Gentlemen,

I am looking for plate curves of the ECL86/PCL86/6GW8/14GW8
pentode section, but being triode strapped.

I am aware that the pentode section of this tube in pentode
mode is known to behave very much like a Pd-derated EL84/6BQ5.

But when comparing triode strapped EL84/6BQ5 plate curves
(which are available in spec sheets) with some selected points
I took of a few triode strapped ECL86/PCL86/6GW8/14GW8 pentode
section samples, there seems to be not so much similarity when
triode mode is used.

Since I don't have suitable PSUs (I would have to use 3
different non-contiguous PSUs to cover the Ea range alone)
to take complete plate curve graphs myself point by point,
I kindly would like to ask if someone has triode strapped
data and plate curves for the triode strapped
ECL86/PCL86/6GW8/14GW8 pentode section at hand or could even
trace/plot them for me w/ a curve tracer?

Tnx2u alot 4 any help!

Tom

--
MS-DOS is the worst text adventure game I have ever played:
Poor vocabulary, weak parser and boring storyline.

Tom Schlangen wrote:

Gentlemen,

I am looking for plate curves of the ECL86/PCL86/6GW8/14GW8
pentode section, but being triode strapped.

I am aware that the pentode section of this tube in pentode
mode is known to behave very much like a Pd-derated EL84/6BQ5.

But when comparing triode strapped EL84/6BQ5 plate curves
(which are available in spec sheets) with some selected points
I took of a few triode strapped ECL86/PCL86/6GW8/14GW8 pentode
section samples, there seems to be not so much similarity when
triode mode is used.

Since I don't have suitable PSUs (I would have to use 3
different non-contiguous PSUs to cover the Ea range alone)
to take complete plate curve graphs myself point by point,
I kindly would like to ask if someone has triode strapped
data and plate curves for the triode strapped
ECL86/PCL86/6GW8/14GW8 pentode section at hand or could even
trace/plot them for me w/ a curve tracer?

I have never seen this tube's triode's curves.

Why not just set one up in triode with Ea = 300v, and Ia = 30 mA.
Use a choke to deliver the DC, and you can cap couple any load value you
like
and use a CRO to monitor what your'e doing.

The gain with 10k and 5k can be measured with a small output voltage of
say
20v, to keep away from much thd, using a 1 kHz sine wave.

Since A = U x RL / ( RL + Ra ) for all tubes,
the two gain readings will allow you to get two equations,
and the values of U and Ra can be found.

The onset of grid current with various loads can be found,
and the anode voltages where this happens, so one can plot
the Ra line where Eg1 = 0v, and the rest of the triode line for Ra can
be approximately
then plotted.

The slope of the Ra at the operating point is only valid for that Ea/Ia
point,
but all triodes have a steepening line for Ra at higher Ea/Ia, and as
Ea/Ia drops the
Ra tends to have less slope, and curve down to the LHS to become
tangents off the
horizontal axis.
What you finish up drawing is an Ra line that
rises according to Child's Law, and the curve follows Y = X taken to the
power of 2/3, or is it
3/2, I have forgotten.

EL84 has Ra = 2,2k at 200v/20 mA, but its minimum Ra is less at higher
Ea/Ia,
and higher at lower Ea/Ia.
EL86 has Ra about 1/2 EL84, but only 1/2 the gain.

6BM8 and 6GW8 are around these figures.
The only reason I don't use the tri-pents for
drivers in my amps is that the small triode section
is like 1/2 a 12AX7, and pretty piddly; if only it was like a 1/2 6CG7,
I might use them more.

Patrick Turner.




Tnx2u alot 4 any help!


Tom

--
MS-DOS is the worst text adventure game I have ever played:
Poor vocabulary, weak parser and boring storyline.

Hi Patrick,

thank you for your advice, I will try to derive some
figures/curves along this method likely next weekend.

The only reason I don't use the tri-pents for
drivers in my amps is that the small triode section
is like 1/2 a 12AX7, and pretty piddly; if only it
was like a 1/2 6CG7, I might use them more.

Well, I found especially ECL82/6BM8/PCL82/16A8 to be
quite interesting to construct some sort of "universal
SE driver" for my experiments, using the pentode
section triode strapped. The result is a realiable
and very compact "module" that features an open loop
gain in the range of 100-150 times and has quite some
driving capabilities under severe (at least in the
range of my usual experiments) load conditions even with
a lowish B+ of, say, 250V.

For my quick breadboard level hacks of output stages,
such a driver module with no need to reinvent the wheel
over and over comes in very handy.

You say you don't like the triode sections of such
compound tubes (probably thinking about your production
amps first), but I found (at my level of knowledge
and interest in tube tech) the triode section of ECL82
bears some interesting features waiting to be explored
by me, notably when using extreme high value grid "leak"
resistors to build up the bias voltage w/o using
cathode bias. Looking at the according distortion
curves being published, this bias method seems to
be an interesting field for some experiments to me.

Seems this bias method is quite forgotten nowadays,
except maybe for some special tubes/circuits especially
intended for condenser mic usage ...

Tom

--
Live is too short to be taken seriously.
- Oscar Wilde

Tom Schlangen said:

by me, notably when using extreme high value grid "leak"
resistors to build up the bias voltage w/o using
cathode bias. Looking at the according distortion
curves being published, this bias method seems to
be an interesting field for some experiments to me.

Seems this bias method is quite forgotten nowadays,
except maybe for some special tubes/circuits especially
intended for condenser mic usage ...

This method is only useful with extremely small input signals.
Also, biasing atube this way may show some deviation from the "ideal"
working point.
With a cathode resistor, local NFB is applied for DC as well as to
linearize the tube's operation and satbilize the DC operation point.
The method you mention doesn't have this advantage, so with the tube
aging, the bias will start to decrease.
Also there's the danger of reverse grid current and the need of
extremely good coupling caps at the input.

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

Hi Sander,

so, probably no interesting bias method even for
home grown/use gear, but nevertheless interesting
to me to explore and I surely will play around with
it when there's no more important tube project at
hand and the TV full of sh*t and some time to spend.

For my ECL82 driver module, I use the usual cathode
bias method anyway.

Tom

--
When in doubt, use brute force.
- Ken Thompson

Tom Schlangen said:

Hi Sander,

so, probably no interesting bias method even for
home grown/use gear, but nevertheless interesting
to me to explore and I surely will play around with
it when there's no more important tube project at
hand and the TV full of sh*t and some time to spend.

There's nothing wrong with fooling around with tubes :-)
The TV is almost always full of ****.
Only things I watch are "Star trek Enterprise" and "The Office".
Oh, and I try to watch everything by Rick Mayall and Adrian Edmondson.
I know, I know, just let me.

The only possible advantage of the grid bias circuit back then would
have been that the grounded cathode couldn't introduce hum from the
filament, especially with very small signals e.g. microphone capsules.
These days with DC on the filaments, the problem no longer exists.
Some would even say: "These days with solid state devices at hand,
what the heck are you talking about?" :-)

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

Tom Schlangen wrote:

Hi Patrick,

thank you for your advice, I will try to derive some
figures/curves along this method likely next weekend.

The only reason I don't use the tri-pents for
drivers in my amps is that the small triode section
is like 1/2 a 12AX7, and pretty piddly; if only it
was like a 1/2 6CG7, I might use them more.

Well, I found especially ECL82/6BM8/PCL82/16A8 to be
quite interesting to construct some sort of "universal
SE driver" for my experiments, using the pentode
section triode strapped. The result is a realiable
and very compact "module" that features an open loop
gain in the range of 100-150 times and has quite some
driving capabilities under severe (at least in the
range of my usual experiments) load conditions even with
a lowish B+ of, say, 250V.

For my quick breadboard level hacks of output stages,
such a driver module with no need to reinvent the wheel
over and over comes in very handy.

Indeed.
My workbench "universal balanced voltage amp"
has a pair of 6CM5 as outputs,
a pair of E280F as drivers, lots of FB,
and bw is from 2Hz to 2 mHz, and max vo
is +/- 100vrms into 10k ohms, Ro = 600 ohms.

Great for testing OPTs.




You say you don't like the triode sections of such
compound tubes (probably thinking about your production
amps first), but I found (at my level of knowledge
and interest in tube tech) the triode section of ECL82
bears some interesting features waiting to be explored
by me, notably when using extreme high value grid "leak"
resistors to build up the bias voltage w/o using
cathode bias. Looking at the according distortion
curves being published, this bias method seems to
be an interesting field for some experiments to me.

I prefer medium U triodes for inputs and drivers.
And in the last few years I have found trioded EL84 to be
quite stunningly excellent to use as LTP halves,
since a pair of EL84 in triode is equal to
five pairs of 6CG7, but you only need one 2 pin sockets, not 10.





Seems this bias method is quite forgotten nowadays,
except maybe for some special tubes/circuits especially
intended for condenser mic usage ...

Having 5M hanging off g1 is OK to develop bias without an Rk,
or fixed bias in a radio, or where there is a preceding stage of low
impedance output. Otherwise you get the noise of the 5M resistor.

1k at the input of a mic amp is usually low enough to keep
noise lower than the grid imput noise of most tubes,
but 5M gives you 70 times more noise.
So the grid can be biased with 5M,
providing there isn't reverse grid current, which sends the grid +ve
instead of -ve, and the grid is cap coupled to a low impedance mic.

Patrick Turner

daio

( that's what Harry Belafonte sang )



Tom

--
Live is too short to be taken seriously.
- Oscar Wilde

Hi Patrick,

My workbench "universal balanced voltage amp"
has a pair of 6CM5 as outputs,

You surely mean 6CW5/EL86? I have a bunch of them and found
them to be interesting for lowish Ea and highish Ia situations
in EL84/6BQ5-alike circuits (using lower Ra or Raa of course).

But this tube has quite a low Eg2 rating of (AFAIR)
only 200 volts only, so triode strapped op might
exceed the specs.

According to your experience, up to which Eg2=Ea could they
be used reliably being triode-strapped?

Tom

--
Okay, maybe i am paranoid. But that doesn't mean
they are not out to get me. - unknown

Tom Schlangen wrote:

Hi Patrick,

My workbench "universal balanced voltage amp"
has a pair of 6CM5 as outputs,

You surely mean 6CW5/EL86?

Nope, I do mean 6CM5, and in beam tetrode mode.
I tested all the usual output tubes but the only one which
gave a bandwidth in a normal two stage R loaded amp was the 6CM5.
It has a 1.8k RL fitted to a heatsink.
The mighty frame grid tube, the E280F, a telefunken,
was the only driver pentode good enough for the bamdwidth.
All done without LR compensation in the anode circuits.

I have a bunch of them and found
them to be interesting for lowish Ea and highish Ia situations
in EL84/6BQ5-alike circuits (using lower Ra or Raa of course).

EL86 was designed to be used at lower B+, higher Ia,
to suit radios, and were used in the Phillips totem pole
amp with an 800 ohm speaker coil cap coupled to the tubes.
It has 1/2 the gain and 1/2 the triode Ra of the EL84,
which at 20 mA of Ia has Ra = 2.2k and U = 20.


But this tube has quite a low Eg2 rating of (AFAIR)
only 200 volts only, so triode strapped op might
exceed the specs.

Triode op allows a higher than normal Ea.
Its because the g2 moves at the same potential at the anode,
so the g2 current with triode averages a lot less than with pentode,
so the g2 dissipation is less.
807 had g2 rated for 300v, but were allowed officially to be operated in
triode of UL amps at
B+ = 450v.

But 200v is quite OK for triode op, and would be ideal
for many drivers, where an RL = 15k with 200v across it might be used
to a supply of +400v.




According to your experience, up to which Eg2=Ea could they
be used reliably being triode-strapped?

I'd say about 250v for EL86.

For EL84, probably 400v.

EveAnna Manley pushes EL84 in near triode op with 66% UL taps
at over 500v for B+.

No need to ever use more than 350v.
For drivers, the little power tubes, EL84, EL86, 6V6, etc, are not
required to have any more than 20 mA
idle current.

There is no need to set them up on the rated 12 watts dissipation,
better to set up for less than 1/2 that.
This assumes class A1 op of the output stage being driven.
So if Ia is 1/2 normal, Ig2 will also be quite low, because
the Ig2 = approx 10% of Ia at most values of Ia.


But EL86 would make a good little driver to transformer couple to a
class AB2 amp,
and avoid the normal direct coupled CF stage.
You'd still get a slight voltage sag when the power becomes class AB2,
but its doable, and NFB will push it along OK.
The 1.5k Ra of the EL86 means the output impedance is low enough
to push into some grid current. A typical 1/2 12AU7 CF has Ro = 600
ohms.

Patrick Turner.






Tom

--
Okay, maybe i am paranoid. But that doesn't mean
they are not out to get me. - unknown

Hi Yves,

6CM5 = EL36 AFAIK

Indeed, got some of them, too. Very few data available ;-)

Tom

--
this is my favourite sig, since there is no reference to
Kibo, Discordianism or The Church of the Subgenius in it.

"Patrick Turner" a écrit dans le message de news:
...

[ . . . ]
Nope, I do mean 6CM5, and in beam tetrode mode.

6CM5 = EL36 AFAIK

[ . . . ]

Yves wrote:

"Patrick Turner" a écrit dans le message de news:
...

[ . . . ]
Nope, I do mean 6CM5, and in beam tetrode mode.

6CM5 = EL36 AFAIK

Yes.

Patrick Turner.



[ . . . ]