Simple audio amp design...

[Cursitor Doom]

On 19/12/2021 20:59, whit3rd wrote:

On Sunday, December 19, 2021 at 10:03:19 AM UTC-8, Cursitor Doom wrote:
On Sun, 19 Dec 2021 09:44:17 -0800, jla...@highlandsniptechnology.com
wrote:
I dislike audio. Why not buy some germanium transistors from Ebay?
No point. The only parts available are NOS which will have whiskers
growing in them anyway, so they\'ll fail eventually guaranteed.

Germanium doesn\'t generally have tin-whisker problems (those are particular to
a case design, not the transistor inside); it has additional failure modes and
aging issues, but so does any technology

Indeed. It\'s the metal case that\'s the primary cause of the problem in
this device. IIRC I believe Ge transistors are also far more easily
damaged by soldering into place than Si. We used to use needle-nose
pliers on the leads close to the case as a heat sink to protect them.

Be aware, though, that Si replacements for Ge transistors does change all the bias
currents, I\'ve seen items burned up because of a naiive replace-with-modern-Si strategy.

Yes, I\'m well aware of that, hence the question about re-jigging the
design to allow for said difference.

 

[Ralph Mowery]

In article <spodeo$emp$1@dont-email.me>, cd@notformail.com says...

Indeed. It\'s the metal case that\'s the primary cause of the problem in
this device. IIRC I believe Ge transistors are also far more easily
damaged by soldering into place than Si. We used to use needle-nose
pliers on the leads close to the case as a heat sink to protect them.

I remember all the heat precautions from years ago. A while back I
started the SMD soldering and using a heat gun. Really had me worried
that playing all that hot air on the devices, but sofar in the few I
have replaced they seem to survived. Most of the SMD are smaller than
the tip of the soldering iron.

 

[Clive Arthur]

On 19/12/2021 20:59, whit3rd wrote:

<snipped>

Germanium doesn\'t generally have tin-whisker problems (those are particular to
a case design, not the transistor inside); it has additional failure modes and
aging issues, but so does any technology.

Yebbut, tin whiskers are the spawn of communism dressed up as
environmentalism, and Crusty Oldloon must rail against them.

--
Cheers
Clive

 

[Phil Allison]

Cursitor Doom wrote:
=================
> Greetings, gentlemen,

** Surely you jest?

I want to remain faithful to the original,
simple design ethos with its five discrete transistors whatever I do.
Any suggestions?

Here\'s the schematic:

https://disk.yandex.com/i/a2xX6_IadjbgIw

** Looks fairly easy:

1. The BC108 can stay
2. Change T2 to a BC178 or similar TO18
3. Change T3 to 2N2904 or similar TO5
3. Change R3 to 820 ohms.
4. Change T4 to TIP31
5. Change T5 to TIP32

Step 2 ( Vbe multiplier) solves the bias issue for the new, Si outputs.


....... Phil

 

[Anthony William Sloman]

On Sunday, December 19, 2021 at 11:40:23 PM UTC+11, Cursitor Doom wrote:

Greetings, gentlemen,

I have a vintage MW/LW/SW 12V battery-powered radio I picked up recently
at a yard sale. Being 50 years old, however, it has a few faults and
needs realignment and whatnot. Before I can proceed with that, I\'ve
discovered that the transistors it uses (mostly obsolete germanium
devices) suffer from tin whiskers so I will need to re-jig the board so
they can be replaced with silicon ones instead. The design is very
simple so it *shouldn\'t* be a big deal. The bias requirements are of
course different between germaniums and silicons so perhaps it would be
better to start from scratch? I want to remain faithful to the original,
simple design ethos with its five discrete transistors whatever I do.
Any suggestions?

Here\'s the schematic:

https://disk.yandex.com/i/a2xX6_IadjbgIw

This is sci.electronics.design. It\'s about designing electronics, not a place where you ask for advice about trivially simple class A/B audio amplifiers which have been around since transistors were first invented.

https://en.wikipedia.org/wiki/Power_amplifier_classes

There are some interesting ways of biasing the amplifier to get an adequate minimum standing current to minimise cross-over distortion without cooking the output transistors, but none of them are antiquated enough to be of interest to Cursitor Doom, nor popular enough for me to find a wikipedia reference.

Douglas Self and other audio guru\'s have spelled it out

http://www.douglas-self.com/ampins/dipa/dipa.htm

You have to scroll through to part 6 - \"the blameless amplifier\" where figure 33 shows an example (wrapped around Tr13).

-- Bill Sloman, Sydney

 

[Phil Allison]

IEEE Bill the FAKE bill....@ieee.org wrote:
===================================

The bias requirements are of
course different between germaniums and silicons so perhaps it would be
better to start from scratch? I want to remain faithful to the original,
simple design ethos with its five discrete transistors whatever I do.
Any suggestions?

Here\'s the schematic:

https://disk.yandex.com/i/a2xX6_IadjbgIw

This is sci.electronics.design. It\'s about designing electronics,

** 100% WRONG !!!!!

Been correctly argued already that the NG name is ambiguous.
So it MUST include examination, analysis and all questions about existing designs..

> not a place where you ask for advice about trivially simple class A/B audio amplifiers

** ROTFL - ain\'t no such animal.


> Douglas Self and other audio guru\'s have spelled it out

** Doug is part \" self appointed \" guru and part bullshitter.

His published mistakes are numerous and NEVER admitted.
Just like YOURS !!


...... Phil

 

[Anthony William Sloman]

On Monday, December 20, 2021 at 11:49:36 AM UTC+11, palli...@gmail.com wrote:

IEEE Bill the FAKE bill....@ieee.org wrote:
==================================
The bias requirements are of
course different between germaniums and silicons so perhaps it would be
better to start from scratch? I want to remain faithful to the original,
simple design ethos with its five discrete transistors whatever I do.
Any suggestions?

Here\'s the schematic:

https://disk.yandex.com/i/a2xX6_IadjbgIw

This is sci.electronics.design. It\'s about designing electronics,
** 100% WRONG !!!!!

Been correctly argued already that the NG name is ambiguous.
So it MUST include examination, analysis and all questions about existing designs..

Or Phil thinks it must. Cursitor Doom\'s aim to stick with just five transistors is a constraint on design - but it isn\'t a rational one.

not a place where you ask for advice about trivially simple class A/B audio amplifiers.

** ROTFL - ain\'t no such animal.

I wonder what Phil meant by that? He probably means he finds the design of class A/B amplifiers complicated, which it isn\'t.

Douglas Self and other audio guru\'s have spelled it out.
http://www.douglas-self.com/ampins/dipa/dipa.htm
You have to scroll through to part 6 - \"the blameless amplifier\" where figure 33 shows an example (wrapped around Tr13).

** Doug is part \" self appointed \" guru and part bullshitter.

According to Phil, who aspires to the same kind of following, but isn\'t good enough at being \"impressively expert\" to make money out of it.

> His published mistakes are numerous and NEVER admitted.

Mostly because what Phil thinks are \"mistakes\" are what other people count as differences of opinion.

> Just like YOURS !!

I do make mistakes from time to time - rather fewer than Phil likes to think - and there are posts (mostly including the word \"oops\") where I admit it..

There\'s even one where I eventually agreed that Phil was probably right - Cursitor Doom and Eeyore are almost certainly different people.

Phil will have forgotten that, in the same way that he seems to forget his own pratfalls. Phil snipped the link to what Douglas Self had had to say on the subject - and skipped the opportunity to tell us what was wrong with it. I wasn\'t all that impressed with output stage of that amplifier - I don\'t find npn and pnp Darlington as nice as complementary Darlngtons - Sziklai pairs - which are what I used in my home built version (back around 1978)..

https://en.wikipedia.org/wiki/Sziklai_pair

--
Bill Sloman, Sydney

 

[whit3rd]

On Sunday, December 19, 2021 at 1:51:21 PM UTC-8, jla...@highlandsniptechnology.com wrote:

On Sun, 19 Dec 2021 12:44:35 -0800 (PST), Phil Allison
palli...@gmail.com> wrote:

On Monday, December 20, 2021 at 3:17:22 AM UTC+11, jla...@highlandsniptechnology.com wrote:

Why are some people so adverse to drawing a simple straight 4-wire
connection? It\'s not as if dots slide off the page or anything.

Think the 45 degree angle convention is exactly because dots can disappear when making copies.

Cheap British dots do fall off the paper. We use macho USA dots.

It\'s mainly a biology-lab problem, that. The interaction of formalin preservative with
the laser-printed labels in the bottle next to the specimen... was to float
the letters off the paper, making a little stack of vowels and consonants.

 

[Rich S]

On Sunday, December 19, 2021 at 12:40:23 PM UTC, Cursitor Doom wrote:

Greetings, gentlemen,

I have a vintage MW/LW/SW 12V battery-powered radio I picked up recently
at a yard sale. Being 50 years old, however, it has a few faults and
needs realignment and whatnot. Before I can proceed with that, I\'ve
discovered that the transistors it uses (mostly obsolete germanium
devices) suffer from tin whiskers so I will need to re-jig the board so
they can be replaced with silicon ones instead. The design is very
simple so it *shouldn\'t* be a big deal. The bias requirements are of
course different between germaniums and silicons so perhaps it would be
better to start from scratch? I want to remain faithful to the original,
simple design ethos with its five discrete transistors whatever I do.
Any suggestions?

Here\'s the schematic:

https://disk.yandex.com/i/a2xX6_IadjbgIw

CD.

well if youre hard-core restorer & remain faithful, and if you cant
get original parts, then you take the old part, carve out its insides
and sneak a new modern part inside. May have to include a
resistor or two, to emulate the old part.

 

[Rich S]

This is sci.electronics.design. It\'s about designing electronics, not a place where you ask for advice about trivially simple class A/B audio amplifiers which have been around since transistors were first invented.

True, CD\'s post is not about \"new design\", its really about restoration
or rehabilitation of an old one. (And he put \"Simple\" in the title, even
though \"simple\" is subjective, as Phil implied).
I do find this area interesting, and suggest it be one of our allowed
topics for discussion.
Practically speaking, we need more good topics. We all want
this usenet group to live, & well, I assume.
There just doesn\'t seem to be enough \"new design\" posts to
keep this group that narrowly defined. So we must evolve.

 

[bitrex]

On 12/20/2021 3:18 PM, Rich S wrote:

This is sci.electronics.design. It\'s about designing electronics, not a place where you ask for advice about trivially simple class A/B audio amplifiers which have been around since transistors were first invented.


True, CD\'s post is not about \"new design\", its really about restoration
or rehabilitation of an old one. (And he put \"Simple\" in the title, even
though \"simple\" is subjective, as Phil implied).
I do find this area interesting, and suggest it be one of our allowed
topics for discussion.
Practically speaking, we need more good topics. We all want
this usenet group to live, & well, I assume.
There just doesn\'t seem to be enough \"new design\" posts to
keep this group that narrowly defined. So we must evolve.

The circuit isn\'t that simple anyway, check out that compensation
network. How the whole amp operates in principle is easy but figuring
out the values so it\'s stable and doesn\'t suck is harder, just
dominant-pole compensating and calling it a day isn\'t an option I don\'t
think, its low-frequency open-loop gain isn\'t huge to begin with.

 

[Rich S]

Is a Vbe multiplier a strange thing to others here?


...... Phil

well not to me, but to many, it could.

My point to everyone is, you don\'t really know
how much our audience knows or doesn\'t know.
There\'s no \"entrance exam\" to joining this group ;-)
I assume for every 1 (semi-)regular poster
there are 100 (semi-)lurkers. And many may
*not* know about a vbe-multipler circuit or
other more esoteric things we delve into.

So I dont my mind, being a little kind, and
welcome unsophisticated questions & responses.
When Man meets The Internet, all sorts of
the sublime & the crazy will ensue. We do
pretty well, here, considering.

 

[Phil Allison]

Rich S wrote:
==================

Is a Vbe multiplier a strange thing to others here?


..
well not to me, but to many, it could.

** OK - so what is the one in that circuit doing exactly ?

How come just swapping the Ge part to Si do the trick ?


...... Phil

 

[Rich S]

On Monday, December 20, 2021 at 9:33:26 PM UTC, palli...@gmail.com wrote:

Rich S wrote:
==================

Is a Vbe multiplier a strange thing to others here?


..
well not to me, but to many, it could.
** OK - so what is the one in that circuit doing exactly ?

How come just swapping the Ge part to Si do the trick ?


..... Phil

(Uh oh, a Phil challenge, do I risk it?)
I assume, doing the Ge > Si swap, on BOTH
the output stage and the \"rubber diode\"*
the proportions setup by the resistors
can remain as is.

*I never heard of this term until now, I just
found the Wikipedia entry:
https://en.wikipedia.org/wiki/Rubber_diode
and it has your mate, Rod, as a
reference.

cheers, RS

 

[Lasse Langwadt Christensen]

mandag den 20. december 2021 kl. 22.33.26 UTC+1 skrev palli...@gmail.com:

Rich S wrote:
==================

Is a Vbe multiplier a strange thing to others here?


..
well not to me, but to many, it could.
** OK - so what is the one in that circuit doing exactly ?

How come just swapping the Ge part to Si do the trick ?

roughly 2*Vbe biasing two Vbe, as long as they are all the same type it adds up

 

[Phil Allison]

lang...@fonz.dk wrote:
==============

Is a Vbe multiplier a strange thing to others here?


..
well not to me, but to many, it could.
** OK - so what is the one in that circuit doing exactly ?

How come just swapping the Ge part to Si do the trick ?

roughly 2*Vbe biasing two Vbe, as long as they are all the same type it adds up

** Give the man a Kewpie doll.


...... Phil

 

[Anthony William Sloman]

On Wednesday, December 22, 2021 at 11:33:41 AM UTC+11, Cursitor Doom wrote:

On 20/12/2021 20:18, Rich S wrote:

This is sci.electronics.design. It\'s about designing electronics, not a place where you ask for advice about trivially simple class A/B audio amplifiers which have been around since transistors were first invented.


True, CD\'s post is not about \"new design\", its really about restoration
or rehabilitation of an old one. (And he put \"Simple\" in the title, even
though \"simple\" is subjective, as Phil implied).
I do find this area interesting, and suggest it be one of our allowed
topics for discussion.
Practically speaking, we need more good topics. We all want
this usenet group to live, & well, I assume.
There just doesn\'t seem to be enough \"new design\" posts to
keep this group that narrowly defined. So we must evolve.

I couldn\'t agree more.

And your opinion couldn\'t be less relevant. Your \"contributions\" are entirely negative - mostly links to \"Russia Today\" and \"Zero Hedge\" disguised as tinyurls.

Too many \'designs\' nowadays are just people modifying existing designs
or interconnecting ICs to do the heavy lifting.

If that solves the problem that needs to be solved, where\'s the problem?

> Both approaches are cop-outs in my view.

How can they be \"cop-outs\"? Design is all about putting something together to solve a problem - sometimes you want the cheapest possible solution, sometimes the one that takes up the least space, and often the one that can be completed as quickly as possible. These are all perfectly legitimate constraints on the design process.

> Sure they get the job done quick, but where\'s the satisfaction in that?

Design isn\'t about leaving the designer feeling satisified. It\'s about leaving the end user feeling satisfied.

> This was a chance for folk here to come up with something original - using discrete components.

Why on earth would anybody bother?

> I\'d have thought the modest challenge of that; hardly insurmountable by any standard, would have appealed at some back-to-basics level.

But who would put in the effort to make Cursitor Doom - of all people - happier.

> But maybe experienced designers here are lost without recourse to their usual go-to ICs. :-/

And maybe Cursitor Doom doesn\'t know what he is talking about.

--
Bill Sloman, Sydney

 

[Jan Panteltje]

On a sunny day (Wed, 22 Dec 2021 00:33:33 +0000) it happened Cursitor Doom
<cd@notformail.com> wrote in <sptrou$nrt$1@dont-email.me>:

On 20/12/2021 20:18, Rich S wrote:

This is sci.electronics.design. It\'s about designing electronics, not a place where you ask for advice about trivially simple
class A/B audio amplifiers which have been around since transistors were first invented.


True, CD\'s post is not about \"new design\", its really about restoration
or rehabilitation of an old one. (And he put \"Simple\" in the title, even
though \"simple\" is subjective, as Phil implied).
I do find this area interesting, and suggest it be one of our allowed
topics for discussion.
Practically speaking, we need more good topics. We all want
this usenet group to live, & well, I assume.
There just doesn\'t seem to be enough \"new design\" posts to
keep this group that narrowly defined. So we must evolve.

I couldn\'t agree more.
Too many \'designs\' nowadays are just people modifying existing designs
or interconnecting ICs to do the heavy lifting. Both approaches are
cop-outs in my view. Sure they get the job done quick, but where\'s the
satisfaction in that? This was a chance for folk here to come up with
something original using discrete components. I\'d have thought the
modest challenge of that; hardly insurmountable by any standard, would
have appealed at some back-to-basics level. But maybe experienced
designers here are lost without recourse to their usual go-to ICs. :-/

There is some truth in that, but in today\'s world you could say build an AM FM radio
with some old transistors, or new ones, but if you go HD TV the codecs and other stuff
need highly integrated chips.
You could code that in an FPGA (for example a mpeg2 decoder) to show YOU can do it too,
but then you are already using existing chips.

But to have some transistors to be ably to send an SOS in case you ran out of solder, why not
Point is I have done all that transistors design for all sort of things, and tubes too before that,
And fixed thousands of circuits, some simple, some very complicated, in broadcasting
and in my repair shop.
I can tell you that (having to look) looking at so many different circuits from so many designers
is a great learning experience.
In my opinion coming out of some EE school and just connecting IC_type1 to I_type2 is a different art perhaps.
It is what it does and how it works that counts,
It would take many years and hands on experience to get to know that,

All electronics designs, be it a simple audio amp or switcher etc. are interesting, there is a lot you can learn
or should have learned ;-)

At least it is on topic, unlike what the slow man just started,

 

[Anthony William Sloman]

On Wednesday, December 22, 2021 at 7:23:52 PM UTC+11, Jan Panteltje wrote:

On a sunny day (Wed, 22 Dec 2021 00:33:33 +0000) it happened Cursitor Doom
c...@notformail.com> wrote in <sptrou$nrt$1...@dont-email.me>:
On 20/12/2021 20:18, Rich S wrote:

This is sci.electronics.design. It\'s about designing electronics, not a place where you ask for advice about trivially simple
class A/B audio amplifiers which have been around since transistors were first invented.


True, CD\'s post is not about \"new design\", its really about restoration
or rehabilitation of an old one. (And he put \"Simple\" in the title, even
though \"simple\" is subjective, as Phil implied).
I do find this area interesting, and suggest it be one of our allowed
topics for discussion.
Practically speaking, we need more good topics. We all want
this usenet group to live, & well, I assume.
There just doesn\'t seem to be enough \"new design\" posts to
keep this group that narrowly defined. So we must evolve.

I couldn\'t agree more.
Too many \'designs\' nowadays are just people modifying existing designs
or interconnecting ICs to do the heavy lifting. Both approaches are
cop-outs in my view. Sure they get the job done quick, but where\'s the
satisfaction in that? This was a chance for folk here to come up with
something original using discrete components. I\'d have thought the
modest challenge of that; hardly insurmountable by any standard, would
have appealed at some back-to-basics level. But maybe experienced
designers here are lost without recourse to their usual go-to ICs. :-/
There is some truth in that, but in today\'s world you could say build an AM FM radio
with some old transistors, or new ones, but if you go HD TV the codecs and other stuff
need highly integrated chips.
You could code that in an FPGA (for example a mpeg2 decoder) to show YOU can do it too,
but then you are already using existing chips.

But to have some transistors to be ably to send an SOS in case you ran out of solder, why not
Point is I have done all that transistors design for all sort of things, and tubes too before that,
And fixed thousands of circuits, some simple, some very complicated, in broadcasting
and in my repair shop.
I can tell you that (having to look) looking at so many different circuits from so many designers
is a great learning experience.
In my opinion coming out of some EE school and just connecting IC_type1 to IC_type2 is a different art perhaps.

I don\'t think that electronic- engineering courses prescribe designing only using integrated circuits - ICs just provide pre-connected transistors, that happen to work in situations that designers have to cope with quite often.

> It is what it does and how it works that counts,

True.

> It would take many years and hands on experience to get to know that,

Application notes can help, for particular applications. Some people can learn quite quickly from a limited amount of experience - others latch onto the NE555 and don\'t seem to want to learn how to use anything else,.

> All electronics designs, be it a simple audio amp or switcher etc. are interesting, there is a lot you can learn or should have learned ;-)

One of the things you learn is that some people are rotten at electronic design.

> At least it is on topic, unlike what sloman just started,

Considering the nut-cases who post here, both those papers from PNAS are decidedly on-topic for the group, though divorced from the electronic design that we are supposed to be posting about (and hardly ever do). Jan Panteltje doesn\'t strike me as one of our saner posters, so maybe he feels that he is being got at.

--
Bill Sloman, Sydney

 

[Tom Gardner]

On 22/12/21 00:33, Cursitor Doom wrote:

On 20/12/2021 20:18, Rich S wrote:

This is sci.electronics.design. It\'s about designing electronics, not a place
where you ask for advice about trivially simple class A/B audio amplifiers
which have been around since transistors were first invented.


True, CD\'s post is not about \"new design\", its really about restoration
or rehabilitation of an old one. (And he put \"Simple\" in the title, even
though \"simple\" is subjective, as Phil implied).
I do find this area interesting, and suggest it be one of our allowed
topics for discussion.
Practically speaking, we need more good topics. We all want
this usenet group to live, & well, I assume.
There just doesn\'t seem to be enough \"new design\" posts to
keep this group that narrowly defined. So we must evolve.

I couldn\'t agree more.
Too many \'designs\' nowadays are just people modifying existing designs or
interconnecting ICs to do the heavy lifting. Both approaches are cop-outs in my
view. Sure they get the job done quick, but where\'s the satisfaction in that?
This was a chance for folk here to come up with something original using
discrete components. I\'d have thought the modest challenge of that; hardly
insurmountable by any standard, would have appealed at some back-to-basics
level. But maybe experienced designers here are lost without recourse to their
usual go-to ICs. :-/

While I have some sympathy for the sentiment, I don\'t agree
with the reasoning and conclusion. OTOH if it is solely for
your sense of achievement, then that\'s a justifiable reason.

But why stop at discrete semiconductors? Why not go a little
further and make your own inductors, capacitors and valves.

My father has a book describing how to make your own
car - starting with how to make the tools necessary to make
the car.