PDA

View Full Version : Large capacitor shoot out



SW1X
15-10-2016, 09:51
SW1X AudioTM Design Large Capacitor Shoot Out

Here is a Shoot Out with a difference. You will be used to auditioning Hi Fi components, but now comes the opportunity to audition just a single component within a DAC and gain an insight into the painstaking process of voicing audio components. The subject of the shoot out is electrolytic capacitors with over 1000uF capacity.

Due to our continuous dedication to improve the quality of sound we routinely carry out voicing for all the parts we are employing in SW1X Audio Design products. We are researching the market for better sounding materials and components. On this occasion audio enthusiasts and music lovers are welcome to participate in judging the sound. It is a unique opportunity to find out how passive components (in this case capacitors) sound, even though buried and often ignored inside your audio equipment. Unfortunately the number of referees is limited to 5, therefore do not hesitate to apply to secure yourself a place.

The approach is quite straight forward. We are looking to control for just one factor i.e. one cap in one position. All burned in caps (for 96 hours as a minimum) are substituted one by one into a power supply of I/V conversion stage of 2 identical standard SW1X DAC 1s (one DAC 1 could be also Signature version as the reference). We will audition all caps in the form of a blind test i.e. all caps will be covered and marked with a number. We then will play all caps sequentially for duration of a short song. The referees write down their impressions (either freestyle or on a predefined form) and either discuss them immediately or later after the end of the run.

Refreshments will be served

The objective is to find a cap that sounds least compromised in all aspects.

Everything being relative, we are interested in opinions on a larger scale. There are currently over 10 capacitors short listed but we are open to suggestions. Current list includes:

Rubycon Black Gate STD, AN Kaisei, AN Standard, ZLH
Elna Silmic II, Cerafine, Tonerex, Startget, RA3, RE3,
Philips/Vishay BC 037, 046, 048, 047, 135, 136 Series
Panasonic Pureism, FM
Nichicon Muse KZ, FG

Of course, the list is far from complete. So feel free to suggest or bring any other capacitor with you for the contest.

The venue is located in Coggeshall, 25min off Stansted Airport

Date and time: noon time, Saturday, the 5th of November.

There will almost certainly be opportunity to listen to a hierarchy of up to three different complete audio systems within the SW1X household.

Should you be interested, please do not hesitate to get in touch with us

http://i1224.photobucket.com/albums/ee376/sw1xad/20161015_134548_zpsu6um7vaq.jpg

Light Dependant Resistor
15-10-2016, 10:51
A schematic might help readers, to show where said mystery capacitor is positioned.

Could I suggest the list also include a capacitance multiplier using a BC547 and a 2.2uf capacitor
arranged appropriately in the internal circuit. Such a small capacitor which could be a 0603 surface
mount then is multiplied by the transistors current gain expressed as hfe.


Cheers / Chris

SW1X
15-10-2016, 11:07
Dear Chris,

Thank you for your input. I am afraid but it would not be possible to publish the circuit design as it is proprietary. All I can say that circuit is elegantly simple and that any change is very audible, so no "golden ears" are needed to hear the differences.

The caps in question are going to be the first capacitor after the rectification (that would be in place of the blue cap in the picture below) before the voltage shunt regulation, which is followed by another cap (which will be exogenous to our experiment).
The rest of the circuit is made of very revealing AN Kaisei capacitors. All caps in question are going to be attached via copper clips, with each cap replaceable in a matter of seconds. Every cap has its sonic signature, which are quite memorable even after prolonged periods of time.

Any other (automated or not) switching method such as relays or mechanical toggle switches are interfering with the results too much- something we would like to avoid happening.

All we are looking for is for enthusiastic volunteers to share, broaden or to confirm our observations.

http://i1224.photobucket.com/albums/ee376/sw1xad/20161015_135904_zpsonz9k51r.jpg

bumpy
15-10-2016, 12:05
I have spent a couple of very interesting afternoons at Slawa's house listening to the audio consequences of changing capacitors, and also power supply diodes. In the comparisons I have experienced, my personal criteria has always been "would I spend good money to own that sound?"

I am past retirement age and my high frequency sensitivity is shot to pieces, but even so the results are far from subtle and a change can easily make or break a Hi-Fi component. Each capacitor has its own signature and I have heard capacitors fitted to £10,000+ DACs sounding pretty awful, so it's not a question of increasing cost automatically bringing you better performance. I suspect we have a few 'manufacturers' on here that know all this, but as a purchaser of equipment, and electronics numpty, I found this a bit of a revelation.

fatmarley
15-10-2016, 15:04
Interesting thread. With capacitors, I've found that what works in one piece of kit may not work well in another. They all seem to act like filters and synergy between all the component choices is key.

Not sure if you're interested but here's my experience playing with caps in my old Onix OA21 integrated amplifier (sorry if it's going off topic).
I keep a diary of my experiences because my memory is bad.

4-pole caps in power supply with 47uF Ansar polyproylene on feedback and Mundorf MCap Supreme
on coupling makes a perfect midrange and treble but slow bass. Without 4-pole caps the
sound is far too bright.

Permanently remove 4-pole caps and fit K52 Wet tants in coupling and feedback. Bass punches much harder. Rolling Stones - Beast
of Burden vocals sound harsh. Same track sounds fine with Cambridge Audio AM1 (electrolytic coupling
and feedback)

Samwa 100uf 100v SG (as used by Rega) electrolytic in feedback with CA AM1 electrolytic coupling cap. Not bright or glassy at all
but lacking a bit of detail.

Wet tant in feedback with 10uf MMK in coupling, very good if a little bit safe, maybe a touch grey.

Put wet tant back in feedback and replace Wet Tant in coupling position for Mundorf Supreme 4.7uF and the
sound is a touch glassy in upper midrange (not harsh though). London Grammar is a good test for glassy vocals.
Replaced Mundorf coupling cap with Jantzen Z superior (Wet tant in feedback) A bit smoother and less glassy (best so far).

Solid Tant Feedback (Nos Kemet 100uF 20v T110), Jantzen Z Superior coupling - Good but a bit flat. Replace Jantzen
with Mundorf Supreme, much more open sounding and glassy sound gone. Not sure I can improve on this but may try
a brand new solid tant (Kemet 100uF 6.3v SMD tants on their way).

Thoughts - Wondering if DCL (dc leakage) can degrade the sound? Naim use a very low voltage tant (6.3v) in feedback position.
The lower the voltage the lower the DCL. I've read that DCL goes up in as tants age (I think that's the only thing that changes in a tant)
and Naim users report a sound quality improvement after having the tants replaced (could be because of the main smoothing caps though)

SW1X
15-10-2016, 16:12
Dear Matt,

Your observations are true and certainly are have their validity. We are only interested in effects of one factor ( i.e one capacitor) that directly affects the sonic signature of the DAC 1. This is not the first time and not the last time we are voicing our products. Of course there is always an interaction between different components, which is the trickiest part of them all. That is why we by all means avoiding to employ any switches or relays, knowing how those parts interfere with the sound. Every passive and active component of our designs underwent a similar procedure. So, we pretty much know how each part, including PCB, IEC and phono sockets sound like. It would be foolish to claim some absolute accuracy of the findings in experiments which are based on individual and subjective perceptions. Sure, different environmental conditions (different circuits, different people and all other variables that we are not going to control for) will certainly bring variations into the results. However, everything is being relative, every cap has its own sonic signature (a common factor) that will affect any circuit in a very similar way. In an extreme case, a rubbish capacitor will always sound rubbish in any circuit just to a different extent. However, in some circuits one would not even notice that. In our circuits one rubbish capacitor spoils the whole sound irrecoverably.

Unlike with somebody else circuit, the biggest advantage in our experiments is that we can control for every factor, we choose to (one by one of course), in an elegantly simple circuits of our own choice. We repeat experiments similar to this but with different components such as resistors, diodes, conductors, inductors etc. on a regular basis anyway. So, here is an opportunity for curious and inquisitive minds.

Slawa

fatmarley
15-10-2016, 17:29
The only slight concern I would have is how a lot of people find it hard to hear differences in unfamiliar systems. I was at a guys house a few years ago and he swapped a preamp over to show me this "large difference in sound" and I can honestly say I couldn't hear any difference, but it was quite obvious by the guys reaction that there was a difference.

Have you heard about the Harbeth challenge? Alan Shaw believes all amplifiers sound the same and he was happy to put his money where his mouth is. He was offering a pair or free M40.1 speakers to anyone who could identify different amplifiers under a blind test. Obviously amplifiers don't sound the same, so I'm guessing that he knows how hard it is to hear differences in unfamiliar systems (or maybe it's his speakers that aren't very revealing :) )

SW1X
15-10-2016, 19:09
The only slight concern I would have is how a lot of people find it hard to hear differences in unfamiliar systems. I was at a guys house a few years ago and he swapped a preamp over to show me this "large difference in sound" and I can honestly say I couldn't hear any difference, but it was quite obvious by the guys reaction that there was a difference.

Have you heard about the Harbeth challenge? Alan Shaw believes all amplifiers sound the same and he was happy to put his money where his mouth is. He was offering a pair or free M40.1 speakers to anyone who could identify different amplifiers under a blind test. Obviously amplifiers don't sound the same, so I'm guessing that he knows how hard it is to hear differences in unfamiliar systems (or maybe it's his speakers that aren't very revealing :) )

Your concern is well justified. It is a question of how revealing the rest of a system is. Cannot speak for anybodies else system but in our system(s), where everything is made of finest materials, one cap (or any other component) makes a dramatic difference. That works without having very trained ears. After all there is always a reference of how it sounded before and relative to other DACs (performance level 1 vs level 2 vs Signature version as a reference).

You are welcome to come over and to hear it for yourself

fatmarley
16-10-2016, 05:58
You are welcome to come over and to hear it for yourself

It's a bit too far for me, otherwise I would.

Hope all goes well. I'll be interested to hear the what people think.

bumpy
16-10-2016, 11:48
On my own system I have never found comparisons after after making changes particularly easy unless they are massive. I think it's all that jumping up to change cables etc, and the additional 'pressure' of judging your own system. At Slawa's however every change seemed much more obvious. I think the reasons for this are several.

1. Slawa does all the jumping about.
2. Tracks are selected that reveal all the required elements of high frequencies, low frequencies, sound stage, clarity, naturalness etc within a short period - maybe a minute or two.
3. Gaps between listening passages are short, which for me with the attention span of a gnat is great.
4. Repeated use of a quality reference component such as a Blackgate capacitor helps to anchor the comparisons.
5. Slawa has a very revealing system, so there is no hiding differences.

fatmarley
16-10-2016, 12:41
2. Tracks are selected that reveal all the required elements of high frequencies, low frequencies, sound stage, clarity, naturalness etc within a short period - maybe a minute or two.

With some capacitors, I've found that you could play a large number of tracks before you notice something wrong. A good track is C2C - Down The Road. Vocals can sound harsh if you get things wrong.

SW1X
16-10-2016, 12:48
It may sound bogus to some folks that one passive component makes a difference but that would be hardly surprising if the rest of a system is made of mediocre materials employed by over-engineered circuit designs.
Those who have done a little bit of experimentation know that quantity and complexity of components/features of an audio product can be quite misleading. The amount of components/features does not necessarily improve the quality of sound as we are lead to believe. On the contrary, since there are no perfect components, parts or materials, the complexity and quantity of components/features contributes only to degradation of sound quality. Therefore, the sound of a recording cannot be improved but only degraded by the quantity of components. Hence the only best possible performance can be achieved when fewer and only highest quality components are employed. One can focus on the choice of materials and components, only when the whole circuit becomes simple and elegant. Discriminating between good and bad parts becomes also easier when there are only a few parts in the circuit.

There are no unimportant or negligible aspects- everything matters when it comes to maximum sound quality! Maximising the sound quality of an audio system is a cumulative process. The improvement in one part of a system may seem to be negligible but as a sum of all parts, it becomes more than significant. That is the core of our philosophy at SW1X Audio Design.

fatmarley
16-10-2016, 13:12
Hi Slawa,

Just out of curiosity, what will be the rest of the system?

SW1X
16-10-2016, 15:25
We have several custom design systems based on 3 performance levels. Each system is based on the same SW1X Audio Design philosophy. In general we employ only the finest components such as tantalum resistors, silver foil, Black Gate & Kaisei capacitors and finest UHiB and Permalloy core transformers wired with fine silver conductors starting from mains, internal wiring, transformers, speaker wiring, etc. The performance level of each system is constrained only by the amount of silver and sophistication of power supplies corresponding to the cost levels. The system, where the voicing takes place is based on SW1X Signature performance level 3 transport, SW1X DAC 2 Signature, SW1X silver wired stepped attenuators based on AN silver tantalum resistors, silver wound Permalloy core transformer based SE 300B mono power amps and SW1X AlNiCo magnet based fully silver wired loudspeakers.

fatmarley
16-10-2016, 18:23
Oh I see. A full SW1X Audio system (didn't see any speakers on your website).

SW1X
17-10-2016, 09:20
We do offer speakers based on vintage AlNiCo and field coil drivers. However, those designs are all custom orders and not in series production. Eventually one of those designs will find its way into small and limited series production some time in the future. Therefore it does not make any sense to advertise them at the moment. One cannot judge and then buy speakers based on their looks without auditioning them anyway.

DSJR
17-10-2016, 16:41
You might hear a supposed difference in the caps when freshly fitted, but there's been enough speak here in the past where fancy-foo caps take a good few hours use to fully bed in - so all you'll be doing is 'comparing' fresh new caps, the fancier of which won't have formed properly. I fear that the preferred item on the day, assuming you think you can hear a difference at all, will be sounding totally different after a few hours' use, where a less favoured one might even be better.

Mind you, the best circuits seem electrolytic cap-proof and try not to use them at all if at all possible...

SW1X
17-10-2016, 16:49
You might hear a supposed difference in the caps when freshly fitted, but there's been enough speak here in the past where fancy-foo caps take a good few hours use to fully bed in - so all you'll be doing is 'comparing' fresh new caps, the fancier of which won't have formed properly. I fear that the preferred item on the day, assuming you think you can hear a difference at all, will be sounding totally different after a few hours' use, where a less favoured one might even be better.

Mind you, the best circuits seem electrolytic cap-proof and try not to use them at all if at all possible...

Dear David,

I can see where you coming from. Not all electrolytic caps are created equally. Some electrolytic caps sound better than most metalized foil caps I came across. And in an application with heavy load where noise is critical, capacitance over 1000uF is an absolute minimum. Film caps are not a solution because of the size constraint. Furthermore, you may have missed our statement. It would be ignorant and prejudiced to voice each cap without sufficient run in time. We are controlling for this kind of bias. Please refer to the first post of this thread.

"....All burned in caps (for 96 hours as a minimum) are substituted one by one into a power supply of I/V conversion stage of 2 identical standard SW1X DAC 1s (one DAC 1 could be also Signature version as the reference)..."

SW1X
19-10-2016, 11:40
We have 3 people so far who would like to participate in the shoot out.

There are 2 places left

SW1X
19-10-2016, 14:05
Just to make it perfectly clear: we are organising a capacitor shoot out. This is not a type of bake off. One and foremost objective is to illustrate of how much one single capacitor in the power supplies can shape the sonic signature of equipment. The other purpose is to gain insight which sound character is more preferable to the audience and to illustrate our approach of designing an audio product.

SW1X
20-10-2016, 17:40
We decided to move the shoot out to Saturday the 5th of November. There are still 2 places left.