[Ian Bell[_2_]]

I just came across this:

http://ax84.com/media/ax84_m276.gif

An PP output stage that is its own phase splitter. Does this work??

Cheers

Ian

[Patrick Turner]

On Nov 20, 7:53*am, Ian Bell wrote:
I just came across this:

http://ax84.com/media/ax84_m276.gif

An PP output stage that is its own phase splitter. Does this work??

Cheers

Ian

Yes, it works, but not as well as having a balanced drive to to both
OP tube grids.
The OPT is 2,250 ohms to 8 ohms presumably for headphones, as the
12AU7 is not able to put much power into such a low primary RL value.

There is a common Rk of only 440R, and when you work out all the Vac
and Iac you will find the bottom triode of the pair shown isn't
contributing the same power as the top driven triode.
Equal power sharing could be done using a CCS instead of the 440
common Rk.

Anode to anode load should be 20k, not the idiotic 2k2. The designer
doesn't know his stuff.

Patrick Turner.

[Patrick Turner]

On Nov 20, 9:05*am, flipper wrote:
On Fri, 19 Nov 2010 20:53:37 +0000, Ian Bell
wrote:

I just came across this:

http://ax84.com/media/ax84_m276.gif

An PP output stage that is its own phase splitter. Does this work??

Yes, Class A. Although that's a primitive one with relatively poor AC
balance just like a 'short tail' stage would have (since that's
exactly what it is). Screen tap is better in that regard. Like this
one..

http://geckoamps.com/magnatone109/

Screen current varies along with anode current so a screen resistor
generates a proportional signal for the opposing grid.

They all depend on Class A or else there's no signal to drive the
opposing tube. I.E. If the tube goes into cutoff there's no change in
current for either cathode or screen drive.

The cathode drive would be better if it were a CCS instead of just an
Rk.


Agreed, but even then there is a different Va for each triode when
working.

Consider you had +/- 50Vrms of Va for each triode at the OPT primary
ends, two phases.

Assume you might have +20Vrms applied to one grid. If there was a CCS
"tail" then you should have near equal Vg to Vk for both tubes, so the
Vk would be +10Vrms if one grid is at 0V.

The phases of the Va are different, and in top triode Va = -50V, so Va-
k = Va - Vk = -50 - ( +10 ) = -60Vrms.

For bottom triode Va-k = Va - vk = +50V - ( +10V) = +40Vrms.

When the two triodes are equally loaded there is a different amount of
power developed by each tube. The amount of power becomes closer to
equal if the tube gain is high, which can only be the case if RL is
high.

All long tailed pairs with CCS tail and equal individual RL or
transformer coupled loads have this analomy. So perfect balancing and
exclusion of 2H generation is impossible, even though Va at each anode
is always equal if RL for each tube is equal.
With CCS tail, the anode currents are always equal, even if you have
different tube types for each side of the LTP. One could have a 12AT7
on one side, 12AU7 on the other, and all that would change would be
the 2H and the Va-k across each would not be equal, nor is Vk exactly
1/2 Vin at the live grid and nor is Vg-k for each tube.

Like everything simple, when you analyse, you often see a dog's
breakfast.

Patrick Turner.


Cheers

Ian- Hide quoted text -

- Show quoted text -

[Patrick Turner]

On Nov 20, 11:40*am, flipper wrote:
On Fri, 19 Nov 2010 16:16:08 -0800 (PST), Patrick Turner





wrote:
On Nov 20, 7:53 am, Ian Bell wrote:
I just came across this:

http://ax84.com/media/ax84_m276.gif

An PP output stage that is its own phase splitter. Does this work??

Cheers

Ian

Yes, it works, but not as well as having a balanced drive to to both
OP tube grids.
The OPT is 2,250 ohms to 8 ohms presumably for headphones, as the
12AU7 is not able to put much power into such a low primary RL value.

There is a common Rk of only 440R, and when you work out all the Vac
and Iac you will find the bottom triode of the pair shown isn't
contributing the same power as the top driven triode.
Equal power sharing could be done using a CCS instead of the 440
common Rk.

Anode to anode load should be 20k, not the idiotic 2k2. The designer
doesn't know his stuff.

The schematic I looked at said it was 22500 (which does match the 125
wiring diagram), not 2k2.



You are dead right, I'm wrong. Too many coffees at breakfast. The
schematic at http://ax84.com/media/ax84_m276.gif
definately shows 22500 : 8, using a Hammond 125A. I didn't think they
went that high with primary load resistance.
http://www.hammondmfg.com/125.htm

So what I said about choosing a load too low was wrong. But the rest I
said about the imperfections of a single drive was right. If the
choice of phase of the 8 ohm winding is done right, the bottom triode
with grounded grid could be driven with the secondary output signal
with R divider from the available higher sec voltage. To get 22.5k:8
ohms, taps 2-4 are used, a small fraction of the sec winding. the sec
voltage 1-6 is much higher.

Thus the input signal to the existing live grid could be halved, so
the drive to the output stage could be balanced from a signal from
itself, and methinks the 2H would then be much less than the existing
would have.

However, the Hammond 125A has only 14.9H for Lp, and if Ra-a = say 25k
at a low Ia idle, then the -3dB point is where RL in parallel with Ra-
a has same reactance as LP. So with RAA = 12k approx, -3dB is at
12,000 / ( 6.28 x 14.9 ) = 128Hz.

This is way too high, and so the 12AU7 is not the right tube, and
there should be a pair of EL86/EL84/6AR5 etc connected in triode to
get Ra-a down to 1/7 of the 12AU7. A couple of paralleled 12BH7 with
7mA in each triode, 28mA total might be OK. or at least more better.

Patrick Turner.

[Patrick Turner]

On Nov 20, 9:24*pm, flipper wrote:
On Fri, 19 Nov 2010 23:02:02 -0800 (PST), Patrick Turner





wrote:
On Nov 20, 11:40*am, flipper wrote:
On Fri, 19 Nov 2010 16:16:08 -0800 (PST), Patrick Turner

wrote:
On Nov 20, 7:53 am, Ian Bell wrote:
I just came across this:

http://ax84.com/media/ax84_m276.gif

An PP output stage that is its own phase splitter. Does this work??

Cheers

Ian

Yes, it works, but not as well as having a balanced drive to to both
OP tube grids.
The OPT is 2,250 ohms to 8 ohms presumably for headphones, as the
12AU7 is not able to put much power into such a low primary RL value.

There is a common Rk of only 440R, and when you work out all the Vac
and Iac you will find the bottom triode of the pair shown isn't
contributing the same power as the top driven triode.
Equal power sharing could be done using a CCS instead of the 440
common Rk.

Anode to anode load should be 20k, not the idiotic 2k2. The designer
doesn't know his stuff.

The schematic I looked at said it was 22500 (which does match the 125
wiring diagram), not 2k2.

You are dead right, I'm wrong. Too many coffees at breakfast. The
schematic at *http://ax84.com/media/ax84_m276.gif
definately shows 22500 : 8, using a Hammond 125A. I didn't think they
went that high with primary load resistance.
http://www.hammondmfg.com/125.htm

No worries. Happens to everyone from time to time.

So what I said about choosing a load too low was wrong. But the rest I
said about the imperfections of a single drive was right. If the
choice of phase of the 8 ohm winding is done right, the bottom triode
with grounded grid could be driven with the secondary output signal
with R divider from the available higher sec voltage. To get 22.5k:8
ohms, taps 2-4 are used, a small fraction of the sec winding. the sec
voltage 1-6 is much higher.

Thus the input signal to the existing live grid could be halved, so
the drive to the output stage could be balanced from a signal from
itself, and methinks the 2H would then be much less than the existing
would have.

Well, that's probably what I would do if self splitting two triodes
but the cathode split saves the bypass and no added components. I just
don't know what his 'logic' was and maybe he thinks that 'adds color'
or something.

He sure as heck wasn't shooting for 'power'.

Of course, he may have simply been unawares copying something, like
the equally dismal 'Compact HI-FI Power Amplifier' (a 1961 DIY
article) Triode Electronics links to from their 'universal OPT' page.

http://www.triodeel.com/compact.html

Claims "good inherent balance" with a 130 Ohm Rk.

Their 'Ultra Compact Club' CCSs the thing.

http://www.diyparadise.com/ultracompact.html

Which, in my opinion, is arm waving since, in exchange for the sand
pebbles, you could use one more blooming tube element for a proper
phase splitter and double the Po along with improving the balance and
distortion.

My 1964 RCA price sheet says a 6C4 was $1.85 with a 6CG7 (duplicating
gain) being only 33% more at $2.45. Plus the additional Rs and Cs, of
course.

But in 1964 wages were rather low, and $2.54 was a considerable sum of
money if you were an apprentice.



Or, if you're dead set on throwing sand in the works use THAT for a
split load phase splitter.

I suspect the author was aiming as much at 'simplicity' (less DIY
'intimidation' factor) as he was cost but he'd have been just as
simple and cheaper pairing a $3.20 6GW8 (6BQ5 plus 1/2 12AX7 in the
same envelope) with a $2.40 6BQ5, plus relieving the cost of one
socket, in the same 'simple self split' topology. And gain would be
higher so you could even wrap a few, but not enough to complicate
things, dBs around it.

A pair of 6GW8s would get you a split load phase splitter and is still
a hair cheaper than two 6BQ5s plus 6C4. That plus saving a socket
would go some of the way to (compared to the original) paying for the
additional Rs and Cs to make a proper amp and if you cleverly inject a
little PFB from the phase splitter into the gain stage you could wrap
20dB around it. Just comes 'natural' if you terminate the split load
into the first stage feedback Rk.

6BM8 with their beam tetrodes strapped as triodes would make an
excellent 4 watt amp, and suitable for phones if there was a
resistance divider, or an extra high OPT ratio. 22.5k : 8 could be
used OK with PP 6BM8 triodes for direct connection to phones because
if you have 100vrms across the primary, there is 1.88 V across the
sec, more than anyone needs to drive phones.
But the 125A has other taps maybe. The very high load on the 6BM8
triodes would reduce their THD to the minimum value, maybe 0.1% at a
watt.

That doesn't hold with today's prices but the difference is not enough
to warrant the performance hit.

Speaking of which, it can be amusing listening to folks talk about
'cheap (and, so, obviously 'junk') TV tubes', like the 6BQ7 people
will darn near give away these days. In 1964 that puppy was GOLD at
$4.95 ea. You could buy BOTH 6BQ5s, with change left over, for that
price and the vaunted 12AX7A was only $2.55 with the 'Hi-Fi' 6EU7
commanding a 15 cent premium at $2.70.

The 6BQ7 might be the only tube you could buy today, hop into a time
machine back to 1964, and double your money with. Terrific deal too
because they'd be paying you in 'silver certificates' worth real
money.

Anybody got a flux capacitor and DeLorean?

Where is Mr H.G. Wells when you need him?

Where are those plans for the time machine....

Probably a wife of an heir chucked them all out last week into a
dumster bin.

And no dumster divers found them.

However, the Hammond 125A has only 14.9H for Lp, and if Ra-a = say 25k
at a low Ia idle, then the -3dB point is where RL in parallel with Ra-
a has same reactance as LP. So with RAA = 12k approx, -3dB is at
12,000 / ( 6.28 x 14.9 ) = 128Hz.

This is way too high, and so the 12AU7 is not the right tube, and
there should be a pair of EL86/EL84/6AR5 etc connected in triode to
get Ra-a down to 1/7 of the 12AU7. A couple of paralleled 12BH7 with
7mA in each triode, 28mA total might be OK. or at least more better.

Don't forget that's a geetar amp. Doesn't need 20Hz. Although, I don't
know what he's thinking because his 'boost' circuit rolls around 289Hz
while the rest seems to be trying for better than that. Maybe he
thinks that sounds 'brighter' but then he 'tweaked' V2a Rk bypass for
'better bass', so I dunno.

He probably wouldn't be too surprised to hear it's -3dB at 128Hz since
the published spec is 150Hz-15kHz +-1dB, which usually translates to
-2dB on the ends.

The 125 series trannies are but toys.......

Patrick Turner.

[Ian Bell[_2_]]

Patrick Turner wrote:
On Nov 20, 7:53 am, Ian wrote:
I just came across this:

http://ax84.com/media/ax84_m276.gif

An PP output stage that is its own phase splitter. Does this work??

Cheers

Ian

Yes, it works, but not as well as having a balanced drive to to both
OP tube grids.
The OPT is 2,250 ohms to 8 ohms presumably for headphones, as the
12AU7 is not able to put much power into such a low primary RL value.


It is a guitar practice amplifier I believe.

There is a common Rk of only 440R, and when you work out all the Vac
and Iac you will find the bottom triode of the pair shown isn't
contributing the same power as the top driven triode.
Equal power sharing could be done using a CCS instead of the 440
common Rk.

Anode to anode load should be 20k, not the idiotic 2k2. The designer
doesn't know his stuff.


Perhaps because he is a guitarist??


Cheers

Ian
Patrick Turner.

[Ian Bell[_2_]]

Patrick Turner wrote:
On Nov 20, 9:05 am, wrote:
On Fri, 19 Nov 2010 20:53:37 +0000, Ian
wrote:

I just came across this:

http://ax84.com/media/ax84_m276.gif

An PP output stage that is its own phase splitter. Does this work??

Yes, Class A. Although that's a primitive one with relatively poor AC
balance just like a 'short tail' stage would have (since that's
exactly what it is). Screen tap is better in that regard. Like this
one..

http://geckoamps.com/magnatone109/

Screen current varies along with anode current so a screen resistor
generates a proportional signal for the opposing grid.

They all depend on Class A or else there's no signal to drive the
opposing tube. I.E. If the tube goes into cutoff there's no change in
current for either cathode or screen drive.

The cathode drive would be better if it were a CCS instead of just an
Rk.


Agreed, but even then there is a different Va for each triode when
working.

Consider you had +/- 50Vrms of Va for each triode at the OPT primary
ends, two phases.

Assume you might have +20Vrms applied to one grid. If there was a CCS
"tail" then you should have near equal Vg to Vk for both tubes, so the
Vk would be +10Vrms if one grid is at 0V.

The phases of the Va are different, and in top triode Va = -50V, so Va-
k = Va - Vk = -50 - ( +10 ) = -60Vrms.

For bottom triode Va-k = Va - vk = +50V - ( +10V) = +40Vrms.

When the two triodes are equally loaded there is a different amount of
power developed by each tube. The amount of power becomes closer to
equal if the tube gain is high, which can only be the case if RL is
high.

All long tailed pairs with CCS tail and equal individual RL or
transformer coupled loads have this analomy. So perfect balancing and
exclusion of 2H generation is impossible, even though Va at each anode
is always equal if RL for each tube is equal.
With CCS tail, the anode currents are always equal, even if you have
different tube types for each side of the LTP. One could have a 12AT7
on one side, 12AU7 on the other, and all that would change would be
the 2H and the Va-k across each would not be equal, nor is Vk exactly
1/2 Vin at the live grid and nor is Vg-k for each tube.

Like everything simple, when you analyse, you often see a dog's
breakfast.

Patrick Turner.


Yes, it was its simplicity that first attracted me to it.

Cheers

Ian

Cheers

Ian- Hide quoted text -

- Show quoted text -

[Ian Bell[_2_]]

Patrick Turner wrote:
On Nov 20, 11:40 am, wrote:
On Fri, 19 Nov 2010 16:16:08 -0800 (PST), Patrick Turner





wrote:
On Nov 20, 7:53 am, Ian wrote:
I just came across this:

http://ax84.com/media/ax84_m276.gif

An PP output stage that is its own phase splitter. Does this work??

Cheers

Ian

Yes, it works, but not as well as having a balanced drive to to both
OP tube grids.
The OPT is 2,250 ohms to 8 ohms presumably for headphones, as the
12AU7 is not able to put much power into such a low primary RL value.

There is a common Rk of only 440R, and when you work out all the Vac
and Iac you will find the bottom triode of the pair shown isn't
contributing the same power as the top driven triode.
Equal power sharing could be done using a CCS instead of the 440
common Rk.

Anode to anode load should be 20k, not the idiotic 2k2. The designer
doesn't know his stuff.

The schematic I looked at said it was 22500 (which does match the 125
wiring diagram), not 2k2.



You are dead right, I'm wrong. Too many coffees at breakfast. The
schematic at http://ax84.com/media/ax84_m276.gif
definately shows 22500 : 8, using a Hammond 125A. I didn't think they
went that high with primary load resistance.
http://www.hammondmfg.com/125.htm

So what I said about choosing a load too low was wrong. But the rest I
said about the imperfections of a single drive was right. If the
choice of phase of the 8 ohm winding is done right, the bottom triode
with grounded grid could be driven with the secondary output signal
with R divider from the available higher sec voltage. To get 22.5k:8
ohms, taps 2-4 are used, a small fraction of the sec winding. the sec
voltage 1-6 is much higher.

Thus the input signal to the existing live grid could be halved, so
the drive to the output stage could be balanced from a signal from
itself, and methinks the 2H would then be much less than the existing
would have.

However, the Hammond 125A has only 14.9H for Lp, and if Ra-a = say 25k
at a low Ia idle, then the -3dB point is where RL in parallel with Ra-
a has same reactance as LP. So with RAA = 12k approx, -3dB is at
12,000 / ( 6.28 x 14.9 ) = 128Hz.

This is way too high, and so the 12AU7 is not the right tube, and
there should be a pair of EL86/EL84/6AR5 etc connected in triode to
get Ra-a down to 1/7 of the 12AU7. A couple of paralleled 12BH7 with
7mA in each triode, 28mA total might be OK. or at least more better.

Patrick Turner.



The site I found the link on had created a modified version by replacing the
12AU7 with an ECC99. Good move??

Cheers

Ian

[Ian Bell[_2_]]

flipper wrote:
On Fri, 19 Nov 2010 23:02:02 -0800 (PST), Patrick Turner
wrote:

On Nov 20, 11:40 am, wrote:
On Fri, 19 Nov 2010 16:16:08 -0800 (PST), Patrick Turner





wrote:
On Nov 20, 7:53 am, Ian wrote:
I just came across this:

http://ax84.com/media/ax84_m276.gif

An PP output stage that is its own phase splitter. Does this work??

Cheers

Ian

Yes, it works, but not as well as having a balanced drive to to both
OP tube grids.
The OPT is 2,250 ohms to 8 ohms presumably for headphones, as the
12AU7 is not able to put much power into such a low primary RL value.

There is a common Rk of only 440R, and when you work out all the Vac
and Iac you will find the bottom triode of the pair shown isn't
contributing the same power as the top driven triode.
Equal power sharing could be done using a CCS instead of the 440
common Rk.

Anode to anode load should be 20k, not the idiotic 2k2. The designer
doesn't know his stuff.

The schematic I looked at said it was 22500 (which does match the 125
wiring diagram), not 2k2.



You are dead right, I'm wrong. Too many coffees at breakfast. The
schematic at http://ax84.com/media/ax84_m276.gif
definately shows 22500 : 8, using a Hammond 125A. I didn't think they
went that high with primary load resistance.
http://www.hammondmfg.com/125.htm

No worries. Happens to everyone from time to time.

So what I said about choosing a load too low was wrong. But the rest I
said about the imperfections of a single drive was right. If the
choice of phase of the 8 ohm winding is done right, the bottom triode
with grounded grid could be driven with the secondary output signal
with R divider from the available higher sec voltage. To get 22.5k:8
ohms, taps 2-4 are used, a small fraction of the sec winding. the sec
voltage 1-6 is much higher.

Thus the input signal to the existing live grid could be halved, so
the drive to the output stage could be balanced from a signal from
itself, and methinks the 2H would then be much less than the existing
would have.

Well, that's probably what I would do if self splitting two triodes
but the cathode split saves the bypass and no added components. I just
don't know what his 'logic' was and maybe he thinks that 'adds color'
or something.

He sure as heck wasn't shooting for 'power'.


I guess I should perhaps have mentioned that the original link that led me to
that circuit was for a guitar practice amp using an ECC99:

http://diyguitarfreak.wordpress.com/...niamp/#more-22

There's a video demo there too - the sound is quite good for a guitar amp.

Cheers

Ian

[Patrick Turner]

On Nov 21, 7:56*am, Ian Bell wrote:
Patrick Turner wrote:
On Nov 20, 7:53 am, Ian *wrote:
I just came across this:

http://ax84.com/media/ax84_m276.gif

An PP output stage that is its own phase splitter. Does this work??

Cheers

Ian

Yes, it works, but not as well as having a balanced drive to to both
OP tube grids.
The OPT is 2,250 ohms to 8 ohms presumably for headphones, as the
12AU7 is not able to put much power into such a low primary RL value.

It is a guitar practice amplifier I believe.

There is a common Rk of only 440R, and when you work out all the Vac
and Iac you will find the bottom triode of the pair shown isn't
contributing the same power as the top driven triode.
Equal power sharing could be done using a CCS instead of the 440
common Rk.

Anode to anode load should be 20k, not the idiotic 2k2. The designer
doesn't know his stuff.

Perhaps because he is a guitarist??

Cheers

Ian

The Guitarist's schematic does show 22500 : 8 OPT, with 125A
( Hammond )

I misread the schematic, and have corrected my mistake in a few posts
ago.

I have also suggested how easy if would be to derive a drive from the
OP secondary seconday back to the bottom triode, thus halving the
drive to the top triode and equalising the power contribution of each
triode, and a CCSink would not be needed.

Maybe he is Jimmy Hendrix's cousin, but he's better on the axe than
Jimmy, who knows.

Patrick T.



Patrick Turner.- Hide quoted text -

- Show quoted text -- Hide quoted text -

- Show quoted text -

[Patrick Turner]

On Nov 21, 8:00*am, Ian Bell wrote:
Patrick Turner wrote:
On Nov 20, 11:40 am, *wrote:
On Fri, 19 Nov 2010 16:16:08 -0800 (PST), Patrick Turner

*wrote:
On Nov 20, 7:53 am, Ian *wrote:
I just came across this:

http://ax84.com/media/ax84_m276.gif

An PP output stage that is its own phase splitter. Does this work??

Cheers

Ian

Yes, it works, but not as well as having a balanced drive to to both
OP tube grids.
The OPT is 2,250 ohms to 8 ohms presumably for headphones, as the
12AU7 is not able to put much power into such a low primary RL value.

There is a common Rk of only 440R, and when you work out all the Vac
and Iac you will find the bottom triode of the pair shown isn't
contributing the same power as the top driven triode.
Equal power sharing could be done using a CCS instead of the 440
common Rk.

Anode to anode load should be 20k, not the idiotic 2k2. The designer
doesn't know his stuff.

The schematic I looked at said it was 22500 (which does match the 125
wiring diagram), not 2k2.

You are dead right, I'm wrong. Too many coffees at breakfast. The
schematic at *http://ax84.com/media/ax84_m276.gif
definately shows 22500 : 8, using a Hammond 125A. I didn't think they
went that high with primary load resistance.
http://www.hammondmfg.com/125.htm

So what I said about choosing a load too low was wrong. But the rest I
said about the imperfections of a single drive was right. If the
choice of phase of the 8 ohm winding is done right, the bottom triode
with grounded grid could be driven with the secondary output signal
with R divider from the available higher sec voltage. To get 22.5k:8
ohms, taps 2-4 are used, a small fraction of the sec winding. the sec
voltage 1-6 is much higher.

Thus the input signal to the existing live grid could be halved, so
the drive to the output stage could be balanced from a signal from
itself, and methinks the 2H would then be much less than the existing
would have.

However, the Hammond 125A has only 14.9H for Lp, and if Ra-a = say 25k
at a low Ia idle, then the -3dB point is where RL in parallel with Ra-
a has same reactance as LP. So with RAA = 12k approx, -3dB is at
12,000 / ( 6.28 x 14.9 ) = 128Hz.

This is way too high, and so the 12AU7 is not the right tube, and
there should be a pair of EL86/EL84/6AR5 etc connected in triode to
get Ra-a down to 1/7 of the 12AU7. A couple of paralleled 12BH7 with
7mA in each triode, 28mA total might be OK. or at least more better.

Patrick Turner.

The site I found the link on had created a modified version by replacing the
12AU7 with an ECC99. Good move??

Could be, but what about a 6BL7?

Patrick Turner.

Cheers

Ian- Hide quoted text -

- Show quoted text -

[Patrick Turner]

On Nov 21, 10:24*am, flipper wrote:
On Sat, 20 Nov 2010 04:28:32 -0800 (PST), Patrick Turner





wrote:
On Nov 20, 9:24*pm, flipper wrote:
On Fri, 19 Nov 2010 23:02:02 -0800 (PST), Patrick Turner

wrote:
On Nov 20, 11:40*am, flipper wrote:
On Fri, 19 Nov 2010 16:16:08 -0800 (PST), Patrick Turner

wrote:
On Nov 20, 7:53 am, Ian Bell wrote:


snip,


But in 1964 wages were rather low, and $2.54 was a considerable sum of
money if you were an apprentice.

Oh, so now all of a sudden you're a bean counter, eh?

Weel,,, someone has to watch the beans, or they dissapear.

The price differential is 60 cents but a buck being worth a buck back
then is why I made mention of the additional Rs and Cs too. That would
be chump change today but perhaps not back then.

At age 13, my father funded me the cost of parts for a kir for an AM
radio, maybe 10 quid then because we had currency like the UK, pounds,
shillings and pennies. Even hapennies, or 1/2 of a penny.

10 Quid would have been a labourers weekly wage. Anyway, I got the
radio working after making the chassis which wasn't included in the
kit, 6AN7, 6N8, 6BM8, and 6X5 i recall, plus a set of coils, OPT, and
all R&C, after I listed them all out made a visit to Geroge Brown P/L
who sold electronic parts all made in Oz, including tubes and the
double tuning gang.

I didn't get time to make a proper radio cabinet and dial, but there
must have been a box for the speaker. It lasted well with a speaker in
a box for years in my dad's vetinary surgery. It kept the mob waiting
in turn with dogs and cats. The assembled crowd with animals sure
needed calming down with songs from Pat Boone and other suitable
crooners.

A new AM radio was a lot more than 10 quid. At that time, as a vet,
dad made much better dough than a labourer, so he reckoned he could
afford the cost.

I later built most of my house myself after learning to be a labourer,
and carpenter, and I worked up to being a foreman. Building my own pad
was much cheaper than working my guts out to pay a builder his wages,
profits and the tax and the interest in the loan needed to fund it
all. I did the work instead of watching telly.

So yeah, I guess count bean like my parents did, and they'd learned
from the Depression, rather like 10% of Americans are learning now. My
father spent the depression time at university learning to become vet,
and when he qualified they wouldn't let him join the army to fight
WW2. He was too valuable keeping farm animals alive. So he had a good
war at home, although he got a lot of white feathers in the mail.

He became a bit arrogant my old man. I guess he never really suffered
much as those who did the Depression as sacked workers, and then did
WW2 to lose an arm, and then find it damn difficult to survive, or buy
an AM radio, let alone a TV set.

People took ages to save for a telly. They would go down town and
watch the sets in the shop windows. Mesmerized they were with mainly
US made sit-coms and utter crap, some much worse than today's telly.


Or, if you're dead set on throwing sand in the works use THAT for
a
split load phase splitter.

I suspect the author was aiming as much at 'simplicity' (less DIY
'intimidation' factor) as he was cost but he'd have been just as
simple and cheaper pairing a $3.20 6GW8 (6BQ5 plus 1/2 12AX7 in the
same envelope) with a $2.40 6BQ5, plus relieving the cost of one
socket, in the same 'simple self split' topology. And gain would be
higher so you could even wrap a few, but not enough to complicate
things, dBs around it.

A pair of 6GW8s would get you a split load phase splitter and is still
a hair cheaper than two 6BQ5s plus 6C4. That plus saving a socket
would go some of the way to (compared to the original) paying for the
additional Rs and Cs to make a proper amp and if you cleverly inject a
little PFB from the phase splitter into the gain stage you could wrap
20dB around it. Just comes 'natural' if you terminate the split load
into the first stage feedback Rk.

6BM8 with their beam tetrodes strapped as triodes would make an
excellent 4 watt amp, and suitable for phones if there was a
resistance divider, or an extra high OPT ratio. 22.5k : 8 could be
used OK with PP 6BM8 triodes for direct connection to phones because
if you have 100vrms across the primary, there is 1.88 V across the
sec, more than anyone needs to drive phones.
But the 125A has other taps maybe. The very high load on the 6BM8
triodes would reduce their THD to the minimum value, maybe 0.1% at a
watt.

Perhaps but you're working in the wrong direction making the amp less
powerful.

Huh?

6BM8 would give more than 12AU7....

It wouldn't be a bad choice for a phase splitter version but it has
less gain (a 'feature' he's touting) and less plate power than the
6BQ5(6GW8)... and you couldn't pair it with the 6BQ5 for the 'even
cheaper' version.

He's pushing the envelope of 'hi-fi' there but he's shooting for rock
bottom cost.

Hmm, a single 6BQ5 and an old radio OPT would be cheaper than what
he's doing.

More power, nice 2H in over driven sound.



The 6BQ7 might be the only tube you could buy today, hop into a time
machine back to 1964, and double your money with. Terrific deal too
because they'd be paying you in 'silver certificates' worth real
money.

Anybody got a flux capacitor and DeLorean?

Where is Mr H.G. Wells when you need him?

Hobnobbing with the Eloy, last I head



The 125 series trannies are but toys.......

Yeah, I noticed. They're not really even good enough for a guitar amp.
More like 40's bargain basement table radio class.

Oh well old traditions die hard.

Patrick Turner.

[Patrick Turner]

On Nov 21, 11:03*pm, flipper wrote:
On Sun, 21 Nov 2010 02:10:38 -0800 (PST), Patrick Turner


snip,


The site I found the link on had created a modified version by replacing the
12AU7 with an ECC99. Good move??

Could be, but what about a 6BL7?

1.5A heaters and 12W plates for a 1W thereabouts amp?

A bit like attacking fireflies with hand grenades, isn't it?

But you don't need to run the Pda at 12W if RL is rather high.

Hand grenades? Blunderbuss perhaps? But anything done with tubes is
like travelling on a steam driven train rather than flying. So what, I
know folks who KILL to get on a steam train journey.

Patrick Turner

[Patrick Turner]

On Nov 21, 11:17*pm, flipper wrote:
On Sun, 21 Nov 2010 02:38:09 -0800 (PST), Patrick Turner





wrote:
On Nov 21, 10:24*am, flipper wrote:
On Sat, 20 Nov 2010 04:28:32 -0800 (PST), Patrick Turner

wrote:
On Nov 20, 9:24*pm, flipper wrote:
On Fri, 19 Nov 2010 23:02:02 -0800 (PST), Patrick Turner

wrote:
On Nov 20, 11:40*am, flipper wrote:
On Fri, 19 Nov 2010 16:16:08 -0800 (PST), Patrick Turner

wrote:
On Nov 20, 7:53 am, Ian Bell wrote:

snip,

But in 1964 wages were rather low, and $2.54 was a considerable sum of
money if you were an apprentice.

Oh, so now all of a sudden you're a bean counter, eh?

Weel,,, someone has to watch the beans, or they dissapear.

The price differential is 60 cents but a buck being worth a buck back
then is why I made mention of the additional Rs and Cs too. That would
be chump change today but perhaps not back then.

At age 13, my father funded me the cost of parts for a kir for an AM
radio, maybe 10 quid then because we had currency like the UK, pounds,
shillings and pennies. Even hapennies, or 1/2 of a penny.

10 Quid would have been a labourers weekly wage. Anyway, I got the
radio working after making the chassis which wasn't included in the
kit, 6AN7, 6N8, 6BM8, and 6X5 i recall, plus a set of coils, OPT, and
all R&C, after I listed them all out made a visit to Geroge Brown P/L
who sold electronic parts all made in Oz, including tubes and the
double tuning gang.

I didn't get time to make a proper radio cabinet and dial, but there
must have been a box for the speaker. It lasted well with a speaker in
a box for years in my dad's vetinary surgery. It kept the mob waiting
in turn with dogs and cats. The assembled crowd with animals sure
needed calming down with songs from Pat Boone and other suitable
crooners.

A new AM radio was a lot more than 10 quid. At that time, as a vet,
dad made much better dough than a labourer, so he reckoned he could
afford the cost.

I later built most of my house myself after learning to be a labourer,
and carpenter, and I worked up to being a foreman. Building my own pad
was much cheaper than working my guts out to pay a builder his wages,
profits and the tax and the interest in the loan needed to fund it
all. I did the work instead of watching telly.

So yeah, I guess count bean like my parents did, and they'd learned
from the Depression, rather like 10% of Americans are learning now. My
father spent the depression time at university learning to become vet,
and when he qualified they wouldn't let him join the army to fight
WW2. He was too valuable keeping farm animals alive. So he had a good
war at home, although he got a lot of white feathers in the mail.

He became a bit arrogant my old man. I guess he never really suffered
much as those who did the Depression as sacked workers, and then did
WW2 to lose an arm, and then find it damn difficult to survive, or buy
an AM radio, let alone a TV set.

People took ages to save for a telly. They would go down town and
watch the sets in the shop windows. Mesmerized they were with mainly
US made sit-coms and utter crap, some much worse than today's telly.

Someone scraping for hapennies wouldn't be building a 'hi-fi' amp.

The better comparison would be to see how much it is in relation to
the overall cost of the project vs benefit gained and I gave both
directions. One a little more and one a little less than the 'compact
amplifier'.





Or, if you're dead set on throwing sand in the works use THAT for
a
split load phase splitter.

I suspect the author was aiming as much at 'simplicity' (less DIY
'intimidation' factor) as he was cost but he'd have been just as
simple and cheaper pairing a $3.20 6GW8 (6BQ5 plus 1/2 12AX7 in the
same envelope) with a $2.40 6BQ5, plus relieving the cost of one
socket, in the same 'simple self split' topology. And gain would be
higher so you could even wrap a few, but not enough to complicate
things, dBs around it.

A pair of 6GW8s would get you a split load phase splitter and is still
a hair cheaper than two 6BQ5s plus 6C4. That plus saving a socket
would go some of the way to (compared to the original) paying for the
additional Rs and Cs to make a proper amp and if you cleverly inject a
little PFB from the phase splitter into the gain stage you could wrap
20dB around it. Just comes 'natural' if you terminate the split load
into the first stage feedback Rk.

6BM8 with their beam tetrodes strapped as triodes would make an
excellent 4 watt amp, and suitable for phones if there was a
resistance divider, or an extra high OPT ratio. 22.5k : 8 could be
used OK with PP 6BM8 triodes for direct connection to phones because
if you have 100vrms across the primary, there is 1.88 V across the
sec, more than anyone needs to drive phones.
But the 125A has other taps maybe. The very high load on the 6BM8
triodes would reduce their THD to the minimum value, maybe 0.1% at a
watt.

Perhaps but you're working in the wrong direction making the amp less
powerful.

Huh?

6BM8 would give more than 12AU7....

The 12AU7 isn't the topic here. The topic is the 'compact amplifier',
6BQ5s, 6GW8s, self split 6 Watts and split load 12 Watts.

It wouldn't be a bad choice for a phase splitter version but it has
less gain (a 'feature' he's touting) and less plate power than the
6BQ5(6GW8)... and you couldn't pair it with the 6BQ5 for the 'even
cheaper' version.

He's pushing the envelope of 'hi-fi' there but he's shooting for rock
bottom cost.

Hmm, a single 6BQ5 and an old radio OPT would be cheaper than what
he's doing.

More power, nice 2H in over driven sound.

An SE would not be more power.

Weel, 1 x 6BQ5 can do 5W which is more than 12AU7 in PP.

But yep, not as much as two 6BQ5 tweaked right up in PP.

The self split is sort of 'half way' between SE and Class AB because
it's stuck, like SE, in Class A. That, plus better distortion and
enough gain for some NFB is why I think the 60 cents is worth the
cost.





The 6BQ7 might be the only tube you could buy today, hop into a time
machine back to 1964, and double your money with. Terrific deal too
because they'd be paying you in 'silver certificates' worth real
money.

Anybody got a flux capacitor and DeLorean?

Where is Mr H.G. Wells when you need him?

Hobnobbing with the Eloy, last I head

The 125 series trannies are but toys.......

Yeah, I noticed. They're not really even good enough for a guitar amp.
More like 40's bargain basement table radio class.

Oh well old traditions die hard.

Yeah, well, people buy 'em and if you wrap some NFB around the thing
it's not so bad for guitar. That's essentially what the triode drive
does, 'local NFB' rather than global.

My old neighbour, an electronics enginneer, was young once and in
about 1964 the uni dept where he trained had a competition for the
best compact amp which fitted in a cube 4" x 4" x 4".

He said the guy won it used two 6BQ5, one 12AX7, PT and OPT, crammed
rather tightly.

Nowdays you could get 50W instead of 10W using digital PWM amps.
And it'd run a lot cooler.

Patrick Turner.

[Bret L]

Philco was fond of self phase splitting circuits in their outputs.
They didn't work very well and still don't.

[Patrick Turner]

On Nov 22, 1:40*pm, flipper wrote:
On Sun, 21 Nov 2010 16:59:41 -0800 (PST), Patrick Turner





wrote:
On Nov 21, 11:03 pm, flipper wrote:
On Sun, 21 Nov 2010 02:10:38 -0800 (PST), Patrick Turner

snip,

The site I found the link on had created a modified version by replacing the
12AU7 with an ECC99. Good move??

Could be, but what about a 6BL7?

1.5A heaters and 12W plates for a 1W thereabouts amp?

A bit like attacking fireflies with hand grenades, isn't it?

But you don't need to run the Pda at 12W if RL is rather high.

I understand that but there's no reason to have it either and you
don't have a choice on powering the 1.5A heater.

Hand grenades? Blunderbuss perhaps? But anything done with tubes is
like travelling on a steam driven train rather than flying. So what, I
know folks who KILL to get on a steam train journey.

Uh huh. But if you want a choo choo for going round the Christmas tree
buying a full size locomotive is like attacking fireflies with hand
grenades.

I know where some full sized locos are available. One is a Garatt,
with two articulated lots of driving wheels with a huge boiler in the
middle. One still runs somewhere.

http://www.gettyimages.com.au/detail/90774269/SSPL

It was one of Oz's most powerful locos, but now the trains used to
pull the iron ore to the ships waiting to fill up for China are
unreal, but lack the atmosphere and animal appeal of the Garratt which
was a visually and aurally stunning iron animal, especially on a cold
day. They ran across a paddock which we could see out the chool
window. But demand for such toys is at an all time low now, and the
local historical railway society and where the tracks are is all under
attack from real estate developers..

But usually wives prevent the purchase of full sized personal locos to
get around the Xmas tree and the appropriate farm sized house block
and railway station like home required. Worse than trying to own hi-fi
with the wrong kinda wife.

Steam fetishistas ultimately have to come to terms with their
passions.

Perhaps make do with wind up Hornby toy train sets which indeed can
get around the plug together tracks about 3 times per wind up, while
mouthing nice choo choo sounds and laying on the floor and making 1947
come right back.

But I recall in about 1956 my father taking me with him while school
holidays were on when he went to see the Heinz food company manager's
sick dog on a home visit. The attraction was the train set the old guy
had. Vast. Impressive. Competition winning. Quite fascinating - many
trains and locos, musta had complex controls and electronics. Set up
to look great at night with hundreds of tiny lights, even though
outside the huge underhouse garage it was a hot boiling day in summer.
Lots of Marklin german crafted models, and little model people walking
around little railway stations, and snow on the mountains. These
things were collectors items then and lord knows what they are worth
now. Anyway, it was a fine hour I had there that day. But dad never
did buy me a train set.

Nostalgia.

And tubes do good music.

Patrick Turner.

[Patrick Turner]

On Nov 22, 3:50*pm, Bret L wrote:
Philco was fond of self phase splitting circuits in their outputs.
They didn't work very well and still don't.

There are samples in RDH4. One problem is maintaining drive balance
because the drive to second PP input is derived from someplace where
the voltage is subject to change with load, ie, from an OPT.

And of course it is inevitable that self derived phase inversion leads
to doubling some distortions ( not all ) because it is effectively
positive feedback which doubles the open loop gain, as it is in the
case of a paraphase phase splitter where the anode output of one
triode stage is fed into a second triode's grid to create an inverted
phase.

Patrick Turner.

[Big Bad Bob]

Patrick Turner wrote:
When the two triodes are equally loaded there is a different amount of
power developed by each tube. The amount of power becomes closer to
equal if the tube gain is high, which can only be the case if RL is
high.

With CCS tail, the anode currents are always equal, even if you have
different tube types for each side of the LTP.

interesting point. So in effect you'll get twice the power output of a
single end class A, with performance based on the first triode's operating
curve.

some people want that 2nd harmonic THD so maybe this circuit would be useful
for them...

[Big Bad Bob]

flipper wrote:
On Fri, 19 Nov 2010 20:53:37 +0000, Ian Bell
wrote:

I just came across this:

http://ax84.com/media/ax84_m276.gif

An PP output stage that is its own phase splitter. Does this work??

Yes, Class A. Although that's a primitive one with relatively poor AC
balance just like a 'short tail' stage would have (since that's
exactly what it is). Screen tap is better in that regard. Like this
one..

http://geckoamps.com/magnatone109/

ewww.... can't say I much like what THAT is going to do to THD without some
kind of feedback. Pray you don't enter a reverse resistance region that
way. I can also see the possibility of a single component failure taking
multiple components with it.

Screen current varies along with anode current so a screen resistor
generates a proportional signal for the opposing grid.

er, not entirely proportional, but yeah it's "ok for gummint work"
proportional. Careful choice of components would be needed, as well as
tubes that do not degrade from their design specs.

They all depend on Class A or else there's no signal to drive the
opposing tube. I.E. If the tube goes into cutoff there's no change in
current for either cathode or screen drive.

I think the magnatone 109 circuit would work in AB mode. The curve for the
cathode/screen tap circuit is probably going to act MORE LIKE A TRIODE, even
when plate current gets really close to zero. So it's my guess that you
could cross into the AB range without degrading the output too much, at
least to the point where screen current 'pinches off'.

[Big Bad Bob]

flipper wrote:
I can't follow your logic there but if things were 'perfect' then +10
on the top grid would raise cathode to +10 because it's a cathode
follower. The effective grid on the lower is then 0-10V. It balances
and, if the tubes were linear, current simply shifts from one side to
the other.

in an ideal world the voltage between the screen grid and the cathode would
remain the same. In fact this is NOT the case, and that goes double for
"the other half" with the grounded grid. You need a delta voltage between
the cathode and the grid in order to change the plate current. Since the
cathode voltage is NOT a constant, the 'driven' tube's signal will always be
LOWER than the driving tube. It is necessary to create an increase in
cathode voltage on the 'upswing' (dropping the plate current of the other
tube from the INCREASE in total current across the shared cathode resistor),
and vice versa. So the current swing on the plate of one tube must be
higher than that of the other. High gains on both tubes will help keep this
from being a problem, but it will still be there, no matter what. And,
naturally, you'll be forced to operate class 'A'.

Having a constant current source on the cathode would probably address this,
but good luck finding something that would operate as a CCS in the tube
world at the kinds of current needed by power output tubes.

If you're going to add a tube to the circuit, a voltage following phase
splitter would be a better choice. And cheaper. Dual triodes make nice
pre-amp/splitters, especially 12AX7

[Big Bad Bob]

Bret L wrote:
Philco was fond of self phase splitting circuits in their outputs.
They didn't work very well and still don't.

if you insist on class A output and want to reduce the number of components,
it'll do the job. But yeah, it's inferior to 'hi-fi' circuits.

[Big Bad Bob]

Ian Bell wrote:
There is a common Rk of only 440R, and when you work out all the Vac
and Iac you will find the bottom triode of the pair shown isn't
contributing the same power as the top driven triode.
Equal power sharing could be done using a CCS instead of the 440
common Rk.

Anode to anode load should be 20k, not the idiotic 2k2. The designer
doesn't know his stuff.


Perhaps because he is a guitarist??

Booo. I'm a guitarist, and now you've gained my wrath. Be prepared to have
a song published about you. Muahahahaha! Just kidding, heh.

FYI Leo Fender's designs are much better, like the Fender Twin (as one
example). You can find a schematic he

http://www.freeinfosociety.com/elect...iew.php?id=961

V6 (12AT7) combines cathode coupling, biasing, and negative feedback for
phase inversion. The resistor values are difficult to read, but it looks
like there is a built-in compensation for any imbalance created by the
cathode coupling.

[Big Bad Bob]

On 11/22/10 18:31, flipper so witilly quipped:
I don't understand why you snipped out where I said the same thing.

probably didn't follow your logic, and thought you were saying something
different.

[Big Bad Bob]

On 11/22/10 20:14, flipper so witilly quipped:
I think the magnatone 109 circuit would work in AB mode. The curve for the
cathode/screen tap circuit is probably going to act MORE LIKE A TRIODE, even
when plate current gets really close to zero.

I don't know what you mean by that as the screen current curves mirror
the plate curves and it's in the cutoff region where screen as a fixed
percentage of plate current holds the best.

I would actually expect a number of interactions to happen in the
cathode-screen circuit. Screen current is not always a linear
proportion of plate current, especially when you're not using 'screen
taps' on the transformer. I am speculating that there is an AB bias
level that you could use in which plate current would swing sufficiently
low (with plate voltage nearly twice the B+) but with screen current
still 'linear enough' to drive the control grid of the other tube.
You'll obviously hit a point where this won't work, but it's my guess
that this will be FAR more linear than using a common cathode resistor
and grounded grid to drive the other tube. Keep in mind that a loose
definition of AB simply means "something between class A and B" and so
it would be a way of tweeking the bias levels to the point just before
screen current becomes too non-linear at the low end, maybe as low as
80% or 70% of class A. It might be a nice experiment, in any case, to
see how close to 50% you could go before THD was too much.

[John L Stewart]

Originally Posted by Ian Bell[_2_]

I just came across this:

http://ax84.com/media/ax84_m276.gif

An PP output stage that is its own phase splitter. Does this work??

Cheers

Ian

Hi Ian- The circuit as shown will work........sort of! But there will be lots of even order harmonics (2nd, 4th, 6th.....Etc). Just like the SE guys like.

An obvious & easy way out would be to replace that output stage cathode resistor with a low cost choke. I've found no mention of that fix anywhere in this thread. Perhaps I'm missing something.

If you like I can send you a couple of simulations I did a few years ago on that fix. Works very well. Not sure why it was never used in the 30s when several trick phase inverters showed up, all to save a few pennys. And shows up nowhere in RDH4.

Cheers, John