Capacitors. Audible differences?
 

In article , Iain M Churches
wrote:

"Don Pearce" wrote in message
... (huge snip)

So the question of the possible audible effects of different capacitor
types is a nonsense. Unless you specify the circuit and exact
operating conditions, it doesn't admit of an answer. And in most
situations where one might be found in an audio chain, there can be no
effect.

Most seem to consider the psu also as an inetgral part of the audio
chain, so changing just the interstage coupling capacitors will not be
enough. To do it properly, we should perhaps also change the cathode
bypass caps

A problem here is that you may be introducing complications by choosing to
focus on (low feedback?) simple valve designs. For reasons like those you
indicate these may be sensitive to a number of sources of problems which
other designs may not be prone to.

If you wish to check capacitors as such, I would personally think it
necessary to consider the actual circuit design/topology again to avoid one
effect looking like something else, and risk giving mileading results, or
masking something. This would improve the chance that you can just change
*one* component and have that produce an effect (or not!) without other
effects coming into play.

Slainte,

Jim

--
Electronics http://www.st-and.ac.uk/~www_pa/Scot...o/electron.htm
Audio Misc http://www.st-and.demon.co.uk/AudioMisc/index.html
Armstrong Audio http://www.st-and.demon.co.uk/Audio/armstrong.html
Barbirolli Soc. http://www.st-and.demon.co.uk/JBSoc/JBSoc.html

Capacitors. Audible differences?
 

On Mon, 3 Jan 2005 13:18:47 -0000, "Mike Gilmour"
wrote:

Agreed :-) I should point out that previous tests had determined that no
sonic differences between monoblocks could be identified blindly prior to
modification (but I think thats so obvious it should go without saying).

Actually, that's not at all obvious. Valvies keep banging on about how
different makes of the same valve sound vastly different, and of
course we all know that valves age, so it's absolutely *not* a given
that two valve monoblocs will sound the same. Hence the suggestion
that if you're *really* interested in whether capacitors are audibly
different, you need a truly transparent amplifier, i.e a good SS one.

Basically, valve amps are just too variable (and hence *by definition*
inaccurate), to be trusted for such subtle differences as *may* exist
among capacitors.

Valves can make a lot of difference to the sound and we select by listening
tests as well as by test parameters - quick n' easy HT off, hot swap signal
valves, wait, ht on. Output valves can be changed out but it takes a bit
longer. I always stick to the same make/type of valve between monoblocks
(which have been preselected pairings anyway for sonic quality).
I found interstage coupling changes made much more difference than cathode
bypass but as Iain says it depends on the cathode resistor values. The PSU
caps are already bypassed so that was not part of the experimentation.
It has been a number of years since I changed these components and since
then many of my fellow enthusiasts have done likewise with similar results.

But hardly ever in *blind* tests, which is the whole point.
--

Stewart Pinkerton | Music is Art - Audio is Engineering

Capacitors. Audible differences?
 

On Mon, 3 Jan 2005 15:09:17 -0000, "Mike Gilmour"
wrote:

"Iain M Churches" wrote in message
...

I have just spoken with the Jensen chap, in fact I read your post out to him
over the phone. He thought the methodology of your test was correct, and
agreed with your findings that interstage caps have the most audible effect.
So I wonder if this would be the place to start?

Well, IMHO the place to start would be with two sonically transparent
amps which sound the same, i.e. good SS ones, but if you *insist* on
using valve amps........................

We are thinking to use only a single valve stage, a 6SN7 in µ-follower
topology and have a good NOS supply from the same maker
(Westinghouse) from which we can select closely matched pair on the
Avo tester.

M tells me from his own experience that the difference
between signal caps as opposed to the difference between
two closely matched valves (not usually discernible) is much more
dramatic.

Iain

Hello Iain,

I'd agree with that. I also think that starting with the interstage
capacitors will give you a good demonstation of the differences. Try
comparing the Jensens to say RS's axial leaded Polyesters - (manufactured
for RS by Vishay-Roederstein) at only about UKP 0.50 each. I'm sure you'll
hear a large improvement with the Jensens.

Obviously, you'll *first* be comparing the two amps with RS (or
Jensen) caps fitted to *both* amps, to ensure that they really do
sound the same, before making any changes..............
--

Stewart Pinkerton | Music is Art - Audio is Engineering

Capacitors. Audible differences?
 

In article , Don Pearce
wrote:
On Mon, 3 Jan 2005 13:27:49 +0200, "Iain M Churches"
wrote:


To simplify matters, we plan to build one unit with Jensen only (psu
caps included) and one with say BHC/Aerovox or Siemens in the psu, and
Evox/Rifa or Arcotronics signal caps, which seem to be commonly used.

As the psu will be a separate unit, it will be possible to evaluate the
Jensen power supply with the "standard component" pre-amp, and vice
versa.

The difficulty I can see here is that with two separate versions there
are factors involved with vastly more part-to-part variation than you
will ever see between capacitors - the valves for example. What do you
propose to do about that? If you are trying to find a subtle difference,
you must confine the changes to the component under test only. A single
amplifier with switchable capacitors would be one way to achieve this.

I would agree. I'd also be concerned with how closely the various caps
shared 'similar' capcitance values. If we are going to worry about possible
tan delta effects of the order of 1E-4 or less then the implication is that
we'd also need to match the gains, capacitance values, etc, to at least
that level of precision as well to exclude other effects masking or
mimicing possible 'material' effects. Otherwise we have no way to interpret
the results as telling us about different 'types' of cap as opposed to
different values of some uncontrolled mix of parameters.

If using valves as gain elements, I'd also wonder about microphonic
differences. These should be small, but I am not sure if they would be
'small' compared with 1E-4 changes...

My general concern about this area, Iain, is that it a minefield of such
uncontrolled minor effects. Hence any result can become quite hard to
interpret unless a lot of care and thought goes into the measurements and
experimental equipment/design. This would be the case just as much if the
result was "can't hear a change" as if "hear a change" as other changes
might be hiding what you are trying to hear.

Slainte,

Jim

--
Electronics http://www.st-and.ac.uk/~www_pa/Scot...o/electron.htm
Audio Misc http://www.st-and.demon.co.uk/AudioMisc/index.html
Armstrong Audio http://www.st-and.demon.co.uk/Audio/armstrong.html
Barbirolli Soc. http://www.st-and.demon.co.uk/JBSoc/JBSoc.html

Capacitors. Audible differences?
 

In article , Stewart Pinkerton
wrote:
On Mon, 3 Jan 2005 15:09:17 -0000, "Mike Gilmour"
wrote:

"Iain M Churches" wrote in message
...

I have just spoken with the Jensen chap, in fact I read your post out
to him over the phone. He thought the methodology of your test was
correct, and agreed with your findings that interstage caps have the
most audible effect. So I wonder if this would be the place to start?

Well, IMHO the place to start would be with two sonically transparent
amps which sound the same, i.e. good SS ones, but if you *insist* on
using valve amps........................

Is there some reason why the tests cannot be performed with the DUT cap as
an interstage between two SS amp sections? Chances are, these could then be
controlled and have better PSU rejection, etc, than a simple valve follower
stage. Thus reducing uncontrolled signal variables.

Slainte,

Jim

--
Electronics http://www.st-and.ac.uk/~www_pa/Scot...o/electron.htm
Audio Misc http://www.st-and.demon.co.uk/AudioMisc/index.html
Armstrong Audio http://www.st-and.demon.co.uk/Audio/armstrong.html
Barbirolli Soc. http://www.st-and.demon.co.uk/JBSoc/JBSoc.html

Capacitors. Audible differences?
 

"Stewart Pinkerton" wrote in message
...
On Mon, 3 Jan 2005 15:09:17 -0000, "Mike Gilmour"
wrote:

"Iain M Churches" wrote in message
...

I have just spoken with the Jensen chap, in fact I read your post out to
him
over the phone. He thought the methodology of your test was correct,
and
agreed with your findings that interstage caps have the most audible
effect.
So I wonder if this would be the place to start?

Well, IMHO the place to start would be with two sonically transparent
amps which sound the same, i.e. good SS ones, but if you *insist* on
using valve amps........................

We are thinking to use only a single valve stage, a 6SN7 in µ-follower
topology and have a good NOS supply from the same maker
(Westinghouse) from which we can select closely matched pair on the
Avo tester.

M tells me from his own experience that the difference
between signal caps as opposed to the difference between
two closely matched valves (not usually discernible) is much more
dramatic.

Iain

Hello Iain,

I'd agree with that. I also think that starting with the interstage
capacitors will give you a good demonstation of the differences. Try
comparing the Jensens to say RS's axial leaded Polyesters - (manufactured
for RS by Vishay-Roederstein) at only about UKP 0.50 each. I'm sure you'll
hear a large improvement with the Jensens.

Obviously, you'll *first* be comparing the two amps with RS (or
Jensen) caps fitted to *both* amps, to ensure that they really do
sound the same, before making any changes..............
--

Stewart Pinkerton | Music is Art - Audio is Engineering

I'm sure Iain will, however to quote Arny ;-) I've been there, done that.
Both my amps were originally fitted with RS axial leaded polyesters, thats
why I suggested them for testing purposes.

Mike

Capacitors. Audible differences?
 

"Iain M Churches" wrote in message
...

"Trevor Wilson" wrote in message
...

"Iain M Churches" wrote in message
...
Happy New Year to All:-)

It would be interesting to know how many on this
group think/find that:

a) capacitor "upgrade" results in a clearly audible
improvement.

**It may, depending on a whole bunch of factors.

Yes indeed. Care to elaborate?

**Sure.
* What the cap is to be used for.
* What other characteristics of the cap happen to be relevant.
* The linearity of the rest of what is in the system.


The sales technician to whom I refer below tells me that
polytetrafluoroethylene (Teflon) capacitors are close to
perfect with sonic superiority, and have the lowest value
of "tan d" (I am sure Jim and Don will know what he is
talking about:-) This is also what Andy has discovered
in his own listening tests.

**PTFE caps are useful at HF, but little else, AFAIK. Regular, quality
dielectrics seem to be entirely adequate, IME.



b) a capacitor is a capacitor, and all caps of the same
value and voltage rating sound the same.

**Anyone who believes such a thing is a moron. Capacitors all have
various
losses and characteristics which may adversely affect sound quality, in
various parts of a circuit.

So quite a few morons on this NG then? :-))))

**It would seem so.


I posted a similar question to another (closed) group run by a
broadcasting
organisation. Of the twenty two who replied, seventeen had experienced
or thought they would experience a clearly audible change/improvement,
three were undecided, and two thought that their level of auditory
perception may not yet be high enough to hear what they expected to be
extremely subtle changes.

The reason for my question is that a local valve amp builder/component
salesman who uses Jensen caps, above all others has suggested that
we should build two identical µ-follower pre amps, one with
Jensen and the other with RS or Farnell standard components.


**Why bother? Just use direct coupling. Coupling caps are just dumb.


He says that the reason that Jensen seem to sound better is due to
their "dielectric absorption" factor. Music is made up of a series of
transients of pulses. If we apply such a pulse to a capacitor this is
equivalent to charging and discharging it, and that any voltage left
on the capacitor at the end of the pulse is distortion. He refers us
to Morgan Jones for further reading.

He is confident that a competent listening panel will pick and
prefer the amp with the Jensen caps, type for type.
If not, some interesting eating of hats will follow:-))

Mike G seems to have already done a similar test, with
Jensen silver foil caps in one channel of a power amplifier.
His group picked out the Jensen channel every time.
I am not suggesting that these audible differences only apply
to Jensen. There are many other makers whose products
may well fall into the same category. Interesting.

**Without knowing the details of the tests, I can't comment further. I
suggest that direct coupling makes far more sense, however.


--
Trevor Wilson
www.rageaudio.com.au

Capacitors. Audible differences?
 

"Stewart Pinkerton" wrote in message
...
On Mon, 3 Jan 2005 13:18:47 -0000, "Mike Gilmour"
wrote:

Agreed :-) I should point out that previous tests had determined that no
sonic differences between monoblocks could be identified blindly prior to
modification (but I think thats so obvious it should go without saying).

Actually, that's not at all obvious. Valvies keep banging on about how
different makes of the same valve sound vastly different, and of
course we all know that valves age, so it's absolutely *not* a given
that two valve monoblocs will sound the same. Hence the suggestion
that if you're *really* interested in whether capacitors are audibly
different, you need a truly transparent amplifier, i.e a good SS one.

Thats why serious valvies use valve testers. In my long experience with
valve amplifiers I have never known the sound to significantly change unless
a valve is reaching the end of its useful life i.e. can no longer maintain
the required cathode current. With signal valves this can be of many
thousands of hours. With careful choice of valves and preventative checks
there is no identifiable 'variability' of performance and there is
absolutely no reason why a pair of monoblocks can't be told apart in blind
testing conditions.


Basically, valve amps are just too variable (and hence *by definition*
inaccurate), to be trusted for such subtle differences as *may* exist
among capacitors.

Please explain your definitions of variability of a properly maintained
valve amplifier under normal conditions?



Valves can make a lot of difference to the sound and we select by
listening
tests as well as by test parameters - quick n' easy HT off, hot swap
signal
valves, wait, ht on. Output valves can be changed out but it takes a bit
longer. I always stick to the same make/type of valve between monoblocks
(which have been preselected pairings anyway for sonic quality).
I found interstage coupling changes made much more difference than cathode
bypass but as Iain says it depends on the cathode resistor values. The PSU
caps are already bypassed so that was not part of the experimentation.
It has been a number of years since I changed these components and since
then many of my fellow enthusiasts have done likewise with similar
results.

But hardly ever in *blind* tests, which is the whole point.

Yes, surprise - blind tests in all these cases, that is the point. We are
all aware of sighted expectation bias which usually works the other way
around as far as I'm concerned anyways e.g. the Melos amplifiers looked
wonderful but IMO sounded crap plus many other examples. Some of the
god-awful ugliest amps have sounded the best eg the old P100 sure ain't
pretty but IMO and many others it sounds great .

--

Stewart Pinkerton | Music is Art - Audio is Engineering

Capacitors. Audible differences?
 

Stewart Pinkerton wrote:
On Mon, 3 Jan 2005 13:18:47 -0000, "Mike Gilmour"
wrote:


Agreed :-) I should point out that previous tests had determined that no
sonic differences between monoblocks could be identified blindly prior to
modification (but I think thats so obvious it should go without saying).


Actually, that's not at all obvious. Valvies keep banging on about how
different makes of the same valve sound vastly different, and of
course we all know that valves age, so it's absolutely *not* a given
that two valve monoblocs will sound the same. Hence the suggestion
that if you're *really* interested in whether capacitors are audibly
different, you need a truly transparent amplifier, i.e a good SS one.

Basically, valve amps are just too variable (and hence *by definition*
inaccurate), to be trusted for such subtle differences as *may* exist
among capacitors.


*By definition*?! LOL! Where on earth do you get that? Within your
sample (ie global) are you saying that there isn't a single type of
'accurate' valve amplifier?

Rob

Capacitors. Audible differences?
 

Rob wrote:
Stewart Pinkerton wrote:

On Mon, 3 Jan 2005 13:18:47 -0000, "Mike Gilmour"
wrote:


Agreed :-) I should point out that previous tests had determined
that no sonic differences between monoblocks could be identified
blindly prior to modification (but I think thats so obvious it should
go without saying).



Actually, that's not at all obvious. Valvies keep banging on about how
different makes of the same valve sound vastly different, and of
course we all know that valves age, so it's absolutely *not* a given
that two valve monoblocs will sound the same. Hence the suggestion
that if you're *really* interested in whether capacitors are audibly
different, you need a truly transparent amplifier, i.e a good SS one.

Basically, valve amps are just too variable (and hence *by definition*
inaccurate), to be trusted for such subtle differences as *may* exist
among capacitors.


*By definition*?! LOL! Where on earth do you get that? Within your
sample (ie global) are you saying that there isn't a single type of
'accurate' valve amplifier?

Rob

I think his logic goes this.

If they change with use, then they are not the same at time a and time b
If they are not the same, then even if they were "accurate" at time a,
then they cannot be "accurate" at time b.

This is somewhat self referential as the definition of "accurate" is
what makes this true. A different version of "accurate" where more than
one "accuracy" can exist does not.

--
Nick

--
Nick

"Life has surface noise" - John Peel 1939-2004