[PinkFloyd]

What type of cap do you think sounds best and why?

 

[ppl]

None. With the availability of DC stable circuits there should be no reason to AC couple and if you think this is to protect from sources that might have excessive DC on the output i sugest use another source any source that is outputting DC is defective IMHO. Oh so ya want to reduce turn on and Off transients caps make that worse. to eliminate the Thump un plug you cans first. Just remember Coupling caps in Audio Amplifiers are

and you will not sleep untill each and every one is removed from your Analog signial path

 

[00940]

ppl : that's a theoretical opinion. *

Pinkfloyd : you should perhaps add paper in oil caps.

*edit : coming from a guy who just bought 2x 220uF MKP caps for an otl tube amp

 

[rickcr42]

Capacitors are your friend-try making a tone control/filter/crossfeed without one !

What about RIAA EQ networks ?

Speaker crossovers ?

RF compensation ?

You think there are no caps in those opamps ?

Oh yeah there are !

They hide even better in there !

(the compensation that sets bandwidth and slewing)

While fully balanced discrete designs with matched devices have their strong points there are many times i will take the simple capacitor coupled single ended class-a single stage device ,be it tube or solid state , over the complex

All methods are valid if the implementation is done correctly

Hey pinkster ,you forgot teflon man

 

[PinkFloyd]

Quote:

Originally posted by rickcr42 Capacitors are your friend-try making a tone control/filter/crossfeed without one ! What about RIAA EQ networks ? Speaker crossovers ? RF compensation ? You think there are no caps in those opamps ? Oh yeah there are ! They hide even better in there ! (the compensation that sets bandwidth and slewing) While fully balanced discrete designs with matched devices have their strong points there are many times i will take the simple capacitor coupled single ended class-a single stage device ,be it tube or solid state , over the complex All methods are valid if the implementation is done correctly Hey pinkster ,you forgot teflon man

Can you add teflon into the poll Rick?

 

[ppl]

ricker> Glad to hear from ya please not that RIAA and other such frequency shaping netwoirks are always present by nesasty and as such are a given Evil however the network is often in the Feedback loop where it is less sonicaly intrusive that directly in the signial path. Yes op amp do have compensation capacitors inside and capacitors in monolithic for are indeed sonicaly bad but again thay are a nessary evil. but my point is why add more just to complicate things

00940> Yes this is as i stated MHO and YMMV but remember every time you add an uneeded component in the signial path you intentionaly by you own work add Distortion and noise to the circuit.

Everyone> You all Know by now i dont like slew limit caps and compensation caps because just like DC blocking caps thay can be eliminated by proper Design. (Your ears may not Mind)

 

[rickcr42]

there are times that having a simple circuit with a cap is better sounding than a complex circuit with dc coupling

everything in audio has a sound and the trick is to get good sound regardless of the topology

Tube stages come to mind as gear that more often than not have cap coupling but yet can sound exeedingly transparent

Same with JFET/MOSFET SE stages unless carefully matched devices are used ,and even then the balance will go "off" over time due to componant aging

Comprimises abound and it is the least intrusive ones that we decide on in the context of the use of the device

I don't think i have one single DC coupled audio device in my residence

Oh yeah , my LH0066 Buffer "clone" done with discrete jfets/bipolars

but the input stage is cap coupled so it really does not count

I just like single ended class a designs for the basic simplicity .If i decide to muck up the design with extras it is usually a CCS or cascode stage for a high psrr

 

[aos]

Unless you have an one-in-all device, it's really recommended to have a capacitor before the final amplifier in the chain. If you are a DIY'er this is an absolute requirement. Especially if you use speakers. Unless you know that you will never, ever connect an untested or breadboarded source / preamp / whatever to it, you need that cap to protect you from your own stupidity. Otherwise you stand a risk of paying a very high price for say a wire that got loose while you transported your latest invention from one room to your main rig.

 

[ppl]

Ok I Know i Live in an alternate sonic universe where accuracy rules over Euphoric coloration and every coupling cap i have used colors the sound to a degree as to totally mask the Micro dynamics that DA thing I think Plus adding a flat 90 deg Phase shift across the audio band with deviations starting at 100 Hz and lower unless your Cap is real large in value. So ya got time smear that is frequency dependent like tone control circuits without the Luxury of turning it off.

aos I am shocked thank god you omitted these Caps in your DAC or i would be using something else.

Rick Please take no offence but you know i gave up on Tubes 20 years ago cause of Micro phonics however I have designed and successfully Built DC coupled Tube amps before.

I also do not belong to the flat earth society

 

[Prune]

Quote:

Originally posted by ppl Ok I Know i Live in an alternate sonic universe where accuracy rules over Euphoric coloration and every coupling cap i have used colors the sound to a degree as to totally mask the Micro dynamics that DA thing

Ha! I find it amusing how many self proclaimed golden ears make these types of claims, only to utterly fail in a real blind test. * **

* This comes from someone that uses pure silver for wiring and teflon for insulation.

** The one argument against real blind tests of components that I accet is that, while one change may not make a noticeable difference, a number of such changes put together may be detectable. Still, I've yet to see of a scientific study that shows a quality coupling capacitor can be heard in the signal path.

Quote:

Originally posted by ppl however the network is often in the Feedback loop where it is less sonicaly intrusive that directly in the signial path

Feedback is part of the signal path. This reminds me of the way audiophiles often erroneously think that components that are not in series with the signal path (such as shunt-to-ground ones) are not as critical and lesser quality parts can be used.

 

[ppl]

Prune. at 50 years of age i do not nore have i ever claimed to Prune. at 50 years of age i do not nor have i ever claimed to have Golden ears however decades of Live Recording and playing that recording Back on assorted High End systems have Taught me what reality sounds like and in what ways Audio systems add Distortion and tone colorations in addition to Compression. I do have a few fellow Head-fiers that i trust to verify my listening test and to make suggestions. In addition Blind ABX Listening tests have proven to be so stressful to the Listener that value Judgments over a wide selection of the Population have difficulty in such a stress test. This same argument has gone on for Decades and probably never really be answered. What is A Known fact is that Two Audio components that Measure the same can and most often do sound quite different.

While Feedback and the power supply are indeed part of the signal path they are secondary as opposed to primary paths. Capacitors must be used in the Power supply and Mandatory EQ networks such as RIAA EQ. These are a given

However just because one must live with a set of uncontrolled variables is no reason to simply throw the Baby out with the Bath water and compromise the entire scheme of things. Yes it is true there is a Difference in sound between any two seemingly similar Audio components most people understand this. If all equipment that measures well below what one should perceive then all Equipment that measured below this threshold would sound the same while in fact the opposite is true.

 

[Prune]

What's your age got to do with it? In science a claim is judged on its merits, not the author's seniority (and yes, I regard this discussion to be of a scientific issue -- is there objectively an audible difference or not).
Quote:

In addition Blind ABX Listening tests have proven to be so stressful to the Listener that value Judgments over a wide selection of the Population have difficulty in such a stress test.

Judging by the numerous discussions at audioasylum, many will not agree that such a thing has been proven; not to mention that ABX is just one approach to get a scientifically valid result (alternatives such as ABC/hidden reference are generally considered superior). Anecdotes from that site and also diyaudio of informal tests have also had negative results. While a negative result does not prove that there is no difference (only a positive result proves anything), multiple negative results, statistically, do strongly suggest that what one is looking for is really not there.
To a non-audiophile your statement would sound like an excuse, and would not be accepted in a scientific circle. Thank goodness pharmaceutical companies are not allowed to use these types of excuses against the blind studies that their drugs are forced to go through. Sure, this is audio and it's not a matter of life and death; however, it is a matter of money -- the money of newbies that listen to and are influenced by the claims of audiophiles.

 

[00940]

On the list, I would pick polypropylene. Wima mkp are ok but not that great, I prefer stuff like audyn. The 630V serie is ok, better than the 400V one.

Teflon is supposed great but very, very hard to find. Polystyrene is also difficult to pick for coupling caps because values are too low usually. Both are supposed to be great.

Polyester is really bad compared to mkp. I tried wima mks-02 and lesser MKT ones and they really can't compare to mkp. Very scratchy sound.

I'm trying to pick some NOS paper in oil caps. They're not that expensive and are supposed to be the best caps ever for coupling. Very big though.

 

[Prune]

00940, I have read on several places that while capacitors such as the paper in oil or the Audionote copper or silver foils are more euphonic than the rest, that is due to the type of coloration they add to sound, and that polypropylenes such as the MIT Multicaps and the Audience Auricaps (and their solid lead predecessors, the Sidereal caps) are the most neutral and transparent. The latter two are also much cheaper, but still not exactly cheap.

 

[rickcr42]

Quote:

While Feedback and the power supply are indeed part of the signal path they are secondary as opposed to primary paths

This is where i have to strongly disagree

When we are talking about the final output coupling stage to the transducer , be it headphones or loudspeakers , the power supply IS the sound !

all an output stage is doing is to convert the power supply to a usable voltage to drive the application

This is where you make or break the entire circuit soundwise

Live recording ?

i guess you don't use condenser microphones 'cause if you do and fail to Ac couple the stage you will be sending a pure +48 VDC directly into the microphone preamp !

Said 48 volts being amplified by a factor of the gain of every gain stage that follows until it hits the speaker/headphone with enough power to kill it

ADC input ?

Fail to have a low or high end cutoff and the error signal will play hell with the audio signal .

again caps

Same with instrument amps , tone controls on the mixing board , limiters ,filters ,etc.

Your signal has already seen more capacitors than you will find in the mouser catalog

DC coupling is nice but no caps at any cost is just unrealistic in the real world and in some cases is detrimental to the final result

and anyone that uses opamps and claims to be cap free is fooling themselves also .

as stated ,while the caps may only be in the feedback loop (not entirely true , there are caps all over the place on the substrate) all opamp connections are signal connections of a type and as such have audible consequences.

the feedback will cancel out some in the same way it cancels out other common mode signals but it is still there and can not be ignored.

And what about instrumentation or balanced monolithic designs ?



Fail to use caps as compensartion in some power amps designs and you will see oscillation , an unstable amp

does this mean poor design ?

It can ,or it can mean a design that attempted to attain a certain thing and in order to satisfy the end requirement choices needed to be made

There are no absolutes in this field or any other and all designs are valid if the final implementation is correct on paper then fine tuned by ear

I have no problem using what i need to get from point A to point B

Caps ,transformers ,inductors , resistors-if it makes for a more stable circuit in the end i go with what i need and if it mucks up the sound i redesign until it is right