Best electrolytic capacitors

Member
Joined 2010
Paid Member
Hi thisusername,
Yes, Teflon are, and we use them for specific things. However, the other capacitor types are more than close enough to being perfect for audio.

Teflon has exceedingly low leakage current, so useful for very high impedance applications. We do not have that situation in audio and chasing this kind of perfection makes zero sense. In other words, over Polypropylene, they ain't going to help you one tiny bit.

Well, anecdotally, I have done the teflon cap upgrades, at the factory, to a Conrad Johnson PV9 and ET3SE... they transformed the sound for the much better. Some parts are actually worth it. like Vishay resistors in certain parts, like gain controls.

The truth is we don't truly understand psycho acoustics and the measurement set we use does not quite correlate to what we hear. You may say that those teflon caps do not measure "better" in audio, but I tell you I hear the difference and the sound is much better. I lived with that PV9 for 25 years before the upgrade and three months with the ET3... so I knew their sound...
 
  • Like
Reactions: 1 user
Member
Joined 2010
Paid Member
This is a great term. I will be using this a lot.

You have to be very careful here.... don't throw the baby with the bathwater.

On the one hand you got the Julian Hirsch / David Ranada / ASR crowd, OTOH, you got the Wilson / d'Agostino audio as jewlery crowd.

The "anti audio fashion" vs. the "audio fashion" crowds.

There is a fine line in between... and just because something is in fashion today, doesn't mean it's just plain fashion and marketing fantasy.

The ASR crowd rejects the validity of the act of listening... but things that today are in "fashion" don't mean that in the future the physics of their validity, with concurrent measuring techniques, won't scientifically justify them.

Ironically, the ASR crowd will then support some of today's audio fashion statements.

Me, I never did got a Tice Clock... I figure I didn't need no fancy Euro metric seconds.... West Coast standard seconds are fine with me.
 
Administrator
Joined 2004
Paid Member
Hi thisusername,
I won't argue it, but you have some misconceptions.

Audiophiles generally don't understand electronics or signal processing. Nor do they understand reliability or even basic measurements. Couple that with how your mind will affect what you think you hear, no wonder there is disagreement. On top of that, beliefs, just like religious arguments. Throw into the mix people who will advance ideas for commercial purposes and it's a wonder the average person hasn't given up entirely ... oh wait! They have.
 
  • Like
Reactions: 3 users
Administrator
Joined 2004
Paid Member
Hi Tony,
Well, the condition of the original parts comes into your experience. Also what is buried in your head, ideas etc. Confirmation bias looms large here.
we don't truly understand psycho acoustics
Well, we actually do. The high end crowd doesn't like the answers! lol! Measurements today actually do correlate to what people hear. But you have to use children or women who don't read the audio press to get subjects with clear minds.

On top of that your auditory memory isn't good except for gross changes. What is in your head makes a bigger difference.

You know what? Good engineers and technicians always listen to the equipment as well as measure it. Two groups of dedicated engineers I know for certain would be the Nakamichi and Cyrus engineering departments. I'm pretty certain Marantz did as well, and Sansui. There is no line. There is a group that doesn't want to spend the money for adequate test equipment , training and learn to use it and do proper tests, and a group that has done that. The group that has also listens 'cause that is cheap and easy to do. I suspect that those who don't understand and want to be heard are your subjectivist core. When they are wrong, they claim we can't measure what they hear. Well, certainly today we certainly can and routinely do. If you can hear something that is real, we can easily measure it. We can measure well over 20 dB lower than the human body can sense through any mechanism. In the 70's and 80's no (but we were improving).

Some of us were looking at the residuals from the monitor output of THD meters (me for one) early on. The HP 3580A was an audio spectrum analyzer available in the late 70's I think (I have two). Better instruments followed and our understanding of what mattered grew rapidly. An oscilloscope can't show much on it's own, and that plus the needle on a THD meter was what the subjectivist crowd points to. To "get it", you have to see what we have now. For that you have to spend a fair amount of money and invest in training (years). I show clients and they are amazed.

I'm sorry, but today the measurements are a science. Interpretation is partly experience and an art to know what is important, but it is all laid bare and visible.
 
  • Like
Reactions: 5 users
Disabled Account
Joined 2002
Hello,
The Exellia/ Sic safco brand looks like a serious player in the " capacitor market"
Weather you wanna make the extra investment to use them is up to you of course.
The French people around Jean Hiraga had a preference for the co38, co39, frs and tfrs types. The last two designed for low and very low series resistance. The first two for long and very long life. It seems that these special " specs" will require more careful designing and manufacturing and eventually giving you better sound too if you are ready to believe it.
You should decide yourself once you need to buy them. Right now i dont have any clue how much a 47000 microfarad 40 volts co39 would cost compared to a " normal grade" Kemet.
Greetings Eduard
 
Audiophiles generally don't understand electronics or signal processing. Nor do they understand reliability or even basic measurements.

I'll help out the ones that want to go forward. Because there is so much that either has been forgotten or lost. There are several approaches to applications, and just like any other tech, I am open to all things.

I'll give people the explanation of why I build something a certain way, or chose one part or another or give them a formula or reference to a book, because if they are asking, they want an answer. I have notice some mainstream transformers are not as great as they could be, but there are other factors that can come into play if they are misused. Certain construction types of parts I will apply in one circuit but not another but its all in application. What I quoted that capacitors are #2 as a coupling device and transformers #1 came from textbooks and mentors like Rupert Neve that taught me about parts and parts construction. So this wasn't any religious or spaghetti monster thing like that.

I know science change, and I know capacitors have gotten better in some areas and worse in others and so has transformers, and since my old arthritic self can't stand there and patently spin a transformer like I used to, I have to rely on others to do it with my design specs. Other things changed in the world and some have missed the boat on it. Like analog i/o conventions and so we have good mix engineers that still think everything is at 600 ohms and get upset when you tell them you have to provide a load for the transformer and adjust to the 10K ohm input z for the scale of their equipment to measure correctly. The other way, mic preamps are now like 2.5K instead of 250 ohms, and some either take it in consideration, and boost later while others give up on the interface preamp and use a seperate analog pre that was designed with the 80 year old specifications. That work with their antique $15K microphone.

But I have noticed that there is a site called Audio Science Review that according to a lot of my engineering customers in the pro audio world should be called Fake Audio Science Review just because the trade paper does a review on an item and the exact opposite opinions ASR does online. So, they kind of joined the ranks of like SOS where their review might be bought and paid for.

As far as other DIY electronics site, GroupDIY turned into a design scab hub. They don't want to know or learn the formula or even your opinion on a part. Because they have some part they want to sell, or want you to show a design so they can copy it so they can sell a kit online. I know a few designers that is upset with them about that.

I'm still evaluating DIYAudio, and I see its not so bad so far. If I need to provide pictures, that's fine, but after looking at some things like a/b tests on youtube videos, I find they are worthless to use any of thier videos. It must be due to the production equipment used and formatting videos. Because I have did the same comparison in a/b/c way and get a different conclusion. I'm guessing it has to do with video formats or the way they record it.
 
Last edited:
Administrator
Joined 2004
Paid Member
Hi thisusername,
I hope you enjoy this site.

I'm not a member of ASR, or many other public places. So you know, I've got almost 50 years in the audio industry from a design and service aspect. Also in the test and measurement industry. I worked in the recording industry and alsom on Neve consoles. I couldn't wind anything with my hands either! lol! Too darned old for that. Besides, machines are more consistent. I had Hammond wind my transformers for me and they do an amazing job.

The two issues I have with transformers are simply phase shift, and BH curves. Really good input transformers don't have these issues nearly as badly as cheap ones. Capacitors don't have these issues, choosing dielectric wisely avoids other distortion mechanisms. That doesn't mean you can't get really low distortion from a transformer coupled circuit, but it will not be as wideband.

Some things are designed to make sound (choice of recording console or effects unit), others to reproduce sound. Personal taste is fine with me. But high fidelity reproduction is rather specific and is an exacting science from my viewpoint. It is certainly something that can be measured and defined. I don't really care what technology is used to get there.

As soon as you say "sounds good", well you are into personal opinion and my hands are off! I just want to help people avoid wasting money or damaging their equipment.
 
Member
Joined 2010
Paid Member
Hi Tony,
Well, the condition of the original parts comes into your experience. Also what is buried in your head, ideas etc. Confirmation bias looms large here.

Well, we actually do. The high end crowd doesn't like the answers! lol! Measurements today actually do correlate to what people hear. But you have to use children or women who don't read the audio press to get subjects with clear minds.

On top of that your auditory memory isn't good except for gross changes. What is in your head makes a bigger difference.

You know what? Good engineers and technicians always listen to the equipment as well as measure it. ...

Some of us were looking at the residuals from the monitor output of THD meters (me for one) early on. The HP 3580A was an audio spectrum analyzer available in the late 70's I think (I have two). Better instruments followed and our understanding of what mattered grew rapidly. An oscilloscope can't show much on it's own, and that plus the needle on a THD meter was what the subjectivist crowd points to. To "get it", you have to see what we have now. For that you have to spend a fair amount of money and invest in training (years). I show clients and they are amazed.

I'm sorry, but today the measurements are a science. Interpretation is partly experience and an art to know what is important, but it is all laid bare and visible.


Me thinks you need to talk to the likes of Nelson Pass a bit, or read his writings!

I've spent most of my career in electronic/computer/metrology R&D labs. By now, I'm ready to retire... so that's something like, hmm... 48 years.

Measurements may not correlate to sound, yes, but why? Nelson himself notices that amps that measure well don't necessary sound well. Why is that? Why is it that we like negative 2nd order harmonic distortion? Huh?

We can say that by empirical methods we know that such is desirable and so we can design products that will behave that way and then we can measure them to ensure we have the final production product... but we STILL DON'T KNOW WHY...

If we knew WHY, then a device that measured "well" per those models would also sound well.

In essence, based on empirical analysis we can determine that a device needs to measure worse in certain ways - contrary to our current scientific model- in some ways because it will sound more realistic.... but we still don't know why... we can't set out via pure engineering at a desktop and design an audio device to sound good based on the current cognitive scientific model.

Why?

Because we don't know WHY something sounds good.

Hence, we have to introduce empirical experience. which you yourself have agreed to as important.

However, I'm neither an engineer nor a mathematician.. I'm a physicist so I have a different take on things. I can program and manipulate all kinds of electronic measurement devices but ultimately I know that it's all a model and likely to change next year.

Today's measurements are indeed NOT a science, they are an engineering discipline. Science is a very different beast. Engineering is following the cook book... Science is developing the models that will eventually yield the cook book.

BTW, my hearing memory is rather good, or rather my ability to recognize a reasonable facsimile of a live acoustic performance, since I've spent quite a bit of time doing mixing and PA (early 80s) and going to classical performances ( since the early 80s... ). Also, for many years my son had a garage band... I know what a drum kit or a good Marshal stack with different kinds of guitars sound...

So, when I tell you that the teflon caps made a difference it's an educated observation: The soundstage of well known recordings, via nothing else but the upgraded preamp changed, became more realistic: instruments sounded more like the real thing in an orchestra hall... or the drum kit and guitar in a rock recording will become clearer: the metallic nature of cymbals and the leading edge of brass instruments -in particular- become very realistic.

If I couldn't detect such things, then I might as well buy a Bose Acoustimass and join ASR.

I suggest you grab a good analog front end and a P3.. and then tell us why it sounds better than a 24/96 Tidal source via a good DAC ( currently an Android bit perfect over USB-OTG with a Burson Play and V6 opamps or RME ADI2 Pro RS). I don't know that we understand why the worse measuring system sounds more accurate to the real thing.

Perhaps you might also do some listening with components that fall into the mumbo jumbo World according to your current orthodoxy... who knows, you might hear the difference too? And for all we know, the psychoacoustic models might be modified to allow for such.

Perhaps the processing of the AD/DAC - following the current models - introduces distortions that we can not measure but that affect the realism of the signal... whereas the worse measuring system, by not adding AD/DAC artifacts actually sounds more realistic.

The same goes for other components in the chain that the model says should not matter, like the aforementioned teflon caps, but in reality they do.

You see, I'm neither tied to engineering orthodoxy nor High End mumbo jumbo... I actually know electronics, measurement techniques and the sound of real instruments...

That's why I'm in this site.
 
Last edited:
  • Like
Reactions: 1 users
Disabled Account
Joined 2002
Hello,
Most " sponsors" here are only here to make us waste money on their latest updates that is said to be a giant leap ahead.
I know loads of people with knowledge of music having a very basic soundsystem and enjoying listening to music much more than the average audiophile who is listening to sounds and has no clue about music.
Kind greetings, Eduard
 
  • Like
Reactions: 1 user
Administrator
Joined 2004
Paid Member
Hi Edward,
Believe me, I listen to music. I deal with live sound and recording studios, plus multiple very good systems at home. I don't listen to sounds, I listen to music.

Hi Tony,
Nelson follows his curiosity, he'll tell you that.

Your exact audio memory isn't that good, and what is in your head has more of an impact than what you heard. Human beings are terrible test instruments. I worked in a cal lab for electronic equipment repair and calibration following ISO and Guide 25 standards. I'm pretty good at that.

Why do imperfect things sound the way they do? Because of their response to stimulus, and some interact with speaker loads differently. Some don't so much. You can tell a great deal from their spectrum response, just two tests at various power levels. A single tone (1 KHz being standard) and a two tone test (19 and 20 KHz standard). Those pretty much tell you everything about a system.

Bit perfect? You know better. Streaming services often send as SIP, no error correction. The DSP only makes it conform to valid data. CD players are not bit perfect, I've been involved from the first consumer units, also for digital tape in studios. Block Error Rate (BLER) exists. Do you know what the actual name for the DSP chip in CD players is? Error detection, correction and concealment. They dropped the last bit of the description and hid the C1 and C2 error flags. If you get a music file as a digital file transmitted the way data is sent, then errors are detected and corrected.
 
Member
Joined 2010
Paid Member
Hi thisusername,
I won't argue it, but you have some misconceptions.

Audiophiles generally don't understand electronics or signal processing. Nor do they understand reliability or even basic measurements. Couple that with how your mind will affect what you think you hear, no wonder there is disagreement. On top of that, beliefs, just like religious arguments. Throw into the mix people who will advance ideas for commercial purposes and it's a wonder the average person hasn't given up entirely ... oh wait! They have.

Cagey, huh? Using the disclaimer "generally".

I'd say you are making this argument in the wrong forum.

In DIYaudio you meet people who are audiophiles and understand electronics... there's very little ASR or snake oil.

Indeed, I think DIYaudio is the most sane, educated audiophile community in the English Speaking world.
 
Member
Joined 2010
Paid Member
Hi Edward,
Believe me, I listen to music. I deal with live sound and recording studios, plus multiple very good systems at home. I don't listen to sounds, I listen to music.

Hi Tony,
Nelson follows his curiosity, he'll tell you that.


Why do imperfect things sound the way they do? Because of their response to stimulus, and some interact with speaker loads differently. Some don't so much. You can tell a great deal from their spectrum response, just two tests at various power levels. A single tone (1 KHz being standard) and a two tone test (19 and 20 KHz standard). Those pretty much tell you everything about a system.

Bit perfect? You know better. Streaming services often send as SIP, no error correction. The DSP only makes it conform to valid data. CD players are not bit perfect, I've been involved from the first consumer units, also for digital tape in studios. Block Error Rate (BLER) exists. Do you know what the actual name for the DSP chip in CD players is? Error detection, correction and concealment. They dropped the last bit of the description and hid the C1 and C2 error flags. If you get a music file as a digital file transmitted the way data is sent, then errors are detected and corrected.

The transmission of data has NOTHING to do... You introducing a red herring, a non sequitur. The differences are introduced in the AD/DAC process.

I was doing error correction for DARPA net before we had standards... I do know better about those algorithms. Did you ever see the pre ringing on square waves in early digital? Come on!

How about the filters and phase shifts used to process the analog signal before the ADC, huh?

You might be able to swing your credentials in front of others, but you can't do that with me because I was in the lab down the street, county, state at the same time as you.

And we're not talking about load interactions either. As I noted, when I noticed a difference when my preamps were updated to teflon caps, the rest of the chain was constant. I was very careful to be rigorous in the process. Therefore, the ffact is that based upon my own carefully controlled listening, the teflon capacitors made the presentation of the recording more realistic (ergo better).

BTW, let me give you two freebies.. (1) over the years, I've found out that the noise floor of a system is important in its realism. Also, (2) playing instruments at their natural sounds ( not lower nor louder ) with their natural dynamics create the most natural presentation of the event.

Yet, we don't even look into that? Just measure THD, IM etc... pffft...

Oh, Nelson is curious... I'm curious, DIYaudio is full of curious people... how do we make audio systems that sound good ( and don't cost an arm and a leg.. ). Aren't you curious?
 
Last edited:
Administrator
Joined 2004
Paid Member
Hi Tony,
The transmission of data to your streamer does impact the sound quality depending on impairments.

I was only letting you know what my experience is, that's all. That was in response to your similar comments. Okay?

I know all about filters, ringing and filter order. Both before and after the A/D and D/A processes. Phase shift is tied to filter type and order and the frequency of interest. EAsy stuff.

I was being general. In your case, I doubt very strongly if you could hear any difference between polypropylene and Teflon unless the Teflon was so big you picked up noise from elsewhere, or transmitted your signal into other areas of the circuit. That's a fact.

Noise floor is important, and if it is below your ability to sense it it no longer matters. We are talking about random noise, for periodic noise you have to go lower. However noise floor is a good indicator of the quality of a circuit.

The electronics hasn't any clue as to what originated a signal. Your brain does. Yes, as mentioned I listen to all phases of music reproduction, including before it is put on any media. But measurements do correlate to what people hear. That's if you have the right equipment and experience. Anyway this comment is more argumentative than anything as I already indicated I listen and measure. Want to get into what goes on in your brain? You're beyond what we can do with physical circuits.

I am curious. I design, enhance and experiment. every single day. It's my job. I love curious people.
 
Last edited:
Member
Joined 2010
Paid Member
Again, before making such statements and off hand reject a particular item.. why don't you go listen to it?

I heard the difference of the teflon caps and trust me, it's not subtle.

Now, if you want to call the Tice Clock or the mPingo disks as foo foo items, I will agree with you. I heard them. The former made no difference, the later did but it wasn't worth the price nor did the sound difference made the resulting presentation more realistic.

Just sayin'... keeping things academically honest and removing hearsay and qualified assumptions.
 
Administrator
Joined 2004
Paid Member
Hi Tony,
Clocks. I use an HP 5372A disciplined to the GPS standard off a timing antenna, the 10 MHz standard runs through an HP 5087A distribution amp (and it has been serviced). I studied DAC clocks among other things. I put it in holdover to run tests. Most of my lab runs off the same clock signal.

I have in fact listened to the Teflon caps compared to Polypropylene and others. Depends a lot on the circuit. I have listened to these things. I don't know if you missed it, but I do "as found" measurements as well as "completed" measurements. Same for listening. I do not assess my own work, I have people who do that. They never know if I have even done anything - so it keeps them honest.

I'm running on practical experience Tony. I am very careful about what conclusions I come to. I run tests several times on different days to ensure they are reproducible.
 
Member
Joined 2010
Paid Member
Look,

(1) Measurements do not reflect the why things sound good. This is because our psyachoustic models are incomplete. If they were, we'd have Perfect Sound Forever.

(2) The accuracy of the clock in your lab is a non sequitur. Repeatability over several days with measurements are irrelevant if you maintain a standard laboratory environment. Would you tape an IC design to a fab if they had inconsistent results day to day?

(3) I strongly believe what I hear, not what I'm told. Sure, if someone with whom I share some common experiences publishes some analysis I will believe them... but I won't put my money into it until I hear it for myself.

Do you buy a car based on a review on Car and Driver? Or do you go test drive it?

Whatever, I made my point, based on not conflating science with engineering. It's very simple: I heard the difference in my own system. I don't have to justify or defend my decision, that's the part of your argument that fails.

I don't deny you didn't hear a difference, but why then must you deny (or invalidate or belittle or ridicule) that others have heard it?

QED.
 
  • Like
Reactions: 2 users
Administrator
Joined 2004
Paid Member
Fine Tony.
1.) Measurements actually do point to why things sound good or bad. However, personal opinion isn't something you can measure, and it will change. Therefore there is no point in measuring it as it's a variable.

2.) I check to make sure my cables and equipment are in order and that I didn't make a mistake. I'm also then sure anyone can repeat my experiment and report on what they find. It's just simply good practice.

3.) Me as well, or we wouldn't be having this discussion. I have never once denied someone else hasn't "heard" a change. However there is much that can interfere with an honest appraisal of what exactly they did hear. Not once have I engaged in belittling or ridiculing what others have heard. However it is helpful to interject some reality and have them think. Without this, audio becomes the wild west fashion show so beloved by those very folks who created the situation in the first place. There is profit in confusion in case you haven't noticed.

One thing I have noticed over decades. If someone spends a lot of money to improve something, they most will definitely see or hear an improvement. It is very rare for a human to admit they wasted money doing something. The more money they spend, the more invested they are in the changes "they heard". It's a mind game and the victim has no possible way to see it. Now that is physiology (which I also studied).

Of note, I have zero to sell. No name to make, nothing. I just pass on what I have learned, and I didn't learn these things in a vacuum.

I have no problem with any disagreement between us at all. It also has no effect on my appreciation of you or anyone else as a person.
 
Only a fool discounts measurements and one can be just as foolish putting too much faith in them.

If you cannot discern the difference a good capacitor makes consider yourself lucky. You have saved yourself lots of money that would have been wasted in your setup. It is as simple as that.
 
  • Like
Reactions: 1 users