[[email protected]]

I have tested 32 6AU6 tubes in the circuit described in my earlier post.
Wired as a triode with anode, suppressor grid and screen grid connected
together. In all cases Ia is ~3.8mA and Va is ~100V. The results a

average mu 36.92
mu std dev 3.4 (9.2%)
average ra 9.82K
ra std dev 1.49K (15.1%)
average gm 3.81
gm std dev 0.43 (11.2%)

I then tested the same 32 tubes wired with anode and suppressor grid to
ground and using the screen grid as anode. I have not worked out the
averages etc yet, but the following is clear from the data.

In this mode (compared to the other triode mode):

ra is higher (looks around 5%)
gm is lower (again around 5%)
mu is little changed

Cheers

Ian

[Patrick Turner]

wrote:

I have tested 32 6AU6 tubes in the circuit described in my earlier post.
Wired as a triode with anode, suppressor grid and screen grid connected
together. In all cases Ia is ~3.8mA and Va is ~100V. The results a

average mu 36.92
mu std dev 3.4 (9.2%)
average ra 9.82K
ra std dev 1.49K (15.1%)
average gm 3.81
gm std dev 0.43 (11.2%)

I then tested the same 32 tubes wired with anode and suppressor grid to
ground and using the screen grid as anode. I have not worked out the
averages etc yet, but the following is clear from the data.

In this mode (compared to the other triode mode):

ra is higher (looks around 5%)
gm is lower (again around 5%)
mu is little changed

Cheers

Ian

Good work there.

The tube obviously works well with screen as the anode.

What about noise? is it any quiter using the screen
as anode?

What about pentode mode?

Patrick Turner.

[[email protected]]

Patrick Turner wrote:



wrote:

I have tested 32 6AU6 tubes in the circuit described in my earlier post.
Wired as a triode with anode, suppressor grid and screen grid connected
together. In all cases Ia is ~3.8mA and Va is ~100V. The results a

average mu 36.92
mu std dev 3.4 (9.2%)
average ra 9.82K
ra std dev 1.49K (15.1%)
average gm 3.81
gm std dev 0.43 (11.2%)

I then tested the same 32 tubes wired with anode and suppressor grid to
ground and using the screen grid as anode. I have not worked out the
averages etc yet, but the following is clear from the data.

In this mode (compared to the other triode mode):

ra is higher (looks around 5%)
gm is lower (again around 5%)
mu is little changed

Cheers

Ian

Good work there.


Thanks.

The tube obviously works well with screen as the anode.

What about noise? is it any quiter using the screen
as anode?


Noise is next. The anode/suppressor grounded mode is intended primarily to
reduce hum picked up in the grid circuit because the heaters then have a
grounded pin on either side of them. Whether it is expected to be any
quieter in terms of other noise types I don't know.

To eliminate hum I plan to run the heaters on dc. The big caps I need to
smooth the heater supply have just arrived so I can now update the heater
supply. Obviously there are lots of sources of noise, not least of which is
wring to the grid. So the first test I will do is with the grid shorted and
with the ac 39K load connected. This means I can connect it to my ac
millivoltmeter which has an amplifier output so I can see the noise on a
scope. I know the stage gain for each tube under these conditions so I can
work out ein.

What about pentode mode?


I don't have any immediate plans to use the 6AU6 as a pentode but I will
complete the tests in pentode mode for completeness.

Cheers

ian

[Engineer[_2_]]

On Jul 6, 6:14 pm, wrote:
Patrick Turner wrote:

wrote:

I have tested 32 6AU6 tubes in the circuit described in my earlier post.
Wired as a triode with anode, suppressor grid and screen grid connected
together. In all cases Ia is ~3.8mA and Va is ~100V. The results a

average mu 36.92
mu std dev 3.4 (9.2%)
average ra 9.82K
ra std dev 1.49K (15.1%)
average gm 3.81
gm std dev 0.43 (11.2%)

I then tested the same 32 tubes wired with anode and suppressor grid to
ground and using the screen grid as anode. I have not worked out the
averages etc yet, but the following is clear from the data.

In this mode (compared to the other triode mode):

ra is higher (looks around 5%)
gm is lower (again around 5%)
mu is little changed

Cheers

Ian

Good work there.

Thanks.

The tube obviously works well with screen as the anode.

What about noise? is it any quiter using the screen
as anode?

Noise is next. The anode/suppressor grounded mode is intended primarily to
reduce hum picked up in the grid circuit because the heaters then have a
grounded pin on either side of them. Whether it is expected to be any
quieter in terms of other noise types I don't know.

To eliminate hum I plan to run the heaters on dc. The big caps I need to
smooth the heater supply have just arrived so I can now update the heater
supply. Obviously there are lots of sources of noise, not least of which is
wring to the grid. So the first test I will do is with the grid shorted and
with the ac 39K load connected. This means I can connect it to my ac
millivoltmeter which has an amplifier output so I can see the noise on a
scope. I know the stage gain for each tube under these conditions so I can
work out ein.

What about pentode mode?

I don't have any immediate plans to use the 6AU6 as a pentode but I will
complete the tests in pentode mode for completeness.

Cheers

ian- Hide quoted text -

- Show quoted text -

Nice work, Ian. Looking forward to the pentode figures! Did I see a
reference way back to the 6AU6 as a substitute of sorts for an EF86?
Would it work with 100K in the anode, 470K feeding the screen and 2.2
K (or was it 1.8K, I forget) cathode bias? (OK, I should try it
myself!)

BTW, I've always liked the DC heater idea but with 50/60 Hz
transformers and FW rectifier, the filter is massive for a 3 to 6 tube
preamp - what's your schematic? In words, I can't see the binaries
with my "cheap", i.e. free (I'm cheap), Google newsgroup software.
I'd like to run my pair of Heathkit WAP2 preamps with DC heaters = 1.8
amps as is, without rewiring the heaters in series... but that might
be the way go (19 VDC at 300 mA, each - easy.)
Cheers,
Roger

[Adam Stouffer]

Engineer wrote:


BTW, I've always liked the DC heater idea but with 50/60 Hz
transformers and FW rectifier, the filter is massive for a 3 to 6 tube
preamp - what's your schematic? In words, I can't see the binaries
with my "cheap", i.e. free (I'm cheap), Google newsgroup software.
I'd like to run my pair of Heathkit WAP2 preamps with DC heaters = 1.8
amps as is, without rewiring the heaters in series... but that might
be the way go (19 VDC at 300 mA, each - easy.)
Cheers,
Roger


How about using an LDO regulator like a LD1084V? Good for 5 amps and
only $2.64 at digikey.


Adam

[robert casey]

Noise is next. The anode/suppressor grounded mode is intended primarily to
reduce hum picked up in the grid circuit because the heaters then have a
grounded pin on either side of them. Whether it is expected to be any
quieter in terms of other noise types I don't know.


IIRC, the 6AU6 has a shield that surrounds the plate, and it I think is
connected to the G3 grid. So you may want that tied to ground to keep
crud off the plate.

[Patrick Turner]

wrote:

Patrick Turner wrote:



wrote:

I have tested 32 6AU6 tubes in the circuit described in my earlier post.
Wired as a triode with anode, suppressor grid and screen grid connected
together. In all cases Ia is ~3.8mA and Va is ~100V. The results a

average mu 36.92
mu std dev 3.4 (9.2%)
average ra 9.82K
ra std dev 1.49K (15.1%)
average gm 3.81
gm std dev 0.43 (11.2%)

I then tested the same 32 tubes wired with anode and suppressor grid to
ground and using the screen grid as anode. I have not worked out the
averages etc yet, but the following is clear from the data.

In this mode (compared to the other triode mode):

ra is higher (looks around 5%)
gm is lower (again around 5%)
mu is little changed

Cheers

Ian

Good work there.


Thanks.

The tube obviously works well with screen as the anode.

What about noise? is it any quiter using the screen
as anode?


Noise is next. The anode/suppressor grounded mode is intended primarily to
reduce hum picked up in the grid circuit because the heaters then have a
grounded pin on either side of them. Whether it is expected to be any
quieter in terms of other noise types I don't know.

To eliminate hum I plan to run the heaters on dc. The big caps I need to
smooth the heater supply have just arrived so I can now update the heater
supply. Obviously there are lots of sources of noise, not least of which is
wring to the grid. So the first test I will do is with the grid shorted and
with the ac 39K load connected. This means I can connect it to my ac
millivoltmeter which has an amplifier output so I can see the noise on a
scope. I know the stage gain for each tube under these conditions so I can
work out ein.

What about pentode mode?


I don't have any immediate plans to use the 6AU6 as a pentode but I will
complete the tests in pentode mode for completeness.

Cheers

ian

Your'e on your way to understand the 6AU6.

A simple way to get blamelessly smooth dc for the 6.6V heaters
is to have a roughly smoothed supply of say at least 10Vdc,
and then have a 7805 3 pin 5V regulator, but have a pair of diodes
in the ground pin circuit which gives you 5V + 1.3V = 6.3V which is very
well smoothed.


But otherwise the use of 10,000 uF 25V rated caps and a CRCRC type
filter
where a doubler is used to convert 6.3AC to about 15Vdc, the filtered
down to
about 6.3V works fine, although it pays to
have a 6.8 volt zener across the output to prevent excessively
high Vout if no tube is plugged in.

Patrick Turner.

[[email protected]]

Engineer wrote:

On Jul 6, 6:14 pm, wrote:
Patrick Turner wrote:

wrote:

I have tested 32 6AU6 tubes in the circuit described in my earlier
post. Wired as a triode with anode, suppressor grid and screen grid
connected together. In all cases Ia is ~3.8mA and Va is ~100V. The
results a

average mu 36.92
mu std dev 3.4 (9.2%)
average ra 9.82K
ra std dev 1.49K (15.1%)
average gm 3.81
gm std dev 0.43 (11.2%)

I then tested the same 32 tubes wired with anode and suppressor grid
to ground and using the screen grid as anode. I have not worked out
the averages etc yet, but the following is clear from the data.

In this mode (compared to the other triode mode):

ra is higher (looks around 5%)
gm is lower (again around 5%)
mu is little changed

Cheers

Ian

Good work there.

Thanks.

The tube obviously works well with screen as the anode.

What about noise? is it any quiter using the screen
as anode?

Noise is next. The anode/suppressor grounded mode is intended primarily
to reduce hum picked up in the grid circuit because the heaters then have
a grounded pin on either side of them. Whether it is expected to be any
quieter in terms of other noise types I don't know.

To eliminate hum I plan to run the heaters on dc. The big caps I need to
smooth the heater supply have just arrived so I can now update the heater
supply. Obviously there are lots of sources of noise, not least of which
is wring to the grid. So the first test I will do is with the grid
shorted and with the ac 39K load connected. This means I can connect it
to my ac millivoltmeter which has an amplifier output so I can see the
noise on a scope. I know the stage gain for each tube under these
conditions so I can work out ein.

What about pentode mode?

I don't have any immediate plans to use the 6AU6 as a pentode but I will
complete the tests in pentode mode for completeness.

Cheers

ian- Hide quoted text -

- Show quoted text -

Nice work, Ian. Looking forward to the pentode figures! Did I see a
reference way back to the 6AU6 as a substitute of sorts for an EF86?

Well the F86 is perhaps THE classic audio preamp pentode. However RDH has
quite a few examples of and references to the use of a 6AU6 in professional
audio. Plenty of good quality NOS examples are available at much lower
prices than the EF86.

Would it work with 100K in the anode, 470K feeding the screen and 2.2
K (or was it 1.8K, I forget) cathode bias? (OK, I should try it
myself!)


It would certainly work but whether it would give satisfactory performance
depends and many other factors. Yes you should try it yourself. There is
no substitute for bench work.

BTW, I've always liked the DC heater idea but with 50/60 Hz
transformers and FW rectifier, the filter is massive for a 3 to 6 tube
preamp - what's your schematic? In words, I can't see the binaries
with my "cheap", i.e. free (I'm cheap), Google newsgroup software.

At present nothing more than a 6.3V AC secondary, bridge rectifier and
10,000uF smoothing. This is almost certainly not adequate and I am only
running a single tube at present. I expect to end up with a higher
secondary voltage and a semiconductor regulator but only time and
experimentation will tell.

I'd like to run my pair of Heathkit WAP2 preamps with DC heaters = 1.8
amps as is, without rewiring the heaters in series... but that might
be the way go (19 VDC at 300 mA, each - easy.)

Sounds like a good way to go. I expect to have three tubes per channel. At
present I am making noise measurements in triode mode. My current problem
is residual hum Changing the HT smoothing does not alter the level of hum
in the output so I suspect it gets in from the heaters. Since this is just
a test bed I am thinking of simply using 4x.5V cells for the heater supply.

Cheers

Ian

[[email protected]]

Patrick Turner wrote:



wrote:

Patrick Turner wrote:



wrote:

I have tested 32 6AU6 tubes in the circuit described in my earlier
post. Wired as a triode with anode, suppressor grid and screen grid
connected together. In all cases Ia is ~3.8mA and Va is ~100V. The
results a

average mu 36.92
mu std dev 3.4 (9.2%)
average ra 9.82K
ra std dev 1.49K (15.1%)
average gm 3.81
gm std dev 0.43 (11.2%)

I then tested the same 32 tubes wired with anode and suppressor grid
to ground and using the screen grid as anode. I have not worked out
the averages etc yet, but the following is clear from the data.

In this mode (compared to the other triode mode):

ra is higher (looks around 5%)
gm is lower (again around 5%)
mu is little changed

Cheers

Ian

Good work there.


Thanks.

The tube obviously works well with screen as the anode.

What about noise? is it any quiter using the screen
as anode?


Noise is next. The anode/suppressor grounded mode is intended primarily
to reduce hum picked up in the grid circuit because the heaters then have
a grounded pin on either side of them. Whether it is expected to be any
quieter in terms of other noise types I don't know.

To eliminate hum I plan to run the heaters on dc. The big caps I need to
smooth the heater supply have just arrived so I can now update the heater
supply. Obviously there are lots of sources of noise, not least of which
is wring to the grid. So the first test I will do is with the grid
shorted and with the ac 39K load connected. This means I can connect it
to my ac millivoltmeter which has an amplifier output so I can see the
noise on a scope. I know the stage gain for each tube under these
conditions so I can work out ein.

What about pentode mode?


I don't have any immediate plans to use the 6AU6 as a pentode but I will
complete the tests in pentode mode for completeness.

Cheers

ian

Your'e on your way to understand the 6AU6.

A simple way to get blamelessly smooth dc for the 6.6V heaters
is to have a roughly smoothed supply of say at least 10Vdc,
and then have a 7805 3 pin 5V regulator, but have a pair of diodes
in the ground pin circuit which gives you 5V + 1.3V = 6.3V which is very
well smoothed.


Yes, I think I have come across that idea before. ISTR many moons ago (and I
am talking 30 years here) that adding diodes or resistors in series with
the ground terminal is not recommended by the manufacturers because of
potential instability. Maybe todays devices do not suffer from this problem
or is it advised to decouple the diodes too?


But otherwise the use of 10,000 uF 25V rated caps and a CRCRC type
filter
where a doubler is used to convert 6.3AC to about 15Vdc, the filtered
down to
about 6.3V works fine, although it pays to
have a 6.8 volt zener across the output to prevent excessively
high Vout if no tube is plugged in.


What about higher voltages and series heaters as others have mentioned. The
pros are easier currents, what are the cons?

Cheers

Ian

[Patrick Turner]

wrote:

Patrick Turner wrote:



wrote:

Patrick Turner wrote:



wrote:

I have tested 32 6AU6 tubes in the circuit described in my earlier
post. Wired as a triode with anode, suppressor grid and screen grid
connected together. In all cases Ia is ~3.8mA and Va is ~100V. The
results a

average mu 36.92
mu std dev 3.4 (9.2%)
average ra 9.82K
ra std dev 1.49K (15.1%)
average gm 3.81
gm std dev 0.43 (11.2%)

I then tested the same 32 tubes wired with anode and suppressor grid
to ground and using the screen grid as anode. I have not worked out
the averages etc yet, but the following is clear from the data.

In this mode (compared to the other triode mode):

ra is higher (looks around 5%)
gm is lower (again around 5%)
mu is little changed

Cheers

Ian

Good work there.


Thanks.

The tube obviously works well with screen as the anode.

What about noise? is it any quiter using the screen
as anode?


Noise is next. The anode/suppressor grounded mode is intended primarily
to reduce hum picked up in the grid circuit because the heaters then have
a grounded pin on either side of them. Whether it is expected to be any
quieter in terms of other noise types I don't know.

To eliminate hum I plan to run the heaters on dc. The big caps I need to
smooth the heater supply have just arrived so I can now update the heater
supply. Obviously there are lots of sources of noise, not least of which
is wring to the grid. So the first test I will do is with the grid
shorted and with the ac 39K load connected. This means I can connect it
to my ac millivoltmeter which has an amplifier output so I can see the
noise on a scope. I know the stage gain for each tube under these
conditions so I can work out ein.

What about pentode mode?


I don't have any immediate plans to use the 6AU6 as a pentode but I will
complete the tests in pentode mode for completeness.

Cheers

ian

Your'e on your way to understand the 6AU6.

A simple way to get blamelessly smooth dc for the 6.6V heaters
is to have a roughly smoothed supply of say at least 10Vdc,
and then have a 7805 3 pin 5V regulator, but have a pair of diodes
in the ground pin circuit which gives you 5V + 1.3V = 6.3V which is very
well smoothed.


Yes, I think I have come across that idea before. ISTR many moons ago (and I
am talking 30 years here) that adding diodes or resistors in series with
the ground terminal is not recommended by the manufacturers because of
potential instability. Maybe todays devices do not suffer from this problem
or is it advised to decouple the diodes too?

A small current flows through the chip to 0V via the centre pin of the
device,
and when I have used two diodes to raise the output voltage I have not
encountered any problems.
I've done it with 7812 with one diode to get 12.6V and its very
effective.




But otherwise the use of 10,000 uF 25V rated caps and a CRCRC type
filter
where a doubler is used to convert 6.3AC to about 15Vdc, the filtered
down to
about 6.3V works fine, although it pays to
have a 6.8 volt zener across the output to prevent excessively
high Vout if no tube is plugged in.


What about higher voltages and series heaters as others have mentioned. The
pros are easier currents, what are the cons?

Series heaters are fine and if you had 10 preamp tubes then you'd maybe
need
64Vdc at 300mA.

To stop any stray coupling at HF, a couple of 0.47uF to 0V
at points along the series will decouple, but if the input tubes to a
pre
are closest to 0V then they can have large caps to shunt them so the
tubes themselves form the
R part of the CRCRC set up.

But series heaters ra every inflexible to circuit/tube changes in future
so i never use a series set up.
I often use 12.6Vdc because there will always be a pair of tubes for the
same function in two channels.

And where I have a µ-follower or cathode follower in the tube line up I
may want
to have the whole heater supply biased at say +60V to ensure the heaters
of upper triodes in
µ-followers or CF cathodes are not more than 90V away from the heater
voltage.

In power amps with lots of output tubes I use two phases of 12.6Vac for
heaters
to avoid having the high currents of just one phase of 6.3V.

In my 10 tube preamp I have passive CLCRCRCRC B+ filtering and a two
transistor
discrete device "buffer" emitter follower regulator for the 1.8 amps at
12.6Vdc
for all the tubes involved, and the whole supply is biased at +56V at
the low side of the +12.6Vdc supply.

See my preamp PSU schematic at 1/4 way down the page at
http://www.turneraudio.com.au/preamp...ated-2006.html

Notice how I have the heaters for the phono stage decoupled with L2/3
and C17-20 to prevent
RF entering the phono stage or HF oscillations involving the heater
circuitry.

The religious use of star earthing and absense of loops in 0V rails is
essential
for low hum in preamps.

For MC amp stages I recommend a chassis and bottom cover made of steel
to shunt
stray magnetic fields causing mains hums.

All nearby transformers need to be potted and the PSU with trannies
within
should be in a steel case to reduce magneticly inducted hums.

Patrick Turner.


Cheers

Ian