In article , Rob
wrote:
Jim Lesurf wrote:
There's a relationship between three things: frequency and resistance
(the things you plot) and sound.
All the plots show relationships between frequency and apparent load
'resistance' (actually magnitude of impedance). But that isn't
directly related to 'sound' as such.
Ah fine, that was just my assumption relating to the point of the
article. I see now the point of your discussion is 'risk'. Or have I got
this wrong? There was no specific point - just a series of tests to see
what happens?
The article was to collect, present, and analyse a set of measurements to
help people (who can understand the data and analysis) to consider a number
of 'point'. There are two main ones.
One is the question of what may happen if your amp is not 'uncoditionaly
stable' or is otherwise affected by difficult RF loading. (The classic case
here being a resonant dip in impedance prompting oscillations at RF which
can then damage the amplifier or cause its audio performance to be
altered.)
The other is that the RF measurements provide the data to determine the
series and shunt impedance properties of the cables. This can affect the
audio band frequency response as a result of interaction with the amp
output impedance and/or speaker input impedance *in the audio band*.
Ah, OK, good. But is it *really* risky for any amplifier that doesn't
carry a cable recommendation tag? By risk I assume possibility of
component failure.
It is certainly possible for an amplifier to exhibit uncontrolled RF
oscillations, and for those to then damage the amplifier. Possibly
also the speaker. But I can't tell you any value for the 'risk' of
this happening as it would depend on things we don't know. More likely
is that the audio behaviour may be affected without the amp failing.
I'm sorry - lost here. You explained above that the relationships you
are examining are not related to sound 'as such'. Does 'as such' mean
'except when it is'? ;-)
In effect, yes. :-) *If* your amplifier is designed to be unconditionally
stable then any RF peaks and dips in impedance won't prompt RF
oscillations. That means no oscillations can then cause damage or cause the
audio behaviour to alter.
There is also a rarer chance that RF impedance effects will alter the
intermodulation distortion behaviour at audio. But with a low distortion
unconditionally stable amp, that really should not be a problem. If any
effect arises in such cases it might just mean a slight change in a tiny
value.
The problems would be if your amp design is 'poor'. Then instability or
other effects might occur. The difficulty here is that so far as I can
tell, no-one is even checking this for reviews. And as 'eeyore' wrote, some
amp designs might be '****'.
He stated some are. I don't know, but think it at least possible. I confess
I do wonder about some of the expensive 'valve amps' that show behaviour
like high output impedance (low damping factor) and which may have their
output transformer in the feedback loop, etc. They may well be fine. But
there seems to be no data.
I know you don't know everything about every amplifier, but could you
explain what the average punter should look for in an amplifier to avoid
these issues? The sorts of questions I can ask a manufacturer for
example?
The primary question is, "Is your power amplifier unconditionally stable?"
A "yes" response should go a long way to removing any risk or worry,
although there is still some chance they are wrong, or that there can be
some alteration in behaviour at a low level.
I know of course this is going completely left field, but could you give
real world examples (*an* amplifier*, *some* cable) when sound might be
affected? I suspect your interest is entirely theoretical, but and if I
may say you do appear reluctant to be drawn . . .
I can't do that with any current commercial designs as I haven't tested
them, and have found no data. However a Naim amp some decades ago, using
some 'Monitor Audio 8 Ohm' cable of the kind mentioned in the first HFN
cables article showed clear bursts of RF on audio waveforms. Also with
other cables. IIRC This was using a B&W loudspeaker of the period, but I
can't recall what actual speaker as it was too long ago.
if viewed with a scope with a bandwidth too low to show the RF bursts I
could see little 'kinks' in the audio waveforms, and the distortion rose.
Smaller audio waveforms didn't produce the effect. So here the effect
depended on the details of the waveform, load, and cable.
The problem is that there are so many variables that - unless the amp is
unconditionally stable, etc - you can't always predict what will happen.
Most amp-cable-speaker combinations may be fine, but every now and then one
isn't.
A *useful* theory is one that explains why something is happening. I
follow your theory development to a point, but I don't understand what
use your findings are (exception noted below, remove possibility of
problems) if they're not 'latched' on to real world scenarios.
The two issues a
1) To get people to focus on the need to ensure amps *are* unconditionally
stable* so as to avoid things like RF oscillations. (and to warn that low
inductance cables should perhaps be avoided if your amp is *not*
unconditionally stable.)
2) To note that the cable RCLG values can affect the audio response.
Neither of the above is 'new' to audio engineering. I and other designers
have known about this for decades. The snag is that for ages no-one seems
to have even checked stability, and I'm not certain *all* designers know
this and put it into practice. And I've also lost count of all the
'reviews' that ignore the way cable impedance (usually series impedance)
can alter the audio band frequency response. 'Reviews' sic of cables
often don't even bother to give any values for cable series resistance,
etc.
So if no-one tells you if amplifiers are unconditionally stable, and no-one
tells you the properties of the cables, how can you decide what is either
safe or an optimum choice? All I can do is give some examples and allow
people to think for themselves about the issues.
if you use very short cables none of this is likely to matter. But if you
use cables many metres long, it might. This is another variable that can
alter the results.
So the article is a combination of a 'heads up' warning, and a set of
measurements to put values to the properties of some cables. I deliberately
chose a variety of types of construction, and included Maplin cables as
well as expensive 'audiophile' ones. I also personally still happily use
Maplin loudspeaker cables. But I use short cable runs, and am sure my power
amp is stable, etc, since I designed it in the first place to be so!
Designed it nearly 30 years ago now, so none of this is new. :-)
Alas, it does look like Eeyore and perhaps one or two others have simply
reacted in horror to seeing plots of 'RF band measurements' in an article
and decided that "this must be snake oil" without the bother of actually
ensuring they first understood the purpose of the measurements and what
they can tell a reader who considers them rather than recoiling in horror.
But as I've said, that may be understandable if the reader either doesn't
realise how such measurements can be used to help determine the basic
electrical properties of the cable, or hasn't noticed that there
has been a lack of info on the two topics I have pointed out above.
One interesting effect of the "shock horror" reaction is that this then
means that whilst jumping to the erronious conclusion I am supporting
'snake oil' some of what I was actually saying gets missed. For example,
if you look at the results and comments for the thick multistranded
Maplin cable you can see that this shows a tendency for the series
resistance at RF to be higher than at audio.
The point here is that some might claim that an 'advantage' or
'distinction' of single core cables like the DNM is that they exploit
'skin effect' to obtain higher series resistance at RF - thus reducing
the scale of any RF peaks and dips. Yet the measurements indicate that
the same can happen with multistranded cables like the Maplin one with
loads of tiny strands.
The snag is that the reader has to actually read what I wrote and think
about it, though. Not just gut-react in horror and presume I'm trying to
push fancy cables. As I've said, I still cheerfully use the Maplin
cables and am quite happy with them. But the *measured results* show
specific bases for special case reasons for other choices *if* your
requirements and situation differ from mine. if you understand the
data you can make your own choices. The only drawbacks are the need
to actually consider the data, and to note where you need info that
reviews and manufacturers haven't provided. e.g. is your amplifier
unconditionally stable, and/or does it have a very high output
impedance?
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