In article , John
wrote:
On Thu, 08 Oct 2009 13:38:08 +0100, Jim Lesurf wrote:


Then you may have a problem to worry about. :-) Many of the gain
devices in audio power (and pre) amps have gain at frequencies
reaching up into that region. As a result *unless* the designer/maker
has ensured unconditional stability and no other problems affected by
RF loading, then changing the cables can affect the amplifier
behaviour.

Yeah but bandwith can be tailored.

Indeed. Alas you may be confusing theory with reality, and presuming that
all items of equipment *are* ideally well designed. :-) You are also
perhaps confusing two other things. cf below.

Medical instrumentation amplifiers are an example where attention is
closely paid to when designing equipment. You can fix the bandwith in
any audio amplifier, you can even have software filters with monstrously
steep drop offs/skirts for digital equipment. If an audio amplifier is
comprised of a number of carefully designed stages no single stage
having a very large gain and designed with a roll off beginning at about
25kHz then I don't see why any appreciable rf currents should come into
it.

I appreciate that you may not see why. :-) I fear that may also be the
case for some designers and makers. Ideally, it won't. But the snag is that
reality may not accord with your (or their) presumptions. As I have
personally witnessed in reality on more than one occasion over the years.

Yes, you can design and build amplifiers that are unconditionally stable
and whose audio performance isn't noticably affected by ultrasonic/RF
loading.

And, yes, there are real amplifiers for which the above simply isn't true.
They can and do oscillate, and this can and does affect performance with
audio signals. They can also exhibit behaviours like distortion levels that
vary with RF loading. No need for you to deliberately inject RF for this to
occur. This is a matter of having observed the measured results.


If you wish to avoid the 'madness' then you have two choices...

1) never buy or use any audio amplifiers.

No, it is possible to use an audio amplifier which will not appreciably
boost rf signals.

Yes, it is indeed possible. But the universe does not guarantee that every
design you might use *will* accord with this possibility.

It is possible to take home a box of 6 brown eggs. But that does not
guarantee that *all* boxes of eggs only contain brown ones *when the box
does not tell you they are all brown*. Indeed, even some which claim to
contain brown eggs may not be correctly labelled. The snag here is that
many audio amplifiers don't tell you if they meet the criterion of being
uncoditionally stable or not.


or

2) only buy ones where the maker/sellers/reviewers have told you the
unit is unconditionally stable and that behaviour is unaffected by RF
loading.

I noticed in the 70's how the fashion for extended frequency response in
audio amplifiers was creeping up and up to 200kHz and more and am not
surprised that the uneducated now think higher is better and who now may
be wowed by amps with a high gain response up to 30 mHz or more. It's a
foolish and pointless pursuit.

I am afraid you are confusing having extended gain response of the overall
design with the design being able to oscillate due to poor design and
loading. The two things are not the same. Indeed, tailoring the intended
frequency response - if done poorly - can *increase* the chance of
instability problems. Or just move the frequency of oscillation.

So having a presented audio response that rolls away from about 20 kHz does
not guarantee the amplifier can't misbehave at much higher frequencies into
some loads. The presented response into the rated load does not tell you
about this, I'm afraid.

[snip]

Yes but I suspect you are talking about exceedingly poorly designed
circuits.

Yes, I am. The snag is that they do exist, and have been sold to the
public. I have seen them in the past, and the responses I have had to my
measurements do strengthen my concern that we can't be sure that all modern
designs are immune to the same basic physics involved.

The problem here is that makers and reviewers have largely stopped even
mentioning if a model/design is unconditionally stable or not. So to tell
we would now have to measure the presented gain *and* output impedance *of
every design being sold* over the region where the gain of the unit is
above unity, and measure its complex output impedance across that entire
range, then check if there is any load R+jI (where R= and I can be in the
range from + to minus infinity) that would satisfy the Barkhausen
criterion. if you find a solution the amp is not unconditionally stable.

Also note that 'gain above unity' for a power amp does not just mean
voltage gain. It may mean power gain, which can be a very different matter.

Even that isn't enough to rule out problems like an alteration in
distortion performance short of oscillation.


I have used and tested premium op-amps when designing instrumentation
for medical research and didn't have any difficulty with rf springing up
anywhere along the line. A single high gain block of course is asking
for trouble.

I have spent some years designing and testing audio amplifiers -
commercially, for research, and for personal use. Plus various other uses
of analog amps, signal conditioning, etc, for research and measurement. So
my comments are based on practical experience.

The problem here is partly that domestic hifi amps have to deliver high
powers into undefined loads. The designer has no real idea what the user
will connect to the amplifier - particulary at RF. Thus this situation
isn't like the much better defined one with lab instrumentation. When I've
designed kit for lab use, etc, I could define what the loading and
conditions of use would be. That makes establishing what is OK vastly
easier. You can't do that with domestic audio power amps beyond a vague
range of magnitude. And how many speaker designs have you seen complex
impedance plots for that extend much above 20kHz? Remember that we may
need to know to well above 1MHz in this case.

You get what you pay for.

Always? I'd say you 'pay for what you get' may be more common than 'you get
what you pay for'. :-) Paying a lot of money may sometimes be buying a
fancy designer label and a kewl look to impress the mates. :-)

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