New webpage on loudspeaker cables
 

In article 4a8398e0.708789000@localhost, Don Pearce
wrote:
On Wed, 12 Aug 2009 09:56:54 +0100, Jim Lesurf
wrote:


Yes. I used to keep a small selection of capacitors with values ranging
from a few tens of pF up to a couple of microF to do quick checks on
amplifier stability, etc, when experimenting.

Perhaps worth adding that quite often there was a narrow range of
capacitance values that might upset a design. So, say, 2n2F might cause
oscillations, but 1nF or 4n7F didn't. So the problem often isn't that
'any value above X' causes oscillations.


I think capacitor resonance is important here. If an amp is potentially
unstable (usually with a capacitive load) at say 30MHz, a smallish cap
may provoke it. A big one, though, probably resonates below that
frequency, so at 30MHz it isn't a cap, but an inductor. The amp will be
quite happy with that.

Yes, that is one of the possible Reasons. Another is that it may not always
be the case that a larger capacitance may drag the HF gain too low for the
oscillation condition to be satisfied at HF, but not change the phase at
the point where the loop gain is unity enough to meet the condition.

This sort of brings up another point. It is easy enough to design an amp
that is unconditionally stable - the maths isn't hard. But this makes
some assumptions, one the big ones being that Cdom (usually about 100pF)
has the right phase shift. If the layout isn't great, though, it is
quite easy to end up with perhaps 100nH of parasitic inductance in
series with it. If that happens, Cdom turns into an inductor at 50MHz,
which may well be low enough to start things fizzing.

TBH I always tended to avoid the simple Cdom approach which causes people
to slug the design near the output with a large cap or two. I ended up
preferring using a resistor and cap in series in an earlier stage to drop
the HF gain to a resistor-defined level. Avoids adding to the phase delays
at HF and tends to swamp cap resonance. Means you can also use smaller
value caps that have higher self-resonance frequencies so shift that
problem to where it doesn't matter. I guess in some ways a bit like the
tweaks Bailey did for Radford at one time, but which were missed when
people focused excitement on him using a triode-pentode splitter. :-)

So it is not enough to design an amp which is theoretically
unconditionally stable, a parasitic analysis must be made to ensure that
this also applies to the reality of the physical design.

Again, yes. I also found that designs which use current sources, mirrors,
and long-tail pairs had improved stability as they tend to reject parasitic
paths via the power lines. Helps to keep down noise, distortion, and
crosstalk, and makes the operation more independent of the actual rail
voltages as well.

Slainte,

Jim

--
Please use the address on the audiomisc page if you wish to email me.
Electronics http://www.st-and.ac.uk/~www_pa/Scot...o/electron.htm
Armstrong Audio http://www.audiomisc.co.uk/Armstrong/armstrong.html
Audio Misc http://www.audiomisc.co.uk/index.html

New webpage on loudspeaker cables
 

Jim Lesurf wrote:
In article , mick
wrote:
On Tue, 11 Aug 2009 07:57:10 +0100, David Looser wrote:

"Eiron" wrote in message
...
I wasn't criticizing Jim. I was criticizing you.


Then you are being remarkably obscure. Criticising me for what?



I think his point is that 0.001uF / 1000pF (a figure that I, admittedly,
plucked from the air) across the output is less than some cable loads.
To quote you: "Arggh!... Plonking a 1nF capacitor across the output of
a feedback amplifier is a pretty likely way of making it unstable"

As Eiron points out, that's about 12m of Chord Silver Screen or a foot
of Townshend Isolda cable, which doesn't sound much does it?

Well, to me 12m seems quite long for domestic loudspeaker cables in a UK
context, although maybe not for the USA or elsewhere.

Don't forget the biwirers, especially those with wives who don't allow the
amp to be in the optimum position between the speakers.

--
Eiron.

New webpage on loudspeaker cables
 

David Looser wrote:
"mick" wrote in message
...
On Tue, 11 Aug 2009 07:57:10 +0100, David Looser wrote:

"Eiron" wrote in message
...
I wasn't criticizing Jim. I was criticizing you.


Then you are being remarkably obscure. Criticising me for what?


I think his point is that 0.001uF / 1000pF (a figure that I, admittedly,
plucked from the air) across the output is less than some cable loads. To
quote you:
"Arggh!... Plonking a 1nF capacitor across the output of a feedback
amplifier is a pretty likely way of making it unstable"

As Eiron points out, that's about 12m of Chord Silver Screen or a foot of
Townshend Isolda cable, which doesn't sound much does it?

Now, whether 1nF is enough to make any difference at all I'm not
experienced enough to know. If Jim says:
"if you just put a shunt capacitor across the output that might provoke
oscillations if the amp isn't already stable for other reasons" then
that's good enough for me. :-)

If you are right in your analysis of Eiron's point (and it sems likely) then
clearly he has little experience of this sort of thing. Yes indeed, a 1nF
capacitor connected directly across the output of an amplifier can easily
cause instability if that amplifier is not unconditionally stable.

But that's not what you wrote, is it?

--
Eiron.

New webpage on loudspeaker cables
 

In article , Eiron

wrote:
Jim Lesurf wrote:
In article , mick
wrote:


As Eiron points out, that's about 12m of Chord Silver Screen or a
foot of Townshend Isolda cable, which doesn't sound much does it?

Well, to me 12m seems quite long for domestic loudspeaker cables in a
UK context, although maybe not for the USA or elsewhere.

Don't forget the biwirers, especially those with wives who don't allow
the amp to be in the optimum position between the speakers.

Yes. that's a good point that hadn't occurred to me! From the POV of shunt
capacitance biwired 5m lengths will seem a lot like a conventional 10m run.
Have to have a ponder on that...

Slainte,

Jim

--
Please use the address on the audiomisc page if you wish to email me.
Electronics http://www.st-and.ac.uk/~www_pa/Scot...o/electron.htm
Armstrong Audio http://www.audiomisc.co.uk/Armstrong/armstrong.html
Audio Misc http://www.audiomisc.co.uk/index.html

New webpage on loudspeaker cables
 

"Eiron" wrote in message
...


Yes indeed, a 1nF capacitor connected directly across the output of an
amplifier can easily cause instability if that amplifier is not
unconditionally stable.

But that's not what you wrote, is it?

What I actually wrote was:

quote
Arggh!... Plonking a 1nF capacitor across the output of a feedback
amplifier is a pretty likely way of making it unstable
unquote

OK I didn't qualify that with the caveat about it not applying to
unconditionally stable amplifiers, otherwise I don't see the difference.
Don't forget this was in the context of the someone suggesting that
connecting a 1nF capacitor across the output would get rid of any RF (from
instability). I was just pointing out that adding such a capacitor could
provoke, rather than, remove, instability.

In any case if your objection to my earlier post was that I hadn't added a
caveat about it not applying to unconditionally stable amplifiers why didn't
you say so?, rather than posting some stuff about the capacitance of certain
cables?

Or would I be correct in assuming that you are simply casting around for
something, anything, to disagree with me about?

David.

--
Eiron.

New webpage on loudspeaker cables
 

Don Pearce wrote:

Eeyore wrote:
Jim Lesurf wrote:
Rob wrote:
Jim Lesurf wrote:

FWIW it means very little to me. You seem to assume a correlation
between frequency, resistance and sound.

Not sure what you mean, I'm afraid.

That much is apparent. Doubt you know much about the concept at all, nor the
bases of stability.

Can you explain then? It has me puzzled too. And given Jim's
background I am going to tell you he knows precisely what stability
means, and how to measure and predict it.

How many amps has Jim designed that have sold in the many tens of thousands ? And
been VERY reliable.


Stability circles are the second thing you learn about on the Smith
Chart.

And there's far more than one way to make an amp stable other than simple
dominant pole compensation. Just the output stage configuration and topology ( to
take yet another aspect ) can make a VAST difference. I actually got 'caught out'
once designing a new amp where the DC beta of the output devices was more than
adequate yet the load impedance would reflect back to the last voltage driver
stage, affecting phase margin.

Result, one extra emitter follower stage added and NO load would result in phase
related instability. The voltage driver stage's load was essentially constant as
a result of the change. It's not difficult really once you get the hang of it.

Graham


--
due to the hugely increased level of spam please make the obvious adjustment to
my email address

New webpage on loudspeaker cables
 

David Looser wrote:

"Eeyore" wrote

Not sure what you mean, I'm afraid.

That much is apparent. Doubt you know much about the concept at all, nor
the bases of stability.

I wonder why it is that some people on this NG post simply to be offensive
to others? Of *course* Jim knows the basis of stability, at least as well
as you do, probably a lot better.

Not if he thinks the answer belongs in the CABLE ! That's pure humbug.

JHC !

Graham

--
due to the hugely increased level of spam please make the obvious adjustment
to my email address

New webpage on loudspeaker cables
 

"Eeyore" wrote in message
...


David Looser wrote:

I wonder why it is that some people on this NG post simply to be
offensive
to others? Of *course* Jim knows the basis of stability, at least as
well
as you do, probably a lot better.

Not if he thinks the answer belongs in the CABLE ! That's pure humbug.


But then he doesn't, he was showing that different cables present different
loads, and will thus have different effects on potentially unstable
amplifiers.

So your point was?

David.

New webpage on loudspeaker cables
 

On Thu, 13 Aug 2009 00:26:31 +0100, Eeyore
wrote:



Don Pearce wrote:

Eeyore wrote:
Jim Lesurf wrote:
Rob wrote:
Jim Lesurf wrote:

FWIW it means very little to me. You seem to assume a correlation
between frequency, resistance and sound.

Not sure what you mean, I'm afraid.

That much is apparent. Doubt you know much about the concept at all, nor the
bases of stability.

Can you explain then? It has me puzzled too. And given Jim's
background I am going to tell you he knows precisely what stability
means, and how to measure and predict it.

How many amps has Jim designed that have sold in the many tens of thousands ? And
been VERY reliable.


I don't see any relevance in this answer.


Stability circles are the second thing you learn about on the Smith
Chart.

And there's far more than one way to make an amp stable other than simple
dominant pole compensation. Just the output stage configuration and topology ( to
take yet another aspect ) can make a VAST difference. I actually got 'caught out'
once designing a new amp where the DC beta of the output devices was more than
adequate yet the load impedance would reflect back to the last voltage driver
stage, affecting phase margin.


Really? If the load reflects back through to the voltage amplifier,
the load impedance of the VAs is inadequate. That can't happen if the
beta of the output devices is more than adequate. And of course when
that does happen, what you get is gross distortion more than phase
change.

Result, one extra emitter follower stage added and NO load would result in phase
related instability. The voltage driver stage's load was essentially constant as
a result of the change. It's not difficult really once you get the hang of it.


That is just taken for granted in a good design, and has no impact on
the loop stability of the amplifier, which is set by gain and phase
margin for the lower, easily controlled frequencies, and layout and
device parasitics for the higher frequencies.

d

New webpage on loudspeaker cables
 

In article , Eeyore
wrote:


David Looser wrote:

"Eeyore" wrote

Not sure what you mean, I'm afraid.

That much is apparent. Doubt you know much about the concept at all,
nor the bases of stability.

I wonder why it is that some people on this NG post simply to be
offensive to others? Of *course* Jim knows the basis of stability, at
least as well as you do, probably a lot better.

Not if he thinks the answer belongs in the CABLE ! That's pure humbug.

I agree. cf below. Although part of your confusion here is shown by the
ambiguities of the assertion in your "if" statement. :-)

Alas, you are still persisting in misunderstanding what I wrote and meant.
Still, others have clearly understood, so the good news is that you could
also do so, if you chose. :-)

By your own assertion and in your own terms, some amplifiers are "pure
****". So, if any of them are not unconditionally stable, then the load
presented to them could give rise to busts of oscillations, or perhaps
other problems. In such cases knowing about the load (which includes
impedance transformation by the cable) can matter to avoid problems. And
measuring the cables with some reference load values allows you to
determine the RCLG values that can affect the audio band response due to
interaction with source and loudspeaker impedance.

To return to the confusions in your assertion above.

Your assertion is based on some ambiguities.

Firstly: If something is asserted as an "answer" then you need to define
what the "question" was.

If it was, "Should all commercial amplifiers be unconditionally stable?"
Then my answer is "yes". Not the "answer" you hypothesise.

if it was "Should makers (and/or reviews) tell us which amplifiers are, or
are not, unconditionally stable?" Then again, my "answer" is "yes". Not the
one you hypothesise.

And if the question is "Someone uses an amplifier which may *not* be
unconditionally stable, what should they do?" Then my "answer" is more
complex as they can make a choice. Personally, I'd prefer to either check
and see if the amp is stable (either unconditionally or in the specific
circumstances of their use) and decide on that basis, or to simply replace
the amp with one that is established to be unconditionally stable. Not the
"answer" you hypothesise. However they might prefer to use a choice of one
cable that avoids oscillation over another that might provoke it if that
were possible in their case. I'd regard that as an unsatisfactory band-aid,
and not what I'd prefer myself. But if they were informed, it is their
choice if it is their money, etc.

But if the question is more like "Do users know that it is possible that
amplifiers can be unstable into some loads, and that cables can affect the
behaviour as a result of this?" Or perhaps, "Do users know that the cable
impedance can affect results even with unconditionally stable amplifiers?"
And we then ask "What should be done to let people assess this for
themselves?" Then my "answer" was the series of articles/webpages. Give
people measureable info and explanations of the physics and engneering and
let them use it if they wish. If they don't like it, or don't care, that is
their choice once information is available.

Secondly, your use of "belongs in":

This does largely depend on the earlier ambiguity. But personally, I would
regard it as much better to use an amplifier you are satisfied is stable
and has a low output impedance so you can be confident on that basis. That
makes more sense to me to being in a situation where you have to choose a
cable with an eye to avoiding the risk of instability, etc. So for me the
optimum "answer" would be to choose an amplifier which won't care what
cables you use.

But if a user has no test equipment, only buys on the basis of 'subjective
reviews' and has no information on which designs are unconditionally
stable, etc, then they may well find their choice of cable matters for
reasons which it *should not if their choice of amplifier was
unconditionally stable and had low output impedance*. We may not like
this situation, but it may occur.

*If* that arises in practice - and your assertion that some designs are
"pure ****" seems to indicate you believe it can, then for them avoiding
problems may involve choice of cables. Indeed, if they have already spent a
lot of cash on their (imperfect) amplifier then choosing a different cable
may be a lot cheaper and simpler than buying another amplifier. Howeve this
is *not* a possible situation I am happy with. Hence my interest in giving
people a 'heads up' on the matter.

So...

One of the purposes of the work on the webpage(s) is to assess the effects
of choice of cable could have on the RF loading and to warn people that -
with an amp that you might regard as "pure ****" - could lead pun to
problems.

The other is to show people the data upon which RCGL values can be based.
These in turn can be used to assess any changes in the audio band with a
given amp and speaker.

This is in the context of the previous pages/articles I did on this, and
allowed me to compare measured results on current cables with the theory
and the measurements of others (Jim Moir and Martin Colloms) some years
ago. The actual 'conclusions' I come to should - with some extra work on a
related topic - appear in due course.

So far as I can see, none of what I said is novel, unconventional, or
'rocket science' which only an illuminati can understand. Nor have I said
that we all have to run around like headless chickens or throw around abuse
or blame.

To me it is just boring old engineering and physics. Stuff that I, and I am
sure many other engineers, have understood for decades. But which some
users may not know, and perhaps some designers don't either if their
designs *are* "pure ****" as you assert. Nor does it imply anyone has to
buy fancy cables. Just know what they are doing (particularly reviewers and
designers). Indeed, one advantage of having some knowledge of this is that
it helps people to realise that expensive cables may not be needed, or may
even be a poorer choice.

I'm afraid that almost everything you've written in this thread indicates
you haven't actually understood the meaning and content of the webpage.
Your "if" statement above about what you presume I "think" is an example.
Som far as I can tell from unscrambling your ambiguities I do *not*
"think" what you presumed. Thus much of your argument is based on you
(presumably without realising) adopting a "straw man" approach to what I
wrote. You are attacking things you think I mean, but which I don't.

However it seems clear that others *have* understood what I meant. So you
can still do so if you wish. Hence I am writing this in the hope it may
clarify the situation for you. However having done so, if you prefer to
continue to react as you have thus far, then I am happy to decide I would
be wasting time trying again to explain this to you.

Slainte,

Jim

--
Please use the address on the audiomisc page if you wish to email me.
Electronics http://www.st-and.ac.uk/~www_pa/Scot...o/electron.htm
Armstrong Audio http://www.audiomisc.co.uk/Armstrong/armstrong.html
Audio Misc http://www.audiomisc.co.uk/index.html