Hi, vacuumlanders,
I'm planning to build an "old style" 8-10 watt PP amplifier just for
fun, since I find I have most of the parts lying around in my "junk"
box. Basically, it will be 2 X 6SN7's, 2 X 6V6GT's with a Hammond 468
OPT; about 300 VDC B+ is planned.
My general question is: what is the best way to use the 6SN7's?
(Assume they get a well filtered +250 VDC B+).
Here are some of my thoughts so far:
1. Use just one 6SN7: 1/2 as a voltage amplifier DC coupled to 1/2 as
a concertina phase splitter (very simple, low gain, limited NFB
possible, too few glowing tubes!)
2. Use 1/2 a 6SN7 as a voltage amplifier DC coupled to a complete 6SN7
as Long Tailed Pair (LTP) phase splitter (same problems, also "wastes"
1/2 a 6SN7.)
3. Same as 2, but add the other 1/2 6SN7 as an RC-coupled preamp stage
(more gain but extra LF phase shift possibly still limiting NFB.)
4. Novel? Use a complete 6SN7 as a cascode voltage amplifier stage,
RC-coupled (plate voltage too high for DC coupling, I think) to the
LTP phase splitter above (gets near pentode gain in the first stage;
no more phase shift than 3 (I think); but how noisy is the cascode?)
5. Use 1/2 a 6SN7 as an amplifier DC coupled to 1/2 6SN7 as a
concertina phase splitter, then use two parallel 1/2 6SN7's to amplify
each phase to drive the 6V6's (Balance? Also, the 6V6's hardly needs
the dedicated drivers, IMO.)
6. Use a pentode amplifier stage instead, DC coupled to the LTP (only
RF pentodes, like metal 6SK7's, to hand - and they don't glow!)
7. Other..?
8. Any good pointers to proven 6SN7 schematics?
For all the "cute configs" (cascode, LTP), I would fix the 6.3 VAC
heater supply at +50 volts (decoupled) to stay within the "max 90 VDC"
heater-cathode rating of the 6SN7, but likely could get away with not
doing this for the concertina phase splitter.)
All comments welcomed. Many thanks for all replies
Cheers,
Roger

Engineer wrote:

Hi, vacuumlanders,
I'm planning to build an "old style" 8-10 watt PP amplifier just for
fun, since I find I have most of the parts lying around in my "junk"
box. Basically, it will be 2 X 6SN7's, 2 X 6V6GT's with a Hammond 468
OPT; about 300 VDC B+ is planned.
My general question is: what is the best way to use the 6SN7's?
(Assume they get a well filtered +250 VDC B+).
Here are some of my thoughts so far:
1. Use just one 6SN7: 1/2 as a voltage amplifier DC coupled to 1/2 as
a concertina phase splitter (very simple, low gain, limited NFB
possible, too few glowing tubes!)
2. Use 1/2 a 6SN7 as a voltage amplifier DC coupled to a complete 6SN7
as Long Tailed Pair (LTP) phase splitter (same problems, also "wastes"
1/2 a 6SN7.)
3. Same as 2, but add the other 1/2 6SN7 as an RC-coupled preamp stage
(more gain but extra LF phase shift possibly still limiting NFB.)
4. Novel? Use a complete 6SN7 as a cascode voltage amplifier stage,
RC-coupled (plate voltage too high for DC coupling, I think) to the
LTP phase splitter above (gets near pentode gain in the first stage;
no more phase shift than 3 (I think); but how noisy is the cascode?)
5. Use 1/2 a 6SN7 as an amplifier DC coupled to 1/2 6SN7 as a
concertina phase splitter, then use two parallel 1/2 6SN7's to amplify
each phase to drive the 6V6's (Balance? Also, the 6V6's hardly needs
the dedicated drivers, IMO.)
6. Use a pentode amplifier stage instead, DC coupled to the LTP (only
RF pentodes, like metal 6SK7's, to hand - and they don't glow!)

6SK7 is remote cut-off & really not a good tube for audio. It's close
relative, the 6SJ7 could be used with good results.


7. Other..?
8. Any good pointers to proven 6SN7 schematics?
For all the "cute configs" (cascode, LTP), I would fix the 6.3 VAC
heater supply at +50 volts (decoupled) to stay within the "max 90 VDC"
heater-cathode rating of the 6SN7, but likely could get away with not
doing this for the concertina phase splitter.)

If you intend to run the 6V6GT's cathode biased their cathode return point
is a good place to tie the heater supply & is a common way of doing that.
That simplifies things & gives good results.

All comments welcomed. Many thanks for all replies
Cheers,
Roger

A 2-stage Diff Amp using the 6SN7GT's will give you all the gain you need
to drive the 6V6GT's.
For the LTP's each can be done cheaply with a 2N2222, a common red LED &
two resistors.

Good Luck, John Stewart

On Thu, 16 Jun 2005 20:07:57 -0400, John Stewart
wrote:

A 2-stage Diff Amp using the 6SN7GT's will give you all the gain you need
to drive the 6V6GT's.
For the LTP's each can be done cheaply with a 2N2222, a common red LED &
two resistors.

Excellent plan. Perhaps consider *not* bypassing the 6V6's (common)
cathodes? (sorted for DC balance).

Good fortune,

Chris Hornbeck
Impeach. Nothing less.

"Engineer" wrote

...All comments welcomed.

What's the primary impedance of the Hammond, and what kind of input
sensitivity do you want?

I have been chewing over your problem for years and still can't
quite make my mind up. I use 4 6CH6, not so far from the 6V6 but
higher gain open loop, but with cathode feedback which results in
rather less.

It is still how it began, with direct coupled voltage amp /
concertina, no overall nfb (high open loop gain and CFB of output
stage results in low output impedance anyway), and an empty socket
for some uncertain purpose.

I have an extra and stupidly unforeseen problem in that grid current
begins with lower signal voltage than hoped for, in part a
consequence of the CFB, but also because of class A operation. Even
though I never use much power, it irks me that if I did, the limit
would be imposed by a malfunctioning concertina. If you know you
will stay well within AB1, you can get away with the concertina. In
which case, to add feedback, you should use an extra voltage stage
using one half (not two) of the other 6SN7. If you want stereo you
could share a valve between channels. Or use a mu-follower (not
cascode, IMO), which will increase the gain of that extra first
stage. Every book I have (er, 3) has a circuit like this. Van der
Veen's "Modern High End Valve Amplifiers" takes a detailed look at
adding extra first stage and overall nfb. Morgan Jones also
discusses the circuit in "Valve Amplifiers".

To get more gain for addition of the right amount of overall nfb,
and also to toughen up and make symmetrical the drive to the output
stage, I have nearly decided on a pair of direct-coupled
differential pairs. Obviously this complicates the power supply, but
I think it's worth it. I would use ECC88 for first and ECC82 for
second, dunno if you could get enough gain with two 6SN7 in your
circumstances.
Just ideas. What's "old style"? You might have a different idea of
fun to me. I get a kick out of optimisation. It makes me feel like
an engineer, sometimes.

Really I just want to fill the empty sockets with glowing valves...

Why 250V? That won't be enough for some options. What's the HT for
the output stage?

cheers, Ian

"Ian Iveson" wrote in message
news
"Engineer" wrote

...All comments welcomed.

What's the primary impedance of the Hammond, and what kind of input
sensitivity do you want?

The Hammond #468 OPT is 10Kohm center tapped with 4, 8 and 16 ohm LS
taps.
As for sensitivity at the input, I was thinking of about 100 mV for 10
watts. I have not calculated the required voltage gain yet. One use
is to feed it directly from a modern FM tuner (switched to mono) for
the office, just to impress the younger staff! (Likely not!)

I have been chewing over your problem for years and still can't
quite make my mind up. I use 4 6CH6, not so far from the 6V6 but
higher gain open loop, but with cathode feedback which results in
rather less.

It is still how it began, with direct coupled voltage amp /
concertina, no overall nfb (high open loop gain and CFB of output
stage results in low output impedance anyway), and an empty socket
for some uncertain purpose.

Don't know the 6CH6. I've only ever used the 6V6 (early SE
experiments, used to claim 4 watts but more likely only 2.5!), also
6F6, 6K6, EL84 and ECL82 (all PP) O/P tubes. Oh, and a pair of fat PX4
triodes (directly heated) eons ago.

I have an extra and stupidly unforeseen problem in that grid current
begins with lower signal voltage than hoped for, in part a
consequence of the CFB, but also because of class A operation. Even
though I never use much power, it irks me that if I did, the limit
would be imposed by a malfunctioning concertina. If you know you
will stay well within AB1, you can get away with the concertina.

I intend this project to be "well into AB", so no grid current. I
favour the LTP for the phase splitter as, years ago, I built 2
"Mullard 5-10's" that use an ECC83 this way.
See http://www.bonavolta.ch/hobby/en/audio/el84_7.htm

In which case, to add feedback, you should use an extra voltage
stage using one half (not two) of the other 6SN7. If you want stereo
you could share a valve between channels.

Ah, well... it starts with one for mono as I'm a bit short of parts
for two!

Or use a mu-follower (not cascode, IMO), which will increase the
gain of that extra first stage. Every book I have (er, 3) has a
circuit like this. Van der Veen's "Modern High End Valve Amplifiers"
takes a detailed look at adding extra first stage and overall nfb.
Morgan Jones also discusses the circuit in "Valve Amplifiers".

Do you have a schematic? Or the URL of one?

To get more gain for addition of the right amount of overall nfb,
and also to toughen up and make symmetrical the drive to the output
stage, I have nearly decided on a pair of direct-coupled
differential pairs. Obviously this complicates the power supply, but
I think it's worth it. I would use ECC88 for first and ECC82 for
second, dunno if you could get enough gain with two 6SN7 in your
circumstances.

The notion is to go all octal to use up the 6SN7's and 6V6's to hand
and match the Hammond #468 to hand. My 12AX7's and EF86's are being
saved for a stereo amp. project for which I actually have to buy the
OPT iron with after tax dollars. Painful!

Just ideas. What's "old style"?

All octal, 6SN7's and 6V6's! No miniature tubes, er...valves to you
(and me when I did the "5-10's"!).

You might have a different idea of fun to me. I get a kick out of
optimisation. It makes me feel like an engineer, sometimes.

Really I just want to fill the empty sockets with glowing valves...

Why 250V? That won't be enough for some options. What's the HT for
the output stage?

Anyway, likely about +300 V on the OPT, depending on what P/S I can
brew up. I figure I would filter that down to about +250 V for the
6SN7's. I know, I'd like a bit more for DC coupling.

cheers, Ian

Thanks for the ideas.
Cheers,
Roger

Engineer wrote:

Hi, vacuumlanders,
I'm planning to build an "old style" 8-10 watt PP amplifier just for
fun, since I find I have most of the parts lying around in my "junk"
box. Basically, it will be 2 X 6SN7's, 2 X 6V6GT's with a Hammond 468
OPT; about 300 VDC B+ is planned.
My general question is: what is the best way to use the 6SN7's?
(Assume they get a well filtered +250 VDC B+).
Here are some of my thoughts so far:
1. Use just one 6SN7: 1/2 as a voltage amplifier DC coupled to 1/2 as
a concertina phase splitter (very simple, low gain, limited NFB
possible, too few glowing tubes!)
2. Use 1/2 a 6SN7 as a voltage amplifier DC coupled to a complete 6SN7
as Long Tailed Pair (LTP) phase splitter (same problems, also "wastes"
1/2 a 6SN7.)
3. Same as 2, but add the other 1/2 6SN7 as an RC-coupled preamp stage
(more gain but extra LF phase shift possibly still limiting NFB.)
4. Novel? Use a complete 6SN7 as a cascode voltage amplifier stage,
RC-coupled (plate voltage too high for DC coupling, I think) to the
LTP phase splitter above (gets near pentode gain in the first stage;
no more phase shift than 3 (I think); but how noisy is the cascode?)
5. Use 1/2 a 6SN7 as an amplifier DC coupled to 1/2 6SN7 as a
concertina phase splitter, then use two parallel 1/2 6SN7's to amplify
each phase to drive the 6V6's (Balance? Also, the 6V6's hardly needs
the dedicated drivers, IMO.)
6. Use a pentode amplifier stage instead, DC coupled to the LTP (only
RF pentodes, like metal 6SK7's, to hand - and they don't glow!)
7. Other..?
8. Any good pointers to proven 6SN7 schematics?
For all the "cute configs" (cascode, LTP), I would fix the 6.3 VAC
heater supply at +50 volts (decoupled) to stay within the "max 90 VDC"
heater-cathode rating of the 6SN7, but likely could get away with not
doing this for the concertina phase splitter.)
All comments welcomed. Many thanks for all replies
Cheers,
Roger

You could try the basic circuit with adjusted supply voltages for 6V6
instead of KT88 at
http://www.turneraudio.com.au/htmlwe...0ulabinteg.htm

The CCS tail of the LTP gurantees balance of the LTP outputs.
My circuit uses the bias supply to provide a negative voltage of -87v
to the bottom of the 10k emitter resistor of the MJE340.
But a smaller R would work quite ok, and taken to lesser -ve supply.
I have used 6CG7 in that amp, but 6SN7 are exactly the same
electronically.

Patrick Turner.

On Fri, 17 Jun 2005 07:39:53 GMT, Patrick Turner
wrote:

http://www.turneraudio.com.au/htmlwe...0ulabinteg.htm

That's DOA, man. Kick some butt at your page server; we want
to see what you're talking about.

Thanks,

Chris Hornbeck

Chris Hornbeck said:


http://www.turneraudio.com.au/htmlwe...0ulabinteg.htm

That's DOA, man. Kick some butt at your page server; we want
to see what you're talking about.


??? Page works for me.

--

"Audio as a serious hobby is going down the tubes."
- Howard Ferstler, 25/4/2005

"Chris Hornbeck" wrote in message
...
On Fri, 17 Jun 2005 07:39:53 GMT, Patrick Turner
wrote:

http://www.turneraudio.com.au/htmlwe...0ulabinteg.htm

That's DOA, man. Kick some butt at your page server; we want
to see what you're talking about.

Thanks,

Chris Hornbeck

I accessed it OK.
Cheers,
Roger

On Sat, 18 Jun 2005 17:29:35 +0200, Sander deWaal
wrote:

??? Page works for me.

Broken again for me this morning. Guess my ISP's DNS server has lost
it.

Thanks,

Chris Hornbeck

Here is an amp that I built from accumulated 'junk box' parts.
http://home.comcast.net/+AH4-suptjud/HomemadeHiFi.htm
+ACI-Engineer+ACI- +ADw-fakeaddress+AEA-nowhere.net+AD4- wrote in message
news:abmdndGtFp-OnC+AF8-fRVn-rw+AEA-rogers.com...
+AD4- Hi, vacuumlanders,
+AD4- I'm planning to build an +ACI-old style+ACI- 8-10 watt PP amplifier just for fun,
+AD4- since I find I have most of the parts lying around in my +ACI-junk+ACI- box.
+AD4- Basically, it will be 2 X 6SN7's, 2 X 6V6GT's with a Hammond 468 OPT+ADs-
+AD4- about 300 VDC B is planned.
+AD4- My general question is: what is the best way to use the 6SN7's? (Assume
+AD4- they get a well filtered +250- VDC B).
+AD4- Here are some of my thoughts so far:
+AD4- 1. Use just one 6SN7: 1/2 as a voltage amplifier DC coupled to 1/2 as a
+AD4- concertina phase splitter (very simple, low gain, limited NFB possible,
+AD4- too few glowing tubes+ACE-)
+AD4- 2. Use 1/2 a 6SN7 as a voltage amplifier DC coupled to a complete 6SN7 as
+AD4- Long Tailed Pair (LTP) phase splitter (same problems, also +ACI-wastes+ACI- 1/2 a
+AD4- 6SN7.)
+AD4- 3. Same as 2, but add the other 1/2 6SN7 as an RC-coupled preamp stage
+AD4- (more gain but extra LF phase shift possibly still limiting NFB.)
+AD4- 4. Novel? Use a complete 6SN7 as a cascode voltage amplifier stage,
+AD4- RC-coupled (plate voltage too high for DC coupling, I think) to the LTP
+AD4- phase splitter above (gets near pentode gain in the first stage+ADs- no more
+AD4- phase shift than 3 (I think)+ADs- but how noisy is the cascode?)
+AD4- 5. Use 1/2 a 6SN7 as an amplifier DC coupled to 1/2 6SN7 as a concertina
+AD4- phase splitter, then use two parallel 1/2 6SN7's to amplify each phase to
+AD4- drive the 6V6's (Balance? Also, the 6V6's hardly needs the dedicated
+AD4- drivers, IMO.)
+AD4- 6. Use a pentode amplifier stage instead, DC coupled to the LTP (only RF
+AD4- pentodes, like metal 6SK7's, to hand - and they don't glow+ACE-)
+AD4- 7. Other..?
+AD4- 8. Any good pointers to proven 6SN7 schematics?
+AD4- For all the +ACI-cute configs+ACI- (cascode, LTP), I would fix the 6.3 VAC heater
+AD4- supply at volts (decoupled) to stay within the +ACI-max 90 VDC+ACI-
+AD4- heater-cathode rating of the 6SN7, but likely could get away with not
+AD4- doing this for the concertina phase splitter.)
+AD4- All comments welcomed. Many thanks for all replies
+AD4- Cheers,
+AD4- Roger
+AD4-
+AD4-

"TerryJ" suptjudatcomcastdotnet wrote in message
...
Here is an amp that I built from accumulated 'junk box' parts.
http://home.comcast.net/~suptjud/HomemadeHiFi.htm
"Engineer" wrote in message
...

(snip)
Thanks, Terry. Nice amplifier, some good ideas - all grist for the
mill.
Cheers,
Roger