Have we all stopped arguing? May I ask a question?
What's the best way to check if a power amp is likely to be stable
into a capacitive load such as an ESL?
I could put a 2uF cap on the output and look at the output on a scope
while feeding it with a square wave.
But is it likely to damage anything if the amp fails the test?

One old Rotel amp has a warning in the manual against capacitive loads,
which is a pity as it is a nice, low-power amp and not likely to make
sparks if the kids turn it up to 11.

--
Eiron.

"Eiron"

Have we all stopped arguing? May I ask a question?
What's the best way to check if a power amp is likely to be stable
into a capacitive load such as an ESL?

** ESLs are not capacitive loads - that is an absurd myth.


I could put a 2uF cap on the output and look at the output on a scope
while feeding it with a square wave.

** Creates a load condition that does not exist with any hi-fi speaker.

Totally stupid.


One old Rotel amp has a warning in the manual against capacitive loads,

** Can you quote the actual wording ?

Some amplifiers are not suitable for direct connection to * transformer
input * speakers like the Quad ESL57 - mostly because they lack sub-sonic
filters and current limiting OR because they are simply too powerful.

If you have a combination of speaker and amplifier in mind, then post the
info.



..... Phil

In article , Eiron
wrote:
Have we all stopped arguing? May I ask a question? What's the best way
to check if a power amp is likely to be stable into a capacitive load
such as an ESL? I could put a 2uF cap on the output and look at the
output on a scope while feeding it with a square wave. But is it likely
to damage anything if the amp fails the test?

The old test used to be something like a 2.2uF cap. But TBH you would be
better experimenting by trying various capacitances, and some inductances.
I've know amps that were stable with 2.2uF but oscillated with 10,000pF.

What happens will depend entirely on the amp design. So unless you have
other info the only way to know is to experiment.

If in doubt you can try using a small fuse in the output. Or, better, to
put smaller dc fuses inside on the rails. They may blow needlessly, but
might also save the amp if it would otherwise damage itself. However if it
is OK, you may then need to re-test without the output fuse as that can
alter the loading it saw!

One old Rotel amp has a warning in the manual against capacitive loads,
which is a pity as it is a nice, low-power amp and not likely to make
sparks if the kids turn it up to 11.

If in doubt, put a low-value resistor in series with each output. Say 1.5
Ohms. Alternatively, if you know the amp lacks an output network, add one.
These changes may alter the frequency response, but will help avoid
instability!

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

"Eiron" wrote in message


Have we all stopped arguing? May I ask a question?
What's the best way to check if a power amp is likely to
be stable into a capacitive load such as an ESL?

Hang capacitors across its output.

However, presuming that an ESL is necessarily an unusually capacitive load
is not always correct.

I could put a 2uF cap on the output and look at the
output on a scope while feeding it with a square wave.
But is it likely to damage anything if the amp fails the
test?

If there is damage at all, it is likely to damage the amp, and also the
capacitor if it is too frail. Signfiicant currents can pass through the
capacitor.

IME if an amp is going to be actually unstable, the worst case capacitance
is likely to be far less than 2 uF. 0.01 uF +/- a factor of 10 would be my
recommended values for fishing for instability.

Remember, there are two ways to mess up an amp with capacitive loads. One
involves frying the output transistors if they lack adequate SOA and SAO
limiting. The other involves actually instability. Furthermore, both
inductive and capacitive loads can cause problems.


One old Rotel amp has a warning in the manual against
capacitive loads, which is a pity as it is a nice,
low-power amp and not likely to make sparks if the kids
turn it up to 11.

IME, most modern amps are both stable and durable.

"Jim Lesurf"
Eiron

Have we all stopped arguing? May I ask a question? What's the best way
to check if a power amp is likely to be stable into a capacitive load
such as an ESL? I could put a 2uF cap on the output and look at the
output on a scope while feeding it with a square wave. But is it likely
to damage anything if the amp fails the test?

The old test used to be something like a 2.2uF cap.


** Never was and still is NOT a relevant test of an amp's ability to cope
safety with an ESL speaker like the Quad '57.


But TBH you would be
better experimenting by trying various capacitances,

** Absolute CRAP !!


What happens will depend entirely on the amp design.


** And has nothing to do with Eiron's dopey question - ****HEAD



..... Phil

"Arny Krueger
"Eiron"

Have we all stopped arguing? May I ask a question?
What's the best way to check if a power amp is likely to
be stable into a capacitive load such as an ESL?

Hang capacitors across its output.


** ABSOLUTE BULL**** !!!!!!!!!!!!!


However, presuming that an ESL is necessarily an unusually capacitive load
is not always correct.

** No ESL ever made acts like a pure capacitor.


IME if an amp is going to be actually unstable, the worst case capacitance
is likely to be far less than 2 uF. 0.01 uF +/- a factor of 10 would be
my recommended values for fishing for instability.


** Says Arny the Compewter Geek while fishing for RED HERRINGS !!!!!!!!!!


Remember, there are two ways to mess up an amp with capacitive loads. One
involves frying the output transistors if they lack adequate SOA and SAO
limiting. The other involves actually instability. Furthermore, both
inductive and capacitive loads can cause problems.

** Says Arny the Compewter Geek while fishing for RED HERRINGS !!!!!!!!!!


One old Rotel amp has a warning in the manual against
capacitive loads, which is a pity as it is a nice,
low-power amp and not likely to make sparks if the kids
turn it up to 11.

IME, most modern amps are both stable and durable.


** But Arny has no actual experience.




...... Phil

Phil Allison wrote:
"Eiron"
Have we all stopped arguing? May I ask a question?
What's the best way to check if a power amp is likely to be stable
into a capacitive load such as an ESL?

** ESLs are not capacitive loads - that is an absurd myth.


I could put a 2uF cap on the output and look at the output on a scope
while feeding it with a square wave.

** Creates a load condition that does not exist with any hi-fi speaker.

Totally stupid.


One old Rotel amp has a warning in the manual against capacitive loads,

** Can you quote the actual wording ?

http://www.rotel.com/content/manuals/ra930ax_eng.pdf page 2
"Do not use speakers that have very high capacitive load,
say over 0.1 microfarads, as it may cause a damage to the
power amplifier section. Normally, most speakers are low
capacitance type, but there are a few which do have
excessively high capacitance."

No other Rotel manuals seem to mention that, so the RA930AX must be a lemon.

Probably best not to risk it. Plenty more amps in the junkroom.

--
Eiron.

In article , Arny
Krueger
wrote:
"Eiron" wrote in message


Have we all stopped arguing? May I ask a question? What's the best way
to check if a power amp is likely to be stable into a capacitive load
such as an ESL?

Hang capacitors across its output.

However, presuming that an ESL is necessarily an unusually capacitive
load is not always correct.

That's true. However the old Quad ESL (57) tended to behave like a
microfrarad or two in the treble range. And in the UK for decades that was
what people took for granted was meant by 'ESL' despite some other models
being marketed at various times. I suspect that was why for many years both
UK reviews and some engineers tended to use a 2.2uF cap as a quick
simulation.

That said, above c20kHz the impedance tended to turn over. So the test was
quite likely one of surge current survival as much as stability. As later
experiences showed people, amps that were OK with 2uF could oscillate with
much smaller capacitances. So unless you knew more about the amp simply
relying on a 2.2uF check wasn't very reliable.

Hard to say more since in general almost no-one ever published speaker
impedance plots much beyond 20kHz! Given the effects of the cable as well,
who knows what will happen by the time you get to the order of a megahertz!
:-)


IME if an amp is going to be actually unstable, the worst case
capacitance is likely to be far less than 2 uF. 0.01 uF +/- a factor
of 10 would be my recommended values for fishing for instability.

FWIW my own experience agrees with that! Although again as a 'UK history'
specific, back in the days when 'ESL' meant the Quad 57 most transistor
amps tended to have output zobel with inductor networks and were hence less
likely to be bothered by small capacitance loads.

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

"Jim Lesurf"
Arny Krueger


However, presuming that an ESL is necessarily an unusually capacitive
load is not always correct.

That's true. However the old Quad ESL (57) tended to behave like a
microfrarad or two in the treble range.

** Completely wrong.

One cannot "simulate " an ESL 57 with a single capacitor of any value.


And in the UK for decades that was
what people took for granted was meant by 'ESL' despite some other models
being marketed at various times. I suspect that was why for many years
both
UK reviews and some engineers tended to use a 2.2uF cap as a quick
simulation.

** That test was always stupid and wrong and still is.


That said, above c20kHz the impedance tended to turn over.


** Gobbledegook for it began to rise sharply above 20 kHz.


So the test was
quite likely one of surge current survival as much as stability.

** That test was always stupid and wrong and still is.

As previously posted in this thread - the REAL issues with amps driving
the ESL57 or similar are at LOW and sub-sonic frequencies. The input
transformer easily SATURATES !!



..... Phil

"Eiron"

** Can you quote the actual wording ?

http://www.rotel.com/content/manuals/ra930ax_eng.pdf page 2
"Do not use speakers that have very high capacitive load,
say over 0.1 microfarads, as it may cause a damage to the
power amplifier section. Normally, most speakers are low
capacitance type, but there are a few which do have
excessively high capacitance."


** Nothing there suggests Rotel are referring to ESLs of any kind

No other Rotel manuals seem to mention that, so the RA930AX must be a
lemon.


** You make too many ****ing stupid assumptions.


Probably best not to risk it.


** Risk what ??

All assumptions- no information.

Wot a stupid pommy puke.



...... Phil