class A amplifier

[Kevin Aylward]

John Larkin wrote:

I mean, why should I build it? What possible purpose could that
serve, that would not be known already?



Ha! Thanks for the laugh. Most fun I've had for an hour, almost. [1]

Why?

I have noted that it is routine to design 50,000 transistor, SOC, analogue
i.cs and have the chip quite functional with 1st silicon. Indeed, for some
companies it is a known fact that designers have been paid Ł60k bonuses for
having 1st time success.

So, again, what purpose would it serve to build something, that can be
thoroughally characterised in simulation, when there is clearly no intention
of actually marketing a product? Why would anyone want to waste money and
time to do such a thing? To do so would seem to be quite daft to my mind.

I do understand that many feel the need to, and do, piddle about on the
bench, trying to get a one off to produce some vague waveforms, but many,
many more, do not.

Kevin Aylward
www.kevinaylward.co.uk
www.blonddee.co.uk
www.anasoft.co.uk - SuperSpice

 

[Kevin Aylward]

John Fields wrote:

On Sun, 19 Oct 2008 19:32:49 GMT, "Kevin Aylward"
kaExtractThis@kevinaylward.co.uk> wrote:

I mean, why should I build it? What possible purpose could that
serve, that would not be known already?

---
Wiring has a nasty habit of upsetting perfectly loaded applecarts.

And it what way do the properties of "wiring" transcend the laws of physics
such that such "wiring" is unable to be inclued in simulations?

Kevin Aylward
www.kevinaylward.co.uk
www.blonddee.co.uk
www.anasoft.co.uk - SuperSpice

 

[John Fields]

On Mon, 20 Oct 2008 17:33:20 GMT, "Kevin Aylward"
<kaExtractThis@kevinaylward.co.uk> wrote:

John Fields wrote:
On Sun, 19 Oct 2008 19:32:49 GMT, "Kevin Aylward"
kaExtractThis@kevinaylward.co.uk> wrote:

I mean, why should I build it? What possible purpose could that
serve, that would not be known already?

---
Wiring has a nasty habit of upsetting perfectly loaded applecarts.


And it what way do the properties of "wiring" transcend the laws of physics
such that such "wiring" is unable to be inclued in simulations?

---
They don't transcend the laws of physics at all, as you very well know,
Mr. Snotty, but the physical and electrical properties of FR-4, for
instance, (thickness of the matrix, thickness of the copper, variation
in the dielectric constant and resistivity of the matrix with
temperature and humidity, and on and on) would all have to be sorted out
and included in the simulation properly.

I think most people who design PCBs don't go to those lengths but,
rather, follow rules which will assure them a pretty good chance of
getting a good board the first time out.

How do you design PCBs?

JF

 

[John Larkin]

On Mon, 20 Oct 2008 17:33:20 GMT, "Kevin Aylward"
<kaExtractThis@kevinaylward.co.uk> wrote:

John Fields wrote:
On Sun, 19 Oct 2008 19:32:49 GMT, "Kevin Aylward"
kaExtractThis@kevinaylward.co.uk> wrote:

I mean, why should I build it? What possible purpose could that
serve, that would not be known already?

---
Wiring has a nasty habit of upsetting perfectly loaded applecarts.


And it what way do the properties of "wiring" transcend the laws of physics
such that such "wiring" is unable to be inclued in simulations?

I'm guessing that the output fets will RF oscillate big time.
Especially with those nice gate-gate caps downstream of the gate
resistors.

Wiring certainly matters here.

John

 

[John Larkin]

On Mon, 20 Oct 2008 17:33:01 GMT, "Kevin Aylward"
<kaExtractThis@kevinaylward.co.uk> wrote:

John Larkin wrote:

I mean, why should I build it? What possible purpose could that
serve, that would not be known already?



Ha! Thanks for the laugh. Most fun I've had for an hour, almost. [1]


Why?

I have noted that it is routine to design 50,000 transistor, SOC, analogue
i.cs and have the chip quite functional with 1st silicon. Indeed, for some
companies it is a known fact that designers have been paid Ł60k bonuses for
having 1st time success.

So, again, what purpose would it serve to build something, that can be
thoroughally characterised in simulation, when there is clearly no intention
of actually marketing a product? Why would anyone want to waste money and
time to do such a thing? To do so would seem to be quite daft to my mind.

I do understand that many feel the need to, and do, piddle about on the
bench, trying to get a one off to produce some vague waveforms, but many,
many more, do not.

Not me. I sell the things I design. By the thousands, preferably.

John

 

[John Larkin]

On Sat, 18 Oct 2008 08:26:03 +0100, Eeyore
<rabbitsfriendsandrelations@hotmail.com> wrote:

John Larkin wrote:

I suppose
the next step would be to modulate the sink current. The efficiency
would go up, but it would retain most of the virtues of the emitter
follower... simplicity, low open-loop distortion, stability, low Zout.

That's what I've been saying all along. It looks like a standard class AB output
stage but with the bias turned right up into the amps region.

Graham

No. The output is an NPN emitter and an NPN collector.

John

 

[Eeyore]

John Fields wrote:

Eeyore wrote:
John Fields wrote:
"Kevin Aylward" wrote:

I mean, why should I build it? What possible purpose could that serve, that
would not be known already?

---
Wiring has a nasty habit of upsetting perfectly loaded applecarts.

Then the applecart wasn't designed right.

---
Whoosh...

The perfectly loaded applecart (properly designed circuit) can easily be
upset if the PCB layout (wiring) is improperly done.

That's *tracking* not 'wiring'.

Graham

 

[Eeyore]

Kevin Aylward wrote:

John Larkin wrote:

I mean, why should I build it? What possible purpose could that
serve, that would not be known already?

Ha! Thanks for the laugh. Most fun I've had for an hour, almost. [1]

Why?

I have noted that it is routine to design 50,000 transistor, SOC, analogue
i.cs and have the chip quite functional with 1st silicon. Indeed, for some
companies it is a known fact that designers have been paid Ł60k bonuses for
having 1st time success.

So, again, what purpose would it serve to build something, that can be
thoroughally characterised in simulation, when there is clearly no intention
of actually marketing a product? Why would anyone want to waste money and
time to do such a thing? To do so would seem to be quite daft to my mind.

I do understand that many feel the need to, and do, piddle about on the
bench, trying to get a one off to produce some vague waveforms, but many,
many more, do not.

I was simulating using MathCad, some critical parts of my 'best to date' mosfet
amp and that's 20 years ago with DOS MathCad before anything got to the bench.
Since then I've only done bipolar amps (cost considerations).

Its performance is still only beaten by a very few.

Graham

 

[Eeyore]

John Larkin wrote:

Eeyore wrote:
John Larkin wrote:

I suppose
the next step would be to modulate the sink current. The efficiency
would go up, but it would retain most of the virtues of the emitter
follower... simplicity, low open-loop distortion, stability, low Zout.

That's what I've been saying all along. It looks like a standard class AB output
stage but with the bias turned right up into the amps region.

No. The output is an NPN emitter and an NPN collector.

They're cruddy. They were called 'quasi-complementary' in their classic configuration
and died an instant death the moment pnp complements arrived. Plus it's a more
complex design. What for ?

Graham

 

[John Larkin]

On Tue, 21 Oct 2008 00:23:23 +0100, Eeyore
<rabbitsfriendsandrelations@hotmail.com> wrote:

John Larkin wrote:

Eeyore wrote:
John Larkin wrote:

I suppose
the next step would be to modulate the sink current. The efficiency
would go up, but it would retain most of the virtues of the emitter
follower... simplicity, low open-loop distortion, stability, low Zout.

That's what I've been saying all along. It looks like a standard class AB output
stage but with the bias turned right up into the amps region.

No. The output is an NPN emitter and an NPN collector.

They're cruddy. They were called 'quasi-complementary' in their classic configuration
and died an instant death the moment pnp complements arrived. Plus it's a more
complex design. What for ?

Graham

Because Oliver wants to play with circuits.

Jonh

 

[Eeyore]

John Larkin wrote:

Eeyore wrote:
John Larkin wrote:
Eeyore wrote:
John Larkin wrote:

I suppose
the next step would be to modulate the sink current. The efficiency
would go up, but it would retain most of the virtues of the emitter
follower... simplicity, low open-loop distortion, stability, low Zout.

That's what I've been saying all along. It looks like a standard class AB output
stage but with the bias turned right up into the amps region.

No. The output is an NPN emitter and an NPN collector.

They're cruddy. They were called 'quasi-complementary' in their classic configuration
and died an instant death the moment pnp complements arrived. Plus it's a more
complex design. What for ?

Because Oliver wants to play with circuits.

Why not play with good ones ?

Graham

 

[John Fields]

On Tue, 21 Oct 2008 00:15:22 +0100, Eeyore
<rabbitsfriendsandrelations@hotmail.com> wrote:

John Fields wrote:

Eeyore wrote:
John Fields wrote:
"Kevin Aylward" wrote:

I mean, why should I build it? What possible purpose could that serve, that
would not be known already?

---
Wiring has a nasty habit of upsetting perfectly loaded applecarts.

Then the applecart wasn't designed right.

---
Whoosh...

The perfectly loaded applecart (properly designed circuit) can easily be
upset if the PCB layout (wiring) is improperly done.

That's *tracking* not 'wiring'.

---
Is that a Right-Pondian term?

I've always heard of it referred to and have called it "wiring" as in:
"You got the wiring wrong on that PCB."

Do you guys say: "You got the tracking wrong on that PCB"?

Sounds kind of ambiguous to me, like you're admonishing FEDEX or
somebody like that.

I take it, though, you take my point?

JF

 

[Kevin Aylward]

John Larkin wrote:

On Mon, 20 Oct 2008 17:33:20 GMT, "Kevin Aylward"
kaExtractThis@kevinaylward.co.uk> wrote:

John Fields wrote:
On Sun, 19 Oct 2008 19:32:49 GMT, "Kevin Aylward"
kaExtractThis@kevinaylward.co.uk> wrote:

I mean, why should I build it? What possible purpose could that
serve, that would not be known already?

---
Wiring has a nasty habit of upsetting perfectly loaded applecarts.


And it what way do the properties of "wiring" transcend the laws of
physics such that such "wiring" is unable to be inclued in
simulations?

I'm guessing that the output fets will RF oscillate big time.
Especially with those nice gate-gate caps downstream of the gate
resistors.

Yes...and er... as now doubt well known in this NG, by now, the mosfet1000
was designed in 1982. Just where do you guess, was the placement of the
mosfet gate resistors, such that the amp actually worked? Unfortunately, no
pack of Guinness for the correct answer.

Wiring certainly matters here.

Indeed, and what part of "And it what way do the properties of "wiring"
transcend the laws of physics such that such "wiring" is unable to be
inclued is simulations?" did you miss?

Kevin Aylward

www.kevinaylward.co.uk

 

[Kevin Aylward]

John Fields wrote:

On Mon, 20 Oct 2008 17:33:20 GMT, "Kevin Aylward"
kaExtractThis@kevinaylward.co.uk> wrote:

John Fields wrote:
On Sun, 19 Oct 2008 19:32:49 GMT, "Kevin Aylward"
kaExtractThis@kevinaylward.co.uk> wrote:

I mean, why should I build it? What possible purpose could that
serve, that would not be known already?

---
Wiring has a nasty habit of upsetting perfectly loaded applecarts.


And it what way do the properties of "wiring" transcend the laws of
physics such that such "wiring" is unable to be inclued in
simulations?

---
They don't transcend the laws of physics at all, as you very well
know, Mr. Snotty, but the physical and electrical properties of
FR-4, for instance, (thickness of the matrix, thickness of the
copper, variation in the dielectric constant and resistivity of the
matrix with temperature and humidity, and on and on) would all have
to be sorted out and included in the simulation properly.

But hardly relevant to audio amp design. Wiring inductance and resistance,
and a few pfs of strays is usually all that's necessary.

I think most people who design PCBs don't go to those lengths but,
rather, follow rules which will assure them a pretty good chance of
getting a good board the first time out.

How do you design PCBs?

I don't. I leave that to techs. Just give them a few pointers here and
there.

Kevin Aylward
www.kevinaylward.co.uk

 

[John Fields]

On Wed, 22 Oct 2008 18:46:34 GMT, "Kevin Aylward"
<kaExtractThis@kevinaylward.co.uk> wrote:

John Fields wrote:
On Mon, 20 Oct 2008 17:33:20 GMT, "Kevin Aylward"
kaExtractThis@kevinaylward.co.uk> wrote:

John Fields wrote:
On Sun, 19 Oct 2008 19:32:49 GMT, "Kevin Aylward"
kaExtractThis@kevinaylward.co.uk> wrote:

I mean, why should I build it? What possible purpose could that
serve, that would not be known already?

---
Wiring has a nasty habit of upsetting perfectly loaded applecarts.


And it what way do the properties of "wiring" transcend the laws of
physics such that such "wiring" is unable to be inclued in
simulations?

---
They don't transcend the laws of physics at all, as you very well
know, Mr. Snotty, but the physical and electrical properties of
FR-4, for instance, (thickness of the matrix, thickness of the
copper, variation in the dielectric constant and resistivity of the
matrix with temperature and humidity, and on and on) would all have
to be sorted out and included in the simulation properly.

But hardly relevant to audio amp design. Wiring inductance and resistance,
and a few pfs of strays is usually all that's necessary.

---
To throw everything off kilter?

Indeed, and paying attention to details is relevant when doing any kind
of design work.

Even audio, and a few pF of stray capacitance or some unintended
inductive coupling, unaccounted for, can easily result in unwanted
oscillations and result in the perfectly loaded applecart's being upset.
---

I think most people who design PCBs don't go to those lengths but,
rather, follow rules which will assure them a pretty good chance of
getting a good board the first time out.

How do you design PCBs?

I don't. I leave that to techs. Just give them a few pointers here and
there.

---
Then you agree that wiring has a nasty habit of upsetting perfectly
loaded applecarts and that your pointers will help to keep that from
happening?

JF

 

[Eeyore]

John Fields wrote:

Eeyore wrote:
John Fields wrote:
Eeyore wrote:
John Fields wrote:
"Kevin Aylward" wrote:

I mean, why should I build it? What possible purpose could that serve, that
would not be known already?

---
Wiring has a nasty habit of upsetting perfectly loaded applecarts.

Then the applecart wasn't designed right.

---
Whoosh...

The perfectly loaded applecart (properly designed circuit) can easily be
upset if the PCB layout (wiring) is improperly done.

That's *tracking* not 'wiring'.
---
Is that a Right-Pondian term?

Well do you call it 'tracing' as in traces ?

I've always heard of it referred to and have called it "wiring" as in:
"You got the wiring wrong on that PCB."

A WIRE is a single usually circular conductor or bundle of same covered by insulation.
I haven't seen many PCBs made of them. None in fact.

Do you guys say: "You got the tracking wrong on that PCB"?

Yes.

Sounds kind of ambiguous to me, like you're admonishing FEDEX or
somebody like that.

I take it, though, you take my point?

That you're being STUPID as usual.

The word these days is PCB *not* PWB.

Graham

 

[Eeyore]

John Fields wrote:

Then you agree that wiring

TRACKING !

has a nasty habit of upsetting perfectly
loaded applecarts and that your pointers will help to keep that from
happening?

 

[John Fields]

On Thu, 23 Oct 2008 03:10:26 +0100, Eeyore
<rabbitsfriendsandrelations@hotmail.com> wrote:

John Fields wrote:

Eeyore wrote:
John Fields wrote:
Eeyore wrote:
John Fields wrote:
"Kevin Aylward" wrote:

I mean, why should I build it? What possible purpose could that serve, that
would not be known already?

---
Wiring has a nasty habit of upsetting perfectly loaded applecarts.

Then the applecart wasn't designed right.

---
Whoosh...

The perfectly loaded applecart (properly designed circuit) can easily be
upset if the PCB layout (wiring) is improperly done.

That's *tracking* not 'wiring'.
---
Is that a Right-Pondian term?

Well do you call it 'tracing' as in traces ?

---
No, we call it wiring as in 'wiring'.

That is, when referring to the conglomeration of traces on the board in
general. When referring to single conductors or to a few, we generally
call them 'traces' or, less commonly, 'tracks'.
---

I've always heard of it referred to and have called it "wiring" as in:
"You got the wiring wrong on that PCB."

A WIRE is a single usually circular conductor or bundle of same covered by insulation.

I haven't seen many PCBs made of them. None in fact.


Do you guys say: "You got the tracking wrong on that PCB"?

Yes.

---
How you manage to mangle the language is beyond me.
---

Sounds kind of ambiguous to me, like you're admonishing FEDEX or
somebody like that.

I take it, though, you take my point?

That you're being STUPID as usual.

The word these days is PCB *not* PWB.

---
Huh???

_Me_ stupid?

At least I can read...

JF

 

[John Fields]

On Thu, 23 Oct 2008 03:11:36 +0100, Eeyore
<rabbitsfriendsandrelations@hotmail.com> wrote:

John Fields wrote:

Then you agree that wiring

TRACKING !

has a nasty habit of upsetting perfectly
loaded applecarts and that your pointers will help to keep that from
happening?

---
Do you consciously persist in your buffoonery, or is it involuntary?

My statement, to Kevin, was:

"Wiring has a nasty habit of upsetting perfectly loaded applecarts.",

not: "PCB wiring has a nasty habit of upsetting perfectly loaded
applecarts."

Can you understand the difference between the two?

Probably not, so I'll explain it to you:

The former case is general enough that it includes improperly shielded
conductors, conductors running too close to each other, and all manner
of problems caused by wiring, including PCB wiring.

The latter applies specifically to PBC's and is too restrictive to apply
to every situation, thereby being the reason I chose the more general
term.

Got it?

JF

 

[Eeyore]

John Fields wrote:

Eeyorewrote:
John Fields wrote:
Eeyore wrote:
John Fields wrote:
Eeyore wrote:
John Fields wrote:
"Kevin Aylward" wrote:

I mean, why should I build it? What possible purpose could that serve, that
would not be known already?

---
Wiring has a nasty habit of upsetting perfectly loaded applecarts.

Then the applecart wasn't designed right.

---
Whoosh...

The perfectly loaded applecart (properly designed circuit) can easily be
upset if the PCB layout (wiring) is improperly done.

That's *tracking* not 'wiring'.
---
Is that a Right-Pondian term?

Well do you call it 'tracing' as in traces ?

---
No, we call it wiring as in 'wiring'.

But it's not a wire.

That is, when referring to the conglomeration of traces on the board in
general. When referring to single conductors or to a few, we generally
call them 'traces' or, less commonly, 'tracks'.
---

I've always heard of it referred to and have called it "wiring" as in:
"You got the wiring wrong on that PCB."

A WIRE is a single usually circular conductor or bundle of same covered by insulation.

I haven't seen many PCBs made of them. None in fact.


Do you guys say: "You got the tracking wrong on that PCB"?

Yes.

---
How you manage to mangle the language is beyond me.

It's called ENGLISH not Americanlish.

Sounds kind of ambiguous to me, like you're admonishing FEDEX or
somebody like that.

I take it, though, you take my point?

That you're being STUPID as usual.

The word these days is PCB *not* PWB.

---
Huh???

_Me_ stupid?

VERY and totally out of touch. Did you play in ZZ Top btw ?

Graham