on g2.

I think you answered your own question. The common purposes of a G2
resistor are to bias, which you note is easy enough to determine from
the existing data, and to kill parasitic oscillations, which are not
'static' characteristics and, so, not something for a static graph.

An unpypassed G2, of the nature you're describing, is not 'typical'
operation.

I have found this does several things, it moves the top bias lines down,
and the bottom bias lines up at the same time raising ( or lowering
depending on how you look at it) the cut off bias voltage. *Usually by
putting a unbypassed g2 resistor on g2 your current drops compared to
bypassing, and a little more wattage is possible for the same operating
point.

G2 current will be less because you are, in effect, applying negative
feedback to the screen with an unbypassed G2 resistor. Plate current
will also be less because you are, in effect, applying negative
feedback to the screen with an unbypassed G2 resistor.

I don't know how you arrive at the 'more wattage' conclusion.

As a side note, the 'opposite polarity' Rg2 derived screen signal was
used for the, so called, "self split" (Class A) PP amplifiers. It
drives the second grid so you don't need 'another tube'.

Negative feedback is also why the SET has lower impedance, lower
distortion, less power and also why it takes more signal to drive it.
Ia of a triode depends on *BOTH* G1 and plate voltages so if you stick
an impedance between plate and B+ the generated signal is negative
feedback to the plate. That is precisely what the 'fixed' screen
voltage in a pentode prevents, plate voltage feedback, hence it's
higher plate impedance, gain, and distortion. Putting an unbypassed
screen resistor on G2 means you're applying 'a little' of the feedback
a triode would see, plus lowering screen voltage.

*It seems most g2's have a certain impedance,

Clearly it has impedance or else screen voltage would be 0 with any
resistance between G2 and B+.

EL84s you can start
with 1.5K and go to above 10K with the resistor have more effect on the
signal traces. *This also lowers distortion it appears,

Negative feedback.

I've always wondered how much good is the "negative feedback" that is
obtained in this very simple manner.

Then I guess it's about time for you to "get away from the PC and go
solder up a circuit and measure it all."

On the other hand, the OP's question was where he could find G2 values
in the datasheets and no amount of bench work will answer that.

snip puffery

but there is
always a point of diminished returns for each tube.

Sure, because the larger Rg2 the lower screen voltage, lower Ia, and
less power.

The 'better' way of doing it is UL. You still get less power but at
least screen voltage isn't also reduced so you only 'sacrifice' that
which goes directly to reducing distortion. (Allegedly there's a
'sweet spot' ratio where 'maximum benefit' of both is achieved).

It sure would be nice to be able to do a few math problem when looking
at datasheets and know where to start with a tube.

The rest is usually contained in the 'typical' operating
characteristics or example circuits.

I will introduce a controversial subject with g2 later, but for now I
just want to know what information on the datasheet will tell me what I
want to know.

I imagine that, with enough perseverance, one could derive screen
current magnitude, the resulting Rg2 voltage, and the degree of
feedback but it's not 'straight on the graphs' because they don't
expect them to be operated that way.

Terry

Something that might be interesting to try would be a sort of 'poor
man's UL' with a tube like the 6CW5 that can use 250V B+ but is screen
limited to 200V. There a 'significant size' unbypassed Rg2 of 10k
would put screen right at the spec'd 200 V.

Ah, the poor man's UL! aka Free Lunch UL.

No one said a blooming thing about a 'free lunch' and a "poor man's
insert anything" always refers to an inferior 'cheap' substitute for
the otherwise preferred solution.

It doesn't work too well
because screen input resistance is fairly low, and where you have RG2
high, then VG2 becomes high, but screen FB applied this way is a non-
linear way of applying FB.

That's the second time you made an arbitrary declaration of
"non-linear," after 'always wondering' about it, and with no
explanation of what the heck that means. Non-linear how? Like Ia is
non-linear?

I don't have any data one way or the other but it was apparently
considered 'linear enough', whatever that criteria might have been,
for self split PP amps. And while that's not entirely equivalent it's
a heck of a lot more substantive than you declaring it 'sort of works
like' Rk and then never again speaking of screen feedback in the
previous rambling.

The conventional UL OPT supplies the screen
input power, ie, Vg2 x Ig2 input from a relatively low source
resistance.

If you are externally *driving* G2 then you need a low source
impedance but a large value Rg2 is not 'driving' G2, the voltage
produced is the *result* of G2 current. G2 does the 'driving'.

That's akin to bitching that Rl on a triode is 'too high an impedance'
to 'drive the plate'.

snip


Patrick Turner.

[Patrick Turner]

snip,


It doesn't work too well
because screen input resistance is fairly low, and where you have RG2
high, then VG2 becomes high, but screen FB applied this way is a non-
linear way of applying FB.

That's the second time you made an arbitrary declaration of
"non-linear," after 'always wondering' about it, and with no
explanation of what the heck that means. Non-linear how? Like Ia is
non-linear?

I often type stuff which baffles many people, sometimes for a second
time, just as you mentioned, so to be sure about what you are doing,
get away from PC and be the the Real Scientist for yourself, in your
workshop, and thus findoutabout all things confusing involving more
than 4 interactive variables.



I don't have any data one way or the other but it was apparently
considered 'linear enough', whatever that criteria might have been,
for self split PP amps. And while that's not entirely equivalent it's
a heck of a lot more substantive than you declaring it 'sort of works
like' Rk and then never again speaking of screen feedback in the
previous rambling.

Anyway, I'm a bit confused by just what you mean now. what was "it"
that was apparently considered "linear enough"? please try to extend
your descriptions and definitions so all the rest of the dummies
reading your post will have some idea what you are saying.

The conventional UL OPT supplies the screen
input power, ie, Vg2 x Ig2 input from a relatively low source
resistance.

If you are externally *driving* G2 then you need a low source
impedance but a large value Rg2 is not 'driving' G2, the voltage
produced is the *result* of G2 current. G2 does the 'driving'.

Unlike G1, G2 has low input impedance. And current flow into isn't
marvellously linear with an applied linear voltage. The larger the
added series R between a voltage source and G2 becomes, the more the
G2 acts independantly and its voltage change is due to Ig2 x series,
and that isn't a very linear application of NFB. I'm saying :- don't
use large series R between any voltage source and G2 because it leads
to worse general operation of the tube in a number of ways. The screen
tap on the OPT at say 40% of P turns would have source impedance of
pentode Ra x 0.4 squared, because there is an Ra transformation here,
so Rout from the g2 tap for 6550 with Ra at 32k = 5.12k. However, as
soon as one connects G2 to the G2 tap at 40% on OPT then whatever
voltage change is applied to the G2 works to reduce Ra, perhaps to
about 3k5, so the G2 tap output resistance becomes 3k5 x 0.4 x 0.4 =
1.k4 approx, and in fact the UL connection lowers the circuit
impedance from any point to 0V. But the screen does have to be driven,
ie, supplied power, both current and voltage to do things to Ia,
whereas G1 does not have to be driven because its input power is
negligible because its input impedance is many megohms, and Iin is
tiny. But this changes with F because of Miller and stray C.

A series R does not drive anything, because its a passive element, but
where there is a series R between G2 tap and G2 on OPT, the anode
drives the OPT, and the tap applies voltage change and hence current
change to G2 and the series R tends to reduce the effectiveness of
what the G2 tap is there for. The ultimate screen NFB occurs when the
pentode or beam tube is triode connected. One never ever sees a high
value R between G2 and a for a triode connected tube. But you should
try it anyway, just to see why nobody uses high value R between a
voltage source and G2, without bypassing G2 to cathode.

That's akin to bitching that Rl on a triode is 'too high an
impedance'
to 'drive the plate'.

I doubt you.

Patrick Turner.

[flipper]

On Wed, 23 May 2012 02:13:59 -0700 (PDT), Patrick Turner
wrote:

snip,


It doesn't work too well
because screen input resistance is fairly low, and where you have RG2
high, then VG2 becomes high, but screen FB applied this way is a non-
linear way of applying FB.

That's the second time you made an arbitrary declaration of
"non-linear," after 'always wondering' about it, and with no
explanation of what the heck that means. Non-linear how? Like Ia is
non-linear?

I often type stuff which baffles many people, sometimes for a second
time, just as you mentioned,

That's because you often babble gibberish.

so to be sure about what you are doing,
get away from PC and be the the Real Scientist for yourself, in your
workshop, and thus findoutabout all things confusing involving more
than 4 interactive variables.

You're the one who "always wondered" so get to it.

I don't have any data one way or the other but it was apparently
considered 'linear enough', whatever that criteria might have been,
for self split PP amps. And while that's not entirely equivalent it's
a heck of a lot more substantive than you declaring it 'sort of works
like' Rk and then never again speaking of screen feedback in the
previous rambling.

Anyway, I'm a bit confused by just what you mean now. what was "it"
that was apparently considered "linear enough"? please try to extend
your descriptions and definitions so all the rest of the dummies
reading your post will have some idea what you are saying.

The "it" is what's been the topic of this thread: the voltage signal
generated by an unbypassed screen resistor.

Look up self split push pull amp.

According to RH4, using a screen resistor was done to get 'less
distortion' than using a divider off the plate signal.

The conventional UL OPT supplies the screen
input power, ie, Vg2 x Ig2 input from a relatively low source
resistance.

If you are externally *driving* G2 then you need a low source
impedance but a large value Rg2 is not 'driving' G2, the voltage
produced is the *result* of G2 current. G2 does the 'driving'.

Unlike G1, G2 has low input impedance. And current flow into isn't
marvellously linear with an applied linear voltage.

You miss the whole point in that one is not 'applying' an external
voltage *to* the screen. Screen current results in a voltage.

The larger the
added series R between a voltage source and G2 becomes, the more the
G2 acts independantly and its voltage change is due to Ig2 x series,

That is 'the point' of it.

and that isn't a very linear application of NFB.

That claim is obviously pulled whole cloth from thin air because the
first blooming thing you said about this subject was you "always
wondered" how good it was.

I'm saying :- don't
use large series R between any voltage source and G2 because it leads
to worse general operation of the tube in a number of ways.

I know what you've been 'saying' and it doesn't make sense to "always
wonder" and then claim to know what you have "always wondered" about.

The screen
tap on the OPT at say 40% of P turns would have source impedance of
pentode Ra x 0.4 squared, because there is an Ra transformation here,
so Rout from the g2 tap for 6550 with Ra at 32k = 5.12k. However, as
soon as one connects G2 to the G2 tap at 40% on OPT then whatever
voltage change is applied to the G2 works to reduce Ra, perhaps to
about 3k5, so the G2 tap output resistance becomes 3k5 x 0.4 x 0.4 =
1.k4 approx, and in fact the UL connection lowers the circuit
impedance from any point to 0V. But the screen does have to be driven,
ie, supplied power, both current and voltage to do things to Ia,

You just said a series screen resistor won't do anything to Ia because
it doesn't 'drive' G2. So much for the claim of 'non-linearity'.

Stop babbling.

whereas G1 does not have to be driven because its input power is
negligible because its input impedance is many megohms, and Iin is
tiny. But this changes with F because of Miller and stray C.

Nobody made a comparison to G1, except your nonsense.

A series R does not drive anything, because its a passive element,

You're the only one stuck on this 'driving' nonsense. Everyone else
has known from the get go it doesn't 'drive' G2 and, as I said, G2
current is what causes the voltage change.

but
where there is a series R between G2 tap and G2 on OPT, the anode
drives the OPT, and the tap applies voltage change and hence current
change to G2 and the series R tends to reduce the effectiveness of
what the G2 tap is there for. The ultimate screen NFB occurs when the
pentode or beam tube is triode connected. One never ever sees a high
value R between G2 and a for a triode connected tube. But you should
try it anyway, just to see why nobody uses high value R between a
voltage source and G2, without bypassing G2 to cathode.

Your the one who "always wondered" so get to it and find out.

That's akin to bitching that Rl on a triode is 'too high an
impedance'
to 'drive the plate'.

I doubt you.

A change in Ia causes a voltage across the load resistor and that
voltage on the plate is the cause of 'inherent triode feedback'.

The 'large' load resistor is not, however, 'driving the plate'. The
voltage, and resulting feedback, is the result of Ia through the load.

Whether that 'works' for an unbypassed screen resistor would likely
depend on how well it tracks plate current and, as I said, I have no
data one way or the other but at least I don't pretend to know what
you "always wondered."

Patrick Turner.

[TerryG]

Originally Posted by Patrick Turner

I say all this because anyone asking the original poster's question
needs to begin to ask 100 MORE QUESTIONS, and get out to his workshop
where he should become at least +20dB MORE ACTIVE with soldering iron,
voltmeters, THD measuring gear, and oscilloscope!

Only then will all answers be revealed about what likely outcomes will
be, and whether any outcomes are better or worse than doing it some
other way, and whether the criteria for citing "better way" is valid
or not.

So after several busy days I'd say any moderately intelligent person
should fill an exercise book up with carefully recorded results which
he/she can apply later when building any amp.


My 3 dracma's worth.

Patrick Turner.

Well I guess it is time to fix my audio signal generator and provide actual circuit measurements here and as much more specific questions.

I would say at this point I have made at least a few hundred pentode circuits and performed tests on them. One thing I did do was what as suggested here, I put a 10K 12watt pot as a reostat on g2, with a 5K OPT load, with an EL84, and ran the pot up and down with another 1K 10 watt pot on the cathode.

What I found was:

1. As I increased the unbypassed resistance value on g2 this lessened the idle current of the circuit, and so had to compensate with Rk to a lower value to keep the idle current.

Obviously g2's unbypassed resistor effected the current, but the question was: Is this because of lower g2 voltage. So I added back voltage on g2 to bring it back up to the original value (without changing the resistor value), this did bring the current back up some, but not to its original value. The unbypassed g2 resistor was effecting the circuit in a static sense, to some degree.

2. As g2 was increased, Rk was reduced to maintain the same current, and voltage on g2 was increased but still within the tubes operational maximum wattage, output wattage on the dummy load resistor would increase. Because the sensitivity of the circuit and tube was effected by the unbypassed resistor more input voltage swing was required, but now the tube would obtain higher output wattages before clipping.

I will set up a circuit and do all these tests again, and measure distortion. My distortion meter only does THD, I do not have a spectrum analyzer to see what order/s are being increased or decreased by the unbypassed g2 resistor.

Also, tubes differ in the effects obtained by an unbypassed g2 resistor. Therefore I will perform measurements on a few different tubes.



Rather than start a new thread on this topic I will post here to this one as it is be relevant to this discussion here.

My findings so far are that the g2 resistor are both dynamic and static in nature, how and why is what I intend to figure out. I guess since this is "unintended" pentode operation, there are no formulas to calculate an unbypassed g2 resistors effect, from a datasheet.

Any suggestions on how to fairly set up my test circuit are welcome.

Terry

[TerryG]

This chart explains a few things. Not of g2 specifically, but of distortion behaviors of Pentodes. An unbypasses resistor on g2 could essentially be helping by lower reflected impedance to the plate, so a little is good, but too much isn't.

This could be why small value resistors (relatively) make things sound better and larger values do not.

Terry

[Patrick Turner]

On May 24, 11:19*am, TerryG wrote:
[image:http://sphotos.xx.fbcdn.net/hphotos-...5421323913...]

This chart explains a few things. *Not of g2 specifically, but of
distortion behaviors of Pentodes. *An unbypasses resistor on g2 could
essentially be helping by lower reflected impedance to the plate, so a
little is good, but too much isn't.

This could be why small value resistors (relatively) make things sound
better and larger values do not.

Terry

You have referred the group to a very interesting graph for 2H and 3H
production in a pentode or beam tetrode. But the graph does not
explain much about having various values of series R between G2 and B
+.

Notice how the 2H is high for low RLa values, falls to ZERO at RLa =
2k2, then rises for RLa 2k2.

How can this be?

Well, the exact mechanism is a little difficult to explain
mathematically or any other way, but when you plot the load line
across the tube curves, you will see Ea change in + and - directions
being equal for RLa = 2k2. What seems to happen is that as RLa is
increased to to 2k2, 2H reduces, and then above 2k2 the PHASE of the
2H reverses 180d, Just one of those lesser known things. In fact for
low Rla values, a triode's THD driving an SE pentode cancels that
produced in the pentode, but only up to RLa = 2k2. Above 2K2, the
triode's 2H adds to the 2H of the pentode. All rather messy, and
needing lots of NFB to fix.

There is more to all this THD business, and I suggest you read RDH4
where a comparison between 2A3 and 6F6 is made, and the 6F6 produces a
much larger number of H than just 2H and 3H.

But harder to see from the anode Ea + Ia Ra curves is the 3H which
occurs at all RLa, increasing with RLa. The 3H in fact usually looks
like flattening of + and - peaks in the sine wave, and is a form of
wave compression. But some tubes produce increased peaking of + and -
sine wave peaks, which simply means the 3H PHASE is 180d opposite to
the 3H flattening wave.

From the curves we see on the picture, we can say that if the RLa-a in
a PP circuit had a fairly LOW value, then 2H current in each tube is
high, and 3H current low, and that if the tubes work in class A then
the 2H can cancel out itself, and 3H won't, but it will remain low.
The result means that a pair of 6L6GC with RLa-a = 4k0 can produce
22Watts of class A PO with only 2% THD, almost all 3H, and that's the
best we might expect from a couple of beam tubes. Its about the same
as can be expected from triode connected tubes in class A.
But in SE mode, the beam tetrode / pentode THD is very much higher,
regardless of load. With triode connection, the NULL in 2H production
does not happen, and the phase the 2H remains constant, and although
2H is high with low RLa for triode, 2H becomes minimal and quite low
when RL becomes maximal, approaching a constant current source. The 3H
produced by the triode output tube, including triode connected
pentodes or beam tubes is relatively much lower.
And where an SE triode drives an SE output triode, the THD always
cancels, although there is a second order effect by way of
intermodulation so that some extra 3H is made in addition to the sum
of the 3H of each of the two triodes.

Fiddling with a series R to G2 probably won't improve anything, ie,
allow lower THD or increased power, or give a better damping factor. I
recall nothing in RDH4 or any other book I've read.

If there was to be any benefit in a large value Rg2, it would have
been well explained in text books used in old days.

Patrick Turner.


--
TerryG

[flipper]

On Thu, 24 May 2012 01:19:25 +0000, TerryG
wrote:


[image:
http://sphotos.xx.fbcdn.net/hphotos-... 009207_n.jpg]


This chart explains a few things. Not of g2 specifically, but of
distortion behaviors of Pentodes. An unbypasses resistor on g2 could
essentially be helping by lower reflected impedance to the plate, so a
little is good, but too much isn't.

An unbypassed screen resistor might lower plate impedance, and lord
knows what else, but it isn't going to do a blessed thing to OPT
impedance.


This could be why small value resistors (relatively) make things sound
better and larger values do not.

Terry

[flipper]

On Wed, 23 May 2012 21:29:06 +0000, TerryG
wrote:


Patrick Turner;956843 Wrote:

I say all this because anyone asking the original poster's question
needs to begin to ask 100 MORE QUESTIONS, and get out to his workshop
where he should become at least +20dB MORE ACTIVE with soldering iron,
voltmeters, THD measuring gear, and oscilloscope!

Only then will all answers be revealed about what likely outcomes will
be, and whether any outcomes are better or worse than doing it some
other way, and whether the criteria for citing "better way" is valid
or not.

So after several busy days I'd say any moderately intelligent person
should fill an exercise book up with carefully recorded results which
he/she can apply later when building any amp.


My 3 dracma's worth.

Patrick Turner.

Well I guess it is time to fix my audio signal generator and provide
actual circuit measurements here and as much more specific questions.

I would say at this point I have made at least a few hundred pentode
circuits and performed tests on them. One thing I did do was what as
suggested here, I put a 10K 12watt pot as a reostat on g2, with a 5K OPT
load, with an EL84, and ran the pot up and down with another 1K 10 watt
pot on the cathode.

What I found was:

1. As I increased the unbypassed resistance value on g2 this lessened
the idle current of the circuit, and so had to compensate with Rk to a
lower value to keep the idle current.

Obviously g2's unbypassed resistor effected the current, but the
question was: Is this because of lower g2 voltage.

The plate curves answer that question, yes. For any value of grid,
lower screen volts is lower Ia.

Btw, in that context the word is "affected."

So I added back
voltage on g2 to bring it back up to the original value (without
changing the resistor value), this did bring the current back up some,
but not to its original value. The unbypassed g2 resistor was effecting
the circuit in a static sense, to some degree.

That doesn't make sense to me. There is nothing in the plate curves
that specify 'at what resistance' screen volts goes through. Screen
volts are volts.

2. As g2 was increased, Rk was reduced to maintain the same current, and
voltage on g2 was increased but still within the tubes operational
maximum wattage, output wattage on the dummy load resistor would
increase. Because the sensitivity of the circuit and tube was effected
by the unbypassed resistor more input voltage swing was required, but
now the tube would obtain higher output wattages before clipping.

'Higher wattage' also doesn't make sense to me. Lets do a quickie
theoretical analysis. Power is I(rms)^2 times R, which can also be
represented for a sine as (Ipp/2.828)^2 times R. Current can go no
lower than 0, which pins down one end of the peak. Imax is then the
current available at 0 grid (since it will grid clamp above 0) but as
Ia increases so does screen current, which lowers screen voltage, and
at any grid value lower screen voltage means lower Ia. Max current,
with an unbypassed Rg2, must be lower than if screen remained static,
so Ipp is less, and, so, power must be lower.

That presumes everything is adjusted for maximum power like, for
example, idle bias at exactly half Ip, and I suspect what's happening
is the 'max power' load line changes when you change the bias
conditions so you're getting 'less than ideal' results.

I will set up a circuit and do all these tests again, and measure
distortion. My distortion meter only does THD, I do not have a spectrum
analyzer to see what order/s are being increased or decreased by the
unbypassed g2 resistor.

Also, tubes differ in the effects obtained by an unbypassed g2 resistor.

That would be because the screen mu factor is different.

Therefore I will perform measurements on a few different tubes.

[image:
http://sphotos.xx.fbcdn.net/hphotos-... 710908_n.jpg]

Rather than start a new thread on this topic I will post here to this
one as it is be relevant to this discussion here.

My findings so far are that the g2 resistor are both dynamic and static
in nature, how and why is what I intend to figure out. I guess since
this is "unintended" pentode operation, there are no formulas to
calculate an unbypassed g2 resistors effect, from a datasheet.

Any suggestions on how to fairly set up my test circuit are welcome.

Terry