'PC Speaker' Tube Amp

'PC Speaker' Tube Amp

That's the latest project I finished, a little 'PC Speaker' Tube
amplifier using, per channel, a pair of 6AW8s PP into 20k OPTs. Does
between 3 and 4 Watts Po per channel.

Similar to my mini guitar amp it's powered off a 12VAC (40VA in this
case) wallwart with a reverse 20VA tranny inside generating about 240V
(loaded) for B+. Screens are at 200V.

PS filtering is 33uF --> OPT common --> 4.7k 33uF ---> screens --->
10k 33uF ---> Preamp. Hum is only detectable if you stick your ear on
the cones but I'm thinking about increasing the primary filter to 47uF
or 100 uF.

Idle is 10mA and, as has become my common practice with cathode bias,
I use a bypassed current mirror to keep a tight DC balance through the
OPT.

Preamp uses a common Rk PFB setup between the gain triode and
concertina phase splitter for a gain of about 140. The amp is then
closed with 18dB of.global NFB.

I used a 1/4 inch jack for the speaker connection to load switch
internal 10W sand block dummies, so the amp is never unloaded, and,
more to the point, the dummies provide a load when using headphones.

Might not be a monster amp but it's cute, compact, and fun as all get
out. The whole ball of wax, sans the wallwart of course, fits in a
5"x7" chassis.

For speakers I worked up a 300cu ported enclosure wrapped around the
MCM 1853 4 inch driver. They seem to have a response peak around 2kHz
but rather then smooth that out with passives in the cab I just use
the PC's equalizer and they sound amazingly good for 14 buck cheapies.


Darndest thing I ever saw, not that I've seen 'everything', but one of
the 6AW8s I got in a batch has a heater short. Not to cathode, or any
other pin, the heater itself. She pulls heavy current but won't light,
as if there's solid wire between the pins. Can't complain too much, I
guess, what with that batch being 50 cents each. Just wish he'd have
had more of them.

I'll post pics in alt.binaries.schematics.electronic if folks would
like to see it.

Re: 'PC Speaker' Tube Amp

Flipper said:

> That's the latest project I finished, a little 'PC
> Speaker' Tube
> amplifier using, per channel, a pair of 6AW8s PP into 20k
> OPTs. Does
> between 3 and 4 Watts Po per channel.
>
> Similar to my mini guitar amp it's powered off a 12VAC
> (40VA in this
> case) wallwart with a reverse 20VA tranny inside
> generating about 240V
> (loaded) for B+. Screens are at 200V.
>
> PS filtering is 33uF --> OPT common --> 4.7k 33uF --->
> screens --->
> 10k 33uF ---> Preamp. Hum is only detectable if you stick
> your ear on
> the cones but I'm thinking about increasing the primary
> filter to 47uF
> or 100 uF.
>
> Idle is 10mA and, as has become my common practice with
> cathode bias,
> I use a bypassed current mirror to keep a tight DC balance
> through the
> OPT.

Is that one mirror or two? I have wondered here in the past
whether it's best to use one of the valves as a reference,
or to use a seperate SS reference and two mirrors.

> Preamp uses a common Rk PFB setup between the gain triode
> and
> concertina phase splitter for a gain of about 140. The amp
> is then
> closed with 18dB of.global NFB.
>
> I used a 1/4 inch jack for the speaker connection to load
> switch
> internal 10W sand block dummies, so the amp is never
> unloaded, and,
> more to the point, the dummies provide a load when using
> headphones.

How does that work out SPL-wise? If you switch from speakers
to 'phones is the apparent volume similar?

> Might not be a monster amp but it's cute, compact, and fun
> as all get
> out. The whole ball of wax, sans the wallwart of course,
> fits in a
> 5"x7" chassis.
>
> For speakers I worked up a 300cu ported enclosure wrapped
> around the
> MCM 1853 4 inch driver. They seem to have a response peak
> around 2kHz
> but rather then smooth that out with passives in the cab I
> just use
> the PC's equalizer and they sound amazingly good for 14
> buck cheapies.
>
>
> Darndest thing I ever saw, not that I've seen
> 'everything', but one of
> the 6AW8s I got in a batch has a heater short. Not to
> cathode, or any
> other pin, the heater itself. She pulls heavy current but
> won't light,
> as if there's solid wire between the pins. Can't complain
> too much, I
> guess, what with that batch being 50 cents each. Just wish
> he'd have
> had more of them.
>
> I'll post pics in alt.binaries.schematics.electronic if
> folks would
> like to see it.

Circuit would be nice too. Don't you have any web space?
Easier for us.

Thanks, Ian
>

Re: 'PC Speaker' Tube Amp

On Wed, 7 May 2008 19:13:08 +0100, "Ian Iveson"
<IanIveson.home@blueyonder.co.uk> wrote:

>Flipper said:
>
>> That's the latest project I finished, a little 'PC
>> Speaker' Tube
>> amplifier using, per channel, a pair of 6AW8s PP into 20k
>> OPTs. Does
>> between 3 and 4 Watts Po per channel.
>>
>> Similar to my mini guitar amp it's powered off a 12VAC
>> (40VA in this
>> case) wallwart with a reverse 20VA tranny inside
>> generating about 240V
>> (loaded) for B+. Screens are at 200V.
>>
>> PS filtering is 33uF --> OPT common --> 4.7k 33uF --->
>> screens --->
>> 10k 33uF ---> Preamp. Hum is only detectable if you stick
>> your ear on
>> the cones but I'm thinking about increasing the primary
>> filter to 47uF
>> or 100 uF.
>>
>> Idle is 10mA and, as has become my common practice with
>> cathode bias,
>> I use a bypassed current mirror to keep a tight DC balance
>> through the
>> OPT.
>
>Is that one mirror or two?

One per channel. One per tube tubes but two in total.

> I have wondered here in the past
>whether it's best to use one of the valves as a reference,

It has an interesting effect, one bad and one good, sort of. If the
tube on Rk is leaky then the current mirror drags the other with it.
On the other hand, if the leaky one is on the 'controlled' side the
current mirror adjusts so it works.

Ran into that on the first one I did, using 6EM7s. Unbeknownst to me
one of them was leaky but, by happenstance, it was on the 'controlled'
side and things appeared fine. But in the process of making changes,
the tubes got swapped and idle current went through the roof on both.

>or to use a seperate SS reference and two mirrors.

Well, that's separate current sources, the mirror just being how you
get two current sources, and bias pulls apart more quickly that way
because the CCS is fixed and 'insistent'. With this approach the Rk
side is as soft as the Rk, meaning it will 'allow' a current increase,
and as bias shifts up from the increased current the mirror expects
more current too. So the mirror would lower bias if not for that tube
also conducting more, like the Rk side, and bias doesn't change as
much (on either side) as it would with an obstinate CCS insisting that
idle is 'it' and no more.


>> Preamp uses a common Rk PFB setup between the gain triode
>> and
>> concertina phase splitter for a gain of about 140. The amp
>> is then
>> closed with 18dB of.global NFB.
>>
>> I used a 1/4 inch jack for the speaker connection to load
>> switch
>> internal 10W sand block dummies, so the amp is never
>> unloaded, and,
>> more to the point, the dummies provide a load when using
>> headphones.
>
>How does that work out SPL-wise? If you switch from speakers
>to 'phones is the apparent volume similar?

Depends on the headphones but they're not straight on the dummy load.
I use a series R, like normal, and you pick it for 'best fit'.


>> Might not be a monster amp but it's cute, compact, and fun
>> as all get
>> out. The whole ball of wax, sans the wallwart of course,
>> fits in a
>> 5"x7" chassis.
>>
>> For speakers I worked up a 300cu ported enclosure wrapped
>> around the
>> MCM 1853 4 inch driver. They seem to have a response peak
>> around 2kHz
>> but rather then smooth that out with passives in the cab I
>> just use
>> the PC's equalizer and they sound amazingly good for 14
>> buck cheapies.
>>
>>
>> Darndest thing I ever saw, not that I've seen
>> 'everything', but one of
>> the 6AW8s I got in a batch has a heater short. Not to
>> cathode, or any
>> other pin, the heater itself. She pulls heavy current but
>> won't light,
>> as if there's solid wire between the pins. Can't complain
>> too much, I
>> guess, what with that batch being 50 cents each. Just wish
>> he'd have
>> had more of them.
>>
>> I'll post pics in alt.binaries.schematics.electronic if
>> folks would
>> like to see it.
>
>Circuit would be nice too. Don't you have any web space?
>Easier for us.

Well, I do but it's all screwed up right now. I'll look at it tonight
and see if something can be made to work.


>Thanks, Ian
>>
>

Re: 'PC Speaker' Tube Amp

On Wed, 07 May 2008 15:12:27 -0500, flipper <flipper@fish . net > wrote:

>On Wed, 7 May 2008 19:13:08 +0100, "Ian Iveson"
><IanIveson.home@blueyonder.co.uk> wrote:
>
>>Flipper said:
>>
>>> That's the latest project I finished, a little 'PC
>>> Speaker' Tube
>>> amplifier using, per channel, a pair of 6AW8s PP into 20k
>>> OPTs. Does
>>> between 3 and 4 Watts Po per channel.
>>>
>>> Similar to my mini guitar amp it's powered off a 12VAC
>>> (40VA in this
>>> case) wallwart with a reverse 20VA tranny inside
>>> generating about 240V
>>> (loaded) for B+. Screens are at 200V.
>>>
>>> PS filtering is 33uF --> OPT common --> 4.7k 33uF --->
>>> screens --->
>>> 10k 33uF ---> Preamp. Hum is only detectable if you stick
>>> your ear on
>>> the cones but I'm thinking about increasing the primary
>>> filter to 47uF
>>> or 100 uF.
>>>
>>> Idle is 10mA and, as has become my common practice with
>>> cathode bias,
>>> I use a bypassed current mirror to keep a tight DC balance
>>> through the
>>> OPT.
>>
>>Is that one mirror or two?
>
>One per channel. One per tube tubes but two in total.
>
>> I have wondered here in the past
>>whether it's best to use one of the valves as a reference,
>
>It has an interesting effect, one bad and one good, sort of. If the
>tube on Rk is leaky then the current mirror drags the other with it.
>On the other hand, if the leaky one is on the 'controlled' side the
>current mirror adjusts so it works.
>
>Ran into that on the first one I did, using 6EM7s. Unbeknownst to me
>one of them was leaky but, by happenstance, it was on the 'controlled'
>side and things appeared fine. But in the process of making changes,
>the tubes got swapped and idle current went through the roof on both.
>
>>or to use a seperate SS reference and two mirrors.
>
>Well, that's separate current sources, the mirror just being how you
>get two current sources, and bias pulls apart more quickly that way
>because the CCS is fixed and 'insistent'. With this approach the Rk
>side is as soft as the Rk, meaning it will 'allow' a current increase,
>and as bias shifts up from the increased current the mirror expects
>more current too. So the mirror would lower bias if not for that tube
>also conducting more, like the Rk side, and bias doesn't change as
>much (on either side) as it would with an obstinate CCS insisting that
>idle is 'it' and no more.
>
>
>>> Preamp uses a common Rk PFB setup between the gain triode
>>> and
>>> concertina phase splitter for a gain of about 140. The amp
>>> is then
>>> closed with 18dB of.global NFB.
>>>
>>> I used a 1/4 inch jack for the speaker connection to load
>>> switch
>>> internal 10W sand block dummies, so the amp is never
>>> unloaded, and,
>>> more to the point, the dummies provide a load when using
>>> headphones.
>>
>>How does that work out SPL-wise? If you switch from speakers
>>to 'phones is the apparent volume similar?
>
>Depends on the headphones but they're not straight on the dummy load.
>I use a series R, like normal, and you pick it for 'best fit'.
>
>
>>> Might not be a monster amp but it's cute, compact, and fun
>>> as all get
>>> out. The whole ball of wax, sans the wallwart of course,
>>> fits in a
>>> 5"x7" chassis.
>>>
>>> For speakers I worked up a 300cu ported enclosure wrapped
>>> around the
>>> MCM 1853 4 inch driver. They seem to have a response peak
>>> around 2kHz
>>> but rather then smooth that out with passives in the cab I
>>> just use
>>> the PC's equalizer and they sound amazingly good for 14
>>> buck cheapies.
>>>
>>>
>>> Darndest thing I ever saw, not that I've seen
>>> 'everything', but one of
>>> the 6AW8s I got in a batch has a heater short. Not to
>>> cathode, or any
>>> other pin, the heater itself. She pulls heavy current but
>>> won't light,
>>> as if there's solid wire between the pins. Can't complain
>>> too much, I
>>> guess, what with that batch being 50 cents each. Just wish
>>> he'd have
>>> had more of them.
>>>
>>> I'll post pics in alt.binaries.schematics.electronic if
>>> folks would
>>> like to see it.
>>
>>Circuit would be nice too. Don't you have any web space?
>>Easier for us.
>
>Well, I do but it's all screwed up right now. I'll look at it tonight
>and see if something can be made to work.

Ok, I found the problem. I don't have the MAC address for that
server's NIC and for some inexplicable reason the DHCP server decided
to 'move' it so the firewall rules didn't route to it any more.

Anyway, that should be back up, assuming the server doesn't crash, and
I moved the schematic and some pics over to it.

Web page is * flipperhome.dyndns.org The 6AW8 amp is the last one
in the projects list.

I did notice one thing, though. I'm not sure that particular zoble is
what I finally ended up with and same with the headphone resistor.
Those were my "start with something" values. But that's minor tweaking
stuff.



>
>
>>Thanks, Ian
>>>
>>

Re: 'PC Speaker' Tube Amp

flipper wrote:

>>>Is that one mirror or two?
>>
>>One per channel. One per tube tubes but two in total.
>>
>>> I have wondered here in the past
>>>whether it's best to use one of the valves as a
>>>reference,
>>
>>It has an interesting effect, one bad and one good, sort
>>of. If the
>>tube on Rk is leaky then the current mirror drags the
>>other with it.
>>On the other hand, if the leaky one is on the 'controlled'
>>side the
>>current mirror adjusts so it works.
>>
>>Ran into that on the first one I did, using 6EM7s.
>>Unbeknownst to me
>>one of them was leaky but, by happenstance, it was on the
>>'controlled'
>>side and things appeared fine. But in the process of
>>making changes,
>>the tubes got swapped and idle current went through the
>>roof on both.
>>
>>>or to use a separate SS reference and two mirrors.
>>
>>Well, that's separate current sources, the mirror just
>>being how you
>>get two current sources, and bias pulls apart more quickly
>>that way
>>because the CCS is fixed and 'insistent'. With this
>>approach the Rk
>>side is as soft as the Rk, meaning it will 'allow' a
>>current increase,
>>and as bias shifts up from the increased current the
>>mirror expects
>>more current too. So the mirror would lower bias if not
>>for that tube
>>also conducting more, like the Rk side, and bias doesn't
>>change as
>>much (on either side) as it would with an obstinate CCS
>>insisting that
>>idle is 'it' and no more.
>>
>>
>>>> Preamp uses a common Rk PFB setup between the gain
>>>> triode
>>>> and
>>>> concertina phase splitter for a gain of about 140. The
>>>> amp
>>>> is then
>>>> closed with 18dB of.global NFB.
>>>>
>>>> I used a 1/4 inch jack for the speaker connection to
>>>> load
>>>> switch
>>>> internal 10W sand block dummies, so the amp is never
>>>> unloaded, and,
>>>> more to the point, the dummies provide a load when
>>>> using
>>>> headphones.
>>>
>>>How does that work out SPL-wise? If you switch from
>>>speakers
>>>to 'phones is the apparent volume similar?
>>
>>Depends on the headphones but they're not straight on the
>>dummy load.
>>I use a series R, like normal, and you pick it for 'best
>>fit'.
>>
>>
>>>> Might not be a monster amp but it's cute, compact, and
>>>> fun
>>>> as all get
>>>> out. The whole ball of wax, sans the wallwart of
>>>> course,
>>>> fits in a
>>>> 5"x7" chassis.
>>>>
>>>> For speakers I worked up a 300cu ported enclosure
>>>> wrapped
>>>> around the
>>>> MCM 1853 4 inch driver. They seem to have a response
>>>> peak
>>>> around 2kHz
>>>> but rather then smooth that out with passives in the
>>>> cab I
>>>> just use
>>>> the PC's equalizer and they sound amazingly good for 14
>>>> buck cheapies.
>>>>
>>>>
>>>> Darndest thing I ever saw, not that I've seen
>>>> 'everything', but one of
>>>> the 6AW8s I got in a batch has a heater short. Not to
>>>> cathode, or any
>>>> other pin, the heater itself. She pulls heavy current
>>>> but
>>>> won't light,
>>>> as if there's solid wire between the pins. Can't
>>>> complain
>>>> too much, I
>>>> guess, what with that batch being 50 cents each. Just
>>>> wish
>>>> he'd have
>>>> had more of them.
>>>>
>>>> I'll post pics in alt.binaries.schematics.electronic if
>>>> folks would
>>>> like to see it.
>>>
>>>Circuit would be nice too. Don't you have any web space?
>>>Easier for us.
>>
>>Well, I do but it's all screwed up right now. I'll look at
>>it tonight
>>and see if something can be made to work.
>
> Ok, I found the problem. I don't have the MAC address for
> that
> server's NIC and for some inexplicable reason the DHCP
> server decided
> to 'move' it so the firewall rules didn't route to it any
> more.
>
> Anyway, that should be back up, assuming the server
> doesn't crash, and
> I moved the schematic and some pics over to it.
>
> Web page is * flipperhome.dyndns.org The 6AW8 amp is
> the last one
> in the projects list.
>
> I did notice one thing, though. I'm not sure that
> particular zoble is
> what I finally ended up with and same with the headphone
> resistor.
> Those were my "start with something" values. But that's
> minor tweaking
> stuff.

Thanks. One or two details might be worth discussing; I'm
always interested to know how ppl make decisions.

Why did you decide not to direct-couple the voltage amp to
the concertina, and to share the valve halves by channel
rather than by function?

What's going on with the connection of the concertina's
cathode load to the cathode of the voltage amp, rather than
to ground? How well is the splitter balanced, given the
different total loads?

I can't see how the output valves' bias is adjusted.

Why do Zobels go to anode supply, rather than to screens? I
have read quite a lot of waffle around this subject but
nothing very convincing. What, exactly, are those Zobels
for?

Using a resistor in series with headphones isn't ideal, to
my mind. My opinion is that a head amp should be have an
output impedance of no more than 5 ohms, or one tenth of the
'phones' impedance.

Nice pictures. Speakers are beautifully done. Amp looks
functional. Valves look a bit gloomy. Transformers look
cheerful but very small. How's your bass?

Ian

Re: 'PC Speaker' Tube Amp

In article <KKLUj.38101$AL4.35754@newsfe23.ams2>,
"Ian Iveson" <IanIveson.home@blueyonder.co.uk> wrote:

> Thanks. One or two details might be worth discussing; I'm
> always interested to know how ppl make decisions.

Hi Ian, I hope you don't mind my jumping into your discussion.

> Why did you decide not to direct-couple the voltage amp to
> the concertina, and to share the valve halves by channel
> rather than by function?

Conversely why would one want to direct couple voltage amp and the
concertina stage? By avoiding direct coupling the design of each stage
can be optimized. The only down side is that somewhat greater care must
be taken in the choice of the three low frequency time constants.

> What's going on with the connection of the concertina's
> cathode load to the cathode of the voltage amp, rather than
> to ground? How well is the splitter balanced, given the
> different total loads?

Flipper explained that it is positive feedback used to increase the open
loop gain of the amplifier. This is common in phono preamp stages, and
not unheard of in power amps. As far as precise balance of the
concertina loads goes, aren't you being a bit anal, they are close
enough.

> I can't see how the output valves' bias is adjusted.

I assumed it isn't adjusted, why should an adjustment be provided?

> Why do Zobels go to anode supply, rather than to screens? I
> have read quite a lot of waffle around this subject but
> nothing very convincing. What, exactly, are those Zobels
> for?

When I use Zobels I connect them to the anode supply as flipper did, why
would you want to connect them to the screens?

> Using a resistor in series with headphones isn't ideal, to
> my mind. My opinion is that a head amp should be have an
> output impedance of no more than 5 ohms, or one tenth of the
> 'phones' impedance.

I'm not into headphones and hence don't understand how to properly
interface them. That said I was surprised that flipper didn't find some
attenuation necessary beyond that provided by a series 51 Ohm resistor?
I would think that a different circuit could be found that would meet
your 5-Ohm impedance criteria, perhaps by simply tapping the headphone
output doen on the 8 Ohm load resistors and eliminating the 51 Ohm
series resistor. Not understanding headphone circuits I would be
interested in learning what flipper's criteria were for the design of
this part of his amplifier?

> Nice pictures. Speakers are beautifully done. Amp looks
> functional. Valves look a bit gloomy. Transformers look
> cheerful but very small. How's your bass?

The transformers do look "cheerful", but I don't understand why you say
the "Valves look a bit gloomy", they look quite bright to me?


Regards,

John Byrns

--
Surf my web pages at, * fmamradios . com /

Re: 'PC Speaker' Tube Amp

John Byrns wrote:
>
> Hi Ian, I hope you don't mind my jumping into your
> discussion.

Not at all, I always hope that's what discussions will be
for.

>> Why did you decide not to direct-couple the voltage amp
>> to
>> the concertina, and to share the valve halves by channel
>> rather than by function?
>
> Conversely why would one want to direct couple voltage amp
> and the
> concertina stage? By avoiding direct coupling the design
> of each stage
> can be optimized. The only down side is that somewhat
> greater care must
> be taken in the choice of the three low frequency time
> constants.

Well, quite. Some folks do it one way and some another. Some
might go to the extreme of direct coupling and use a -ve
supply for the first stage, so that both stages can still be
optimised, presumably.

In general, it is a bit odd that, after so many years, no
particular valve amp design has become ubiquitous, as I am
told it has become in the SS world for a large swathe of the
market.

>> What's going on with the connection of the concertina's
>> cathode load to the cathode of the voltage amp, rather
>> than
>> to ground? How well is the splitter balanced, given the
>> different total loads?
>
> Flipper explained that it is positive feedback used to
> increase the open
> loop gain of the amplifier. This is common in phono
> preamp stages, and
> not unheard of in power amps.

OK, sorry flipper. But why, exactly? Is there some general
consideration that leads a designer, in this or that
situation, to nest a lesser positive loop inside a greater,
negative one? I imagine it may be possible to get more gain
out of a relatively linear section of the amp, so that can
be used to counter the distortion produced by a relatively
non-linear section somewhere else in the global loop. Or
perhaps I am missing some advantage in the frequency domain?

> As far as precise balance of the
> concertina loads goes, aren't you being a bit anal, they
> are close
> enough.

Some folk go to extraordinary lengths to try and AC balance
concertinas precisely at all frequencies. I don't know where
fidelity stops and anal begins for everybody, and it's
interesting to know where other folk think that point is.

>> I can't see how the output valves' bias is adjusted.
>
> I assumed it isn't adjusted, why should an adjustment be
> provided?

I'd want to mess about with it :)

It could be that old valves would be better off with a bit
less current. This is quite an issue, to my mind, especially
as it is relatively simple these days to add some
intelligence to the bias circuit. What is the best way to
keep valves optimally biased as they age?

>> Why do Zobels go to anode supply, rather than to screens?
>> I
>> have read quite a lot of waffle around this subject but
>> nothing very convincing. What, exactly, are those Zobels
>> for?
>
> When I use Zobels I connect them to the anode supply as
> flipper did, why
> would you want to connect them to the screens?

Just asking. Not my amp so I can't know where I would put
them. But there is a more general issue here, about exactly
what such Zobels are designed to achieve. I see a logical
argument for Zobels between anodes and screens but not
between anodes and ground or anode supply. The logic may
well be misplaced...

So what, exactly, are these Zobels for? I had assumed they
are to deal with some problematic local relationship, and
that must be between anode and screen...perhaps because the
anode circuit contains the leakage inductance and winding
capacitance, whereas the screen circuit doesn't. I further
surmised that the most direct approach to damping would be
the best. Maybe there's not much difference.

>> Using a resistor in series with headphones isn't ideal,
>> to
>> my mind. My opinion is that a head amp should be have an
>> output impedance of no more than 5 ohms, or one tenth of
>> the
>> 'phones' impedance.
>
> I'm not into headphones and hence don't understand how to
> properly
> interface them. That said I was surprised that flipper
> didn't find some
> attenuation necessary beyond that provided by a series 51
> Ohm resistor?
> I would think that a different circuit could be found that
> would meet
> your 5-Ohm impedance criteria, perhaps by simply tapping
> the headphone
> output doen on the 8 Ohm load resistors and eliminating
> the 51 Ohm
> series resistor. Not understanding headphone circuits I
> would be
> interested in learning what flipper's criteria were for
> the design of
> this part of his amplifier?

Yes. It could be that the 'phones output was an
afterthought, as they have tended to be over the years. Or
it may be just me, being anal about headphones...

Headphone amps are much more important now, and hi-fidelity
examples will be a big thing in future. Increasingly, music
is designed for heads rather than rooms.

>> Nice pictures. Speakers are beautifully done. Amp looks
>> functional. Valves look a bit gloomy. Transformers look
>> cheerful but very small. How's your bass?
>
> The transformers do look "cheerful", but I don't
> understand why you say
> the "Valves look a bit gloomy", they look quite bright to
> me?

They look a bit drab to me, and the glass looks smoked in
the big photo. Not a serious complaint...some folk are fond
of industrial chic.

Cheers, Ian

Re: 'PC Speaker' Tube Amp

On Fri, 9 May 2008 04:26:34 +0100, "Ian Iveson"
<IanIveson.home@blueyonder.co.uk> wrote:

>John Byrns wrote:
<snip>

>>
>> Flipper explained that it is positive feedback used to
>> increase the open
>> loop gain of the amplifier. This is common in phono
>> preamp stages, and
>> not unheard of in power amps.
>
>OK, sorry flipper. But why, exactly? Is there some general
>consideration that leads a designer, in this or that
>situation, to nest a lesser positive loop inside a greater,
>negative one? I imagine it may be possible to get more gain
>out of a relatively linear section of the amp, so that can
>be used to counter the distortion produced by a relatively
>non-linear section somewhere else in the global loop. Or
>perhaps I am missing some advantage in the frequency domain?
>

Sorry for the delay but you answered the question just as you asked
it.

It's to get more gain for the subsequent 18dB global NFB, with enough
left over for good sensitivity.

It wouldn't do any good if it were over just one stage because PFB
increases distortion and the NFB would just get you back to where you
started. But it's quite useful with multiple stages where one stage,
(or more) is better than the other, as if the case with the gain
triode vs the PP output stage. It has the effect of 'shifting' some of
that linearity from one to the other so there's a net overall
improvement even if NFB=PFB.

The topology I used is almost identical to that in the Harmon Kardon
A300 and A500 except they used 1/2 a 12AX7 for the gain triode and 1/2
a 12AU7 for the phase splitter. They used a plate boot strapped
version in the A224 and A230, both from a single 12AX7.

I've used both, as well as a variant with SS for the phase splitter,
but chose the (cap coupled) A300 version in this case because it's
less susceptible to bias shift over tube variations (note that the PFB
operates on the bias just as it does signal). I.E. The gain triode
plate can end up 'anywhere' without throwing off the phase splitter
operating point and I wanted to be able to substitute 6JW8s for the
6AW8s, which potentially aggravates the tube to tube bias variation.

Re: 'PC Speaker' Tube Amp

On Thu, 15 May 2008 10:31:03 -0500, flipper <flipper@fish . net > wrote:


>On Fri, 9 May 2008 04:26:34 +0100, "Ian Iveson"
><IanIveson.home@blueyonder.co.uk> wrote:
.... snip!....
>
>It wouldn't do any good if it were over just one stage because PFB
>increases distortion and the NFB would just get you back to where you
>started. But it's quite useful with multiple stages where one stage,
>(or more) is better than the other, as if the case with the gain
>triode vs the PP output stage. It has the effect of 'shifting' some of
>that linearity from one to the other so there's a net overall
>improvement even if NFB=PFB.
>
....snip!......

I'm not sure about this... but.... isn't nonlinearity just an
unwanted signal, and in a feedback amp with any number of internal
loops, isn't that signal reduced by the "loop gain", which to my
memory is the overall openloop gain divided by the overall closed loop
gain? That would mean that as far as unwanted signals go, the internal
loops would have no effect.
I'd wonder about slewing problems being reduced, but after a bit of
thought I don't think it would have a lot of effect, unless some of
the internal gains are reduced to the point that the prior stages must
overload, or can't drive input capacitances because of the high levels
required.
Am I all wet?

-Paul

Re: 'PC Speaker' Tube Amp

On Sun, 18 May 2008 22:15:00 GMT, Paul <paulguy@eastlink.ca> wrote:

>On Thu, 15 May 2008 10:31:03 -0500, flipper <flipper@fish . net > wrote:
>
>
>>On Fri, 9 May 2008 04:26:34 +0100, "Ian Iveson"
>><IanIveson.home@blueyonder.co.uk> wrote:
>.... snip!....
>>
>>It wouldn't do any good if it were over just one stage because PFB
>>increases distortion and the NFB would just get you back to where you
>>started. But it's quite useful with multiple stages where one stage,
>>(or more) is better than the other, as if the case with the gain
>>triode vs the PP output stage. It has the effect of 'shifting' some of
>>that linearity from one to the other so there's a net overall
>>improvement even if NFB=PFB.
>>
>....snip!......
>
> I'm not sure about this... but.... isn't nonlinearity just an
>unwanted signal, and in a feedback amp with any number of internal
>loops, isn't that signal reduced by the "loop gain", which to my
>memory is the overall openloop gain divided by the overall closed loop
>gain? That would mean that as far as unwanted signals go, the internal
>loops would have no effect.
> I'd wonder about slewing problems being reduced, but after a bit of
>thought I don't think it would have a lot of effect, unless some of
>the internal gains are reduced to the point that the prior stages must
>overload, or can't drive input capacitances because of the high levels
>required.
> Am I all wet?
>
>-Paul

I don't know how you got the idea that local NFB has no effect. Local
FB works just like global with the difference being how long the loop
is.

Your basic gain vs distortion math is correct but it applies to any
FB, not just global. I.E. Apply local NFB around (or in) one stage
and the distortion of that stage is reduced in the same manner as
overall amp distortion is reduced by global NFB (or vice versa with
PFB). e.g. Take just a single triode. What would the gain be without
the Rk (using, say, fixed bias instead)? Now, what is it with the Rk?
That's how much NFB the Rk adds and distortion is reduced by that
amount. Or, conversely, that's how much increase in distortion there
is when you bypass Rk to get more gain on that stage.

However, it doesn't do you any good to have a squeaky clean front end
and then feed it into a heavily (by comparison) distorted power stage
because distortion over multiple stages is the square root of the sum
of the individual squares so if one section is significantly higher
than the other it dominates the overall distortion figure (and by more
than if it were a linear relationship).

But you also have to have the gain in order to apply global NFB so
it's better to use some local PFB, to get the gain with some increase
in distortion of the squeaky clean stage and use that acquired gain to
apply global, thereby reducing the overall distortion.

The fallacy some engineers fall into is observing that equal PFB and
NFB cancels so they conclude you gain (pun) nothing. I.E. You add
10dB, then subtract 10dB, and you're right back where you started. Or,
when not equal, why add 10dB and subtract 18dB when just subtracting
8dB gets the same result? And that would be true if it was the same
loop but what they miss is the PFB is local, over the much better
stage, while the NFB is global and, as mentioned, the combined
distortion is not linear but the square root of the sum of the
squares.

Of course, there are other ways to achieve the same result but I get a
kick out of this one because, component wise, it's 'free' as the
difference (besides picking good component values) is simply
terminating the phase splitter into Rk rather than ground.

Re: 'PC Speaker' Tube Amp

On Sun, 18 May 2008 22:41:24 -0500, flipper <flipper@fish . net > wrote:

.....snip!......
>
>I don't know how you got the idea that local NFB has no effect. Local
>FB works just like global with the difference being how long the loop
>is.
>
>Your basic gain vs distortion math is correct but it applies to any
>FB, not just global. I.E. Apply local NFB around (or in) one stage
>and the distortion of that stage is reduced in the same manner as
>overall amp distortion is reduced by global NFB (or vice versa with
>PFB). e.g. Take just a single triode. What would the gain be without
>the Rk (using, say, fixed bias instead)? Now, what is it with the Rk?
>That's how much NFB the Rk adds and distortion is reduced by that
>amount. Or, conversely, that's how much increase in distortion there
>is when you bypass Rk to get more gain on that stage.
>
>However, it doesn't do you any good to have a squeaky clean front end
>and then feed it into a heavily (by comparison) distorted power stage
>because distortion over multiple stages is the square root of the sum
>of the individual squares so if one section is significantly higher
>than the other it dominates the overall distortion figure (and by more
>than if it were a linear relationship).
>
>But you also have to have the gain in order to apply global NFB so
>it's better to use some local PFB, to get the gain with some increase
>in distortion of the squeaky clean stage and use that acquired gain to
>apply global, thereby reducing the overall distortion.
>
>The fallacy some engineers fall into is observing that equal PFB and
>NFB cancels so they conclude you gain (pun) nothing. I.E. You add
>10dB, then subtract 10dB, and you're right back where you started. Or,
>when not equal, why add 10dB and subtract 18dB when just subtracting
>8dB gets the same result? And that would be true if it was the same
>loop but what they miss is the PFB is local, over the much better
>stage, while the NFB is global and, as mentioned, the combined
>distortion is not linear but the square root of the sum of the
>squares.
>
>Of course, there are other ways to achieve the same result but I get a
>kick out of this one because, component wise, it's 'free' as the
>difference (besides picking good component values) is simply
>terminating the phase splitter into Rk rather than ground.


Can I summarize this as follows? :
A linear disturbance (no sqrt of sums of squares) would not benefit
from juggling gains of local feedback paths within a global feedback.
This assumes that you haven't messed local gains up so badly that
overloading or slew rate limiting occur.
A disturbance that adds non-linearly (square root of sums of
sqyares) WOULD benefit. That makes sense, you would want to linearize
(more local feedback) the offending section, and boost gain (lower
negative feedback, or POSITIVE feedback) in the other amplifier
sections to make up the required overall global feedback or loop gain.

Now here's MY problem..... it's a mathematical one. It's my belief
that square root of sums of squares applies to uncorreleated signals
like white or thermal noise. Distortion is correlated (it's a
ploynomial base), and I don't think that the non-linear sqr root of
sums of squares (RMS) applies here.
For noise issues, yep, I have no doubt that juggling gains will
help, but for non-linearity, it's still my belief that loop gain
(overall) is what will reduce your distortion.

This is a tricky question, and it gets down to some hairy
theoretical issues..... do you know for sure that non-linearity
(distortion) is in fact added up in the sqrt or sums of squares? You
can decompose nonlinearity into a fourier series, and treat the
distortion as added sine waves, and they aren't random!
It was nice that you clearly backed up your assumptions, so I can
argue the point without getting personal !

Thanks,
Paul

Re: 'PC Speaker' Tube Amp

In article <ar5334desj8go24gh5atqp7f5e4mvr711t@4ax . com >,
Paul <paulguy@eastlink.ca> wrote:

> On Sun, 18 May 2008 22:41:24 -0500, flipper <flipper@fish . net > wrote:
>
> .....snip!......
> >
> >I don't know how you got the idea that local NFB has no effect. Local
> >FB works just like global with the difference being how long the loop
> >is.
> >
> >Your basic gain vs distortion math is correct but it applies to any
> >FB, not just global. I.E. Apply local NFB around (or in) one stage
> >and the distortion of that stage is reduced in the same manner as
> >overall amp distortion is reduced by global NFB (or vice versa with
> >PFB). e.g. Take just a single triode. What would the gain be without
> >the Rk (using, say, fixed bias instead)? Now, what is it with the Rk?
> >That's how much NFB the Rk adds and distortion is reduced by that
> >amount. Or, conversely, that's how much increase in distortion there
> >is when you bypass Rk to get more gain on that stage.
> >
> >However, it doesn't do you any good to have a squeaky clean front end
> >and then feed it into a heavily (by comparison) distorted power stage
> >because distortion over multiple stages is the square root of the sum
> >of the individual squares so if one section is significantly higher
> >than the other it dominates the overall distortion figure (and by more
> >than if it were a linear relationship).
> >
> >But you also have to have the gain in order to apply global NFB so
> >it's better to use some local PFB, to get the gain with some increase
> >in distortion of the squeaky clean stage and use that acquired gain to
> >apply global, thereby reducing the overall distortion.
> >
> >The fallacy some engineers fall into is observing that equal PFB and
> >NFB cancels so they conclude you gain (pun) nothing. I.E. You add
> >10dB, then subtract 10dB, and you're right back where you started. Or,
> >when not equal, why add 10dB and subtract 18dB when just subtracting
> >8dB gets the same result? And that would be true if it was the same
> >loop but what they miss is the PFB is local, over the much better
> >stage, while the NFB is global and, as mentioned, the combined
> >distortion is not linear but the square root of the sum of the
> >squares.
> >
> >Of course, there are other ways to achieve the same result but I get a
> >kick out of this one because, component wise, it's 'free' as the
> >difference (besides picking good component values) is simply
> >terminating the phase splitter into Rk rather than ground.
>
>
> Can I summarize this as follows? :
> A linear disturbance (no sqrt of sums of squares) would not benefit
> from juggling gains of local feedback paths within a global feedback.
> This assumes that you haven't messed local gains up so badly that
> overloading or slew rate limiting occur.
> A disturbance that adds non-linearly (square root of sums of
> sqyares) WOULD benefit. That makes sense, you would want to linearize
> (more local feedback) the offending section, and boost gain (lower
> negative feedback, or POSITIVE feedback) in the other amplifier
> sections to make up the required overall global feedback or loop gain.
>
> Now here's MY problem..... it's a mathematical one. It's my belief
> that square root of sums of squares applies to uncorreleated signals
> like white or thermal noise. Distortion is correlated (it's a
> ploynomial base), and I don't think that the non-linear sqr root of
> sums of squares (RMS) applies here.
> For noise issues, yep, I have no doubt that juggling gains will
> help, but for non-linearity, it's still my belief that loop gain
> (overall) is what will reduce your distortion.
>
> This is a tricky question, and it gets down to some hairy
> theoretical issues..... do you know for sure that non-linearity
> (distortion) is in fact added up in the sqrt or sums of squares? You
> can decompose nonlinearity into a fourier series, and treat the
> distortion as added sine waves, and they aren't random!
> It was nice that you clearly backed up your assumptions, so I can
> argue the point without getting personal !

I was following this thread in the beginning, when it was about an
amplifier, but I haven't been paying close attention since it got into
this PFB/NFB issue, I suppose I will have to go back and carefully read
the posts, now that I am about to comment without having fully read the
relevant posts.

Does it really make any real difference what his assumptions were, i.e.
linear vs. sqrt of sum of squares? Doesn't the strategy of adding PFB
to a relatively blameless stage to restore some or all of the gain lost
in applying NFB to reduce the overall distortion of multiple stages make
sense either way?

That said, without having given it more than 20 seconds thought I would
think the distortion of two stages would add, or subtract, both linearly
and by sqrt of sum of squares at the same time. I would think that the
amplitudes of the individual frequency components of the distortion
would add linearly while the total distortion power would be the sqrt of
sum of squares of the linear "sums" of those individual frequency
components. I have only given this 20 seconds thought, so it is
probably completely wrong.


Regards,

John Byrns

--
Surf my web pages at, * fmamradios . com /

Re: 'PC Speaker' Tube Amp

On Thu, 08 May 2008 20:27:22 -0500, John Byrns <byrnsj@sbcglobal . net >
wrote:

>In article <KKLUj.38101$AL4.35754@newsfe23.ams2>,
> "Ian Iveson" <IanIveson.home@blueyonder.co.uk> wrote:
>
>> Thanks. One or two details might be worth discussing; I'm
>> always interested to know how ppl make decisions.
>
>Hi Ian, I hope you don't mind my jumping into your discussion.
>
>> Why did you decide not to direct-couple the voltage amp to
>> the concertina, and to share the valve halves by channel
>> rather than by function?
>
>Conversely why would one want to direct couple voltage amp and the
>concertina stage? By avoiding direct coupling the design of each stage
>can be optimized. The only down side is that somewhat greater care must
>be taken in the choice of the three low frequency time constants.
>
>> What's going on with the connection of the concertina's
>> cathode load to the cathode of the voltage amp, rather than
>> to ground? How well is the splitter balanced, given the
>> different total loads?
>
>Flipper explained that it is positive feedback used to increase the open
>loop gain of the amplifier. This is common in phono preamp stages, and
>not unheard of in power amps. As far as precise balance of the
>concertina loads goes, aren't you being a bit anal, they are close
>enough.
>
>> I can't see how the output valves' bias is adjusted.
>
>I assumed it isn't adjusted, why should an adjustment be provided?
>
>> Why do Zobels go to anode supply, rather than to screens? I
>> have read quite a lot of waffle around this subject but
>> nothing very convincing. What, exactly, are those Zobels
>> for?
>
>When I use Zobels I connect them to the anode supply as flipper did, why
>would you want to connect them to the screens?

My first thought was obviously the same, since I drew it that way, but
after seeing his post and thinking about it I'm wondering if it's
simply for the tie point convenience.


>> Using a resistor in series with headphones isn't ideal, to
>> my mind. My opinion is that a head amp should be have an
>> output impedance of no more than 5 ohms, or one tenth of the
>> 'phones' impedance.
>
>I'm not into headphones and hence don't understand how to properly
>interface them. That said I was surprised that flipper didn't find some
>attenuation necessary beyond that provided by a series 51 Ohm resistor?

I think in practice I ended up using 56 ohms. That was also just a
'starter' value and I don't use headphones much at all so I haven't
gotten around to 'fine tuning' it but they do work. I was going to pop
them back in and see how close the volume match is but seems I
inadvertently left them at my brother in law's last time I was over
there. So that took care of that, for the moment.

The thing to keep in mind is amp power goes almost exclusively to the
dummy loads with the headphones getting essentially a 'voltage' from
it and at 3 watts Po 32 ohm headphones (mine) will get about 100mW
(after the 56 ohm drop). Now, that's peak and average program
material, with 6dB peaking, will be closer to 3/4 watt, which would
put 25mw into the headphones.

Problem is, SPL in both the headphones and speakers vary so widely
there's no real way to know without listening to the ones in use.

>I would think that a different circuit could be found that would meet
>your 5-Ohm impedance criteria, perhaps by simply tapping the headphone
>output doen on the 8 Ohm load resistors and eliminating the 51 Ohm
>series resistor. Not understanding headphone circuits I would be
>interested in learning what flipper's criteria were for the design of
>this part of his amplifier?

As I mentioned in the other post, 'copy cat' ;)

And the series resistor will 'boost' the voltage up, to some degree,
with higher impedance headphones so power doesn't drop off so much as
it would if fed the same voltage level (like off a tapped 8 ohm load).

Btw, I think 100mW 'max', at the rated amp power, into 32 ohm
headphones is what I used to pick the starter values with the 51 ohms
being back when Po was running a tad low and 56 ohms after some amp
tuning got me to the target 3 watts. In fact, it peaks over 3 watts so
I might increase that series R some more, if I ever find the blooming
headphones.


>> Nice pictures. Speakers are beautifully done. Amp looks
>> functional. Valves look a bit gloomy. Transformers look
>> cheerful but very small. How's your bass?
>
>The transformers do look "cheerful", but I don't understand why you say
>the "Valves look a bit gloomy", they look quite bright to me?
>
>
>Regards,
>
>John Byrns

Re: 'PC Speaker' Tube Amp

On Fri, 09 May 2008 00:43:25 -0500, flipper <flipper@fish . net > wrote:

>On Thu, 08 May 2008 20:27:22 -0500, John Byrns <byrnsj@sbcglobal . net >
>wrote:
>
>>In article <KKLUj.38101$AL4.35754@newsfe23.ams2>,
>> "Ian Iveson" <IanIveson.home@blueyonder.co.uk> wrote:
>>
>>> Thanks. One or two details might be worth discussing; I'm
>>> always interested to know how ppl make decisions.
>>
>>Hi Ian, I hope you don't mind my jumping into your discussion.
>>
>>> Why did you decide not to direct-couple the voltage amp to
>>> the concertina, and to share the valve halves by channel
>>> rather than by function?
>>
>>Conversely why would one want to direct couple voltage amp and the
>>concertina stage? By avoiding direct coupling the design of each stage
>>can be optimized. The only down side is that somewhat greater care must
>>be taken in the choice of the three low frequency time constants.
>>
>>> What's going on with the connection of the concertina's
>>> cathode load to the cathode of the voltage amp, rather than
>>> to ground? How well is the splitter balanced, given the
>>> different total loads?
>>
>>Flipper explained that it is positive feedback used to increase the open
>>loop gain of the amplifier. This is common in phono preamp stages, and
>>not unheard of in power amps. As far as precise balance of the
>>concertina loads goes, aren't you being a bit anal, they are close
>>enough.
>>
>>> I can't see how the output valves' bias is adjusted.
>>
>>I assumed it isn't adjusted, why should an adjustment be provided?
>>
>>> Why do Zobels go to anode supply, rather than to screens? I
>>> have read quite a lot of waffle around this subject but
>>> nothing very convincing. What, exactly, are those Zobels
>>> for?
>>
>>When I use Zobels I connect them to the anode supply as flipper did, why
>>would you want to connect them to the screens?
>
>My first thought was obviously the same, since I drew it that way, but
>after seeing his post and thinking about it I'm wondering if it's
>simply for the tie point convenience.
>
>
>>> Using a resistor in series with headphones isn't ideal, to
>>> my mind. My opinion is that a head amp should be have an
>>> output impedance of no more than 5 ohms, or one tenth of the
>>> 'phones' impedance.
>>
>>I'm not into headphones and hence don't understand how to properly
>>interface them. That said I was surprised that flipper didn't find some
>>attenuation necessary beyond that provided by a series 51 Ohm resistor?
>
>I think in practice I ended up using 56 ohms. That was also just a
>'starter' value and I don't use headphones much at all so I haven't
>gotten around to 'fine tuning' it but they do work. I was going to pop
>them back in and see how close the volume match is but seems I
>inadvertently left them at my brother in law's last time I was over
>there. So that took care of that, for the moment.
>
>The thing to keep in mind is amp power goes almost exclusively to the
>dummy loads with the headphones getting essentially a 'voltage' from
>it and at 3 watts Po 32 ohm headphones (mine) will get about 100mW
>(after the 56 ohm drop). Now, that's peak and average program
>material, with 6dB peaking, will be closer to 3/4 watt, which would
>put 25mw into the headphones.
>
>Problem is, SPL in both the headphones and speakers vary so widely
>there's no real way to know without listening to the ones in use.
>
>>I would think that a different circuit could be found that would meet
>>your 5-Ohm impedance criteria, perhaps by simply tapping the headphone
>>output doen on the 8 Ohm load resistors and eliminating the 51 Ohm
>>series resistor. Not understanding headphone circuits I would be
>>interested in learning what flipper's criteria were for the design of
>>this part of his amplifier?
>
>As I mentioned in the other post, 'copy cat' ;)
>
>And the series resistor will 'boost' the voltage up, to some degree,
>with higher impedance headphones so power doesn't drop off so much as
>it would if fed the same voltage level (like off a tapped 8 ohm load).
>
>Btw, I think 100mW 'max', at the rated amp power, into 32 ohm
>headphones is what I used to pick the starter values with the 51 ohms
>being back when Po was running a tad low and 56 ohms after some amp
>tuning got me to the target 3 watts. In fact, it peaks over 3 watts so
>I might increase that series R some more, if I ever find the blooming
>headphones.
>
>
>>> Nice pictures. Speakers are beautifully done. Amp looks
>>> functional. Valves look a bit gloomy. Transformers look
>>> cheerful but very small. How's your bass?
>>
>>The transformers do look "cheerful", but I don't understand why you say
>>the "Valves look a bit gloomy", they look quite bright to me?
>>
>>
>>Regards,
>>
>>John Byrns

I found the headphones. Turned out they had simply slid out of sight
in the car.

So I tried them again and I'd say they're too loud with the 56 ohms in
series. Works but it would be nicer to have a little more range on the
low end so I'll end up increasing that value.

Re: 'PC Speaker' Tube Amp

On Fri, 09 May 2008 08:50:59 -0500, flipper <flipper@fish . net > wrote:

>On Fri, 09 May 2008 00:43:25 -0500, flipper <flipper@fish . net > wrote:
>
>>On Thu, 08 May 2008 20:27:22 -0500, John Byrns <byrnsj@sbcglobal . net >
>>wrote:
>>
>>>In article <KKLUj.38101$AL4.35754@newsfe23.ams2>,
>>> "Ian Iveson" <IanIveson.home@blueyonder.co.uk> wrote:
>>>
>>>> Thanks. One or two details might be worth discussing; I'm
>>>> always interested to know how ppl make decisions.
>>>
>>>Hi Ian, I hope you don't mind my jumping into your discussion.
>>>
>>>> Why did you decide not to direct-couple the voltage amp to
>>>> the concertina, and to share the valve halves by channel
>>>> rather than by function?
>>>
>>>Conversely why would one want to direct couple voltage amp and the
>>>concertina stage? By avoiding direct coupling the design of each stage
>>>can be optimized. The only down side is that somewhat greater care must
>>>be taken in the choice of the three low frequency time constants.
>>>
>>>> What's going on with the connection of the concertina's
>>>> cathode load to the cathode of the voltage amp, rather than
>>>> to ground? How well is the splitter balanced, given the
>>>> different total loads?
>>>
>>>Flipper explained that it is positive feedback used to increase the open
>>>loop gain of the amplifier. This is common in phono preamp stages, and
>>>not unheard of in power amps. As far as precise balance of the
>>>concertina loads goes, aren't you being a bit anal, they are close
>>>enough.
>>>
>>>> I can't see how the output valves' bias is adjusted.
>>>
>>>I assumed it isn't adjusted, why should an adjustment be provided?
>>>
>>>> Why do Zobels go to anode supply, rather than to screens? I
>>>> have read quite a lot of waffle around this subject but
>>>> nothing very convincing. What, exactly, are those Zobels
>>>> for?
>>>
>>>When I use Zobels I connect them to the anode supply as flipper did, why
>>>would you want to connect them to the screens?
>>
>>My first thought was obviously the same, since I drew it that way, but
>>after seeing his post and thinking about it I'm wondering if it's
>>simply for the tie point convenience.
>>
>>
>>>> Using a resistor in series with headphones isn't ideal, to
>>>> my mind. My opinion is that a head amp should be have an
>>>> output impedance of no more than 5 ohms, or one tenth of the
>>>> 'phones' impedance.
>>>
>>>I'm not into headphones and hence don't understand how to properly
>>>interface them. That said I was surprised that flipper didn't find some
>>>attenuation necessary beyond that provided by a series 51 Ohm resistor?
>>
>>I think in practice I ended up using 56 ohms. That was also just a
>>'starter' value and I don't use headphones much at all so I haven't
>>gotten around to 'fine tuning' it but they do work. I was going to pop
>>them back in and see how close the volume match is but seems I
>>inadvertently left them at my brother in law's last time I was over
>>there. So that took care of that, for the moment.
>>
>>The thing to keep in mind is amp power goes almost exclusively to the
>>dummy loads with the headphones getting essentially a 'voltage' from
>>it and at 3 watts Po 32 ohm headphones (mine) will get about 100mW
>>(after the 56 ohm drop). Now, that's peak and average program
>>material, with 6dB peaking, will be closer to 3/4 watt, which would
>>put 25mw into the headphones.
>>
>>Problem is, SPL in both the headphones and speakers vary so widely
>>there's no real way to know without listening to the ones in use.
>>
>>>I would think that a different circuit could be found that would meet
>>>your 5-Ohm impedance criteria, perhaps by simply tapping the headphone
>>>output doen on the 8 Ohm load resistors and eliminating the 51 Ohm
>>>series resistor. Not understanding headphone circuits I would be
>>>interested in learning what flipper's criteria were for the design of
>>>this part of his amplifier?
>>
>>As I mentioned in the other post, 'copy cat' ;)
>>
>>And the series resistor will 'boost' the voltage up, to some degree,
>>with higher impedance headphones so power doesn't drop off so much as
>>it would if fed the same voltage level (like off a tapped 8 ohm load).
>>
>>Btw, I think 100mW 'max', at the rated amp power, into 32 ohm
>>headphones is what I used to pick the starter values with the 51 ohms
>>being back when Po was running a tad low and 56 ohms after some amp
>>tuning got me to the target 3 watts. In fact, it peaks over 3 watts so
>>I might increase that series R some more, if I ever find the blooming
>>headphones.
>>
>>
>>>> Nice pictures. Speakers are beautifully done. Amp looks
>>>> functional. Valves look a bit gloomy. Transformers look
>>>> cheerful but very small. How's your bass?
>>>
>>>The transformers do look "cheerful", but I don't understand why you say
>>>the "Valves look a bit gloomy", they look quite bright to me?
>>>
>>>
>>>Regards,
>>>
>>>John Byrns
>
>I found the headphones. Turned out they had simply slid out of sight
>in the car.
>
>So I tried them again and I'd say they're too loud with the 56 ohms in
>series. Works but it would be nicer to have a little more range on the
>low end so I'll end up increasing that value.


Did a little research and it turns out IEC has a 'spec', IEC 61938,
for headphone 'interfaces'. Source impedance of 120 ohms driven by
5Vrms max. And, if the load were 8 ohms, 5Vrms would be, drum roll
please, 3.125 Watts... exactly (close enough for government work) what
my amp is putting into the dummy load. So a 120 ohm in series will be
'according to spec'.

Can't get much more convenient than that.