Electrolytic vs film caps for smaller values.

[JohnRoberts]

I just built a fuzz tone pedal and compared a Panasonic FC electrolytic to a WIMA MKS and preferred the electrolytic. I didn't expect to. MKP wasn't going to fit. And the uF wasn't critical, just avoiding a low end rolloff. In some circuits you can hear a difference. Not better, it's a preference, the character of the sound is sometimes different (in my experience).
Electrolytics are limited vs film as they want a voltage difference on the sides. If the circuit doesn't have a voltage difference you need to order BP and then I would go for film unless the uF is really high

when doing listening tests like that it is important to make sure the capacitance values are identical.

JR

 

[dmp]

They were both 1uF with whatever tolerance, 20% probably.
The difference wasn't in the low end, it was a difference in character. But maybe I only thought I heard something. Doesn't matter, it was just a pedal I was building.

 

[Mr. Moscode]

What about possible antenna effects of physically larger caps?

For example, I've not done any tests, but I replaced some tiny .47uF tantalums with big 2X1 inch MKP Wimas. I get nervous sometimes that they might be picking up something . . . .

The "combine input" cap at the input to the master bus channel here:

Sound any better? You can always scope it to see if it's picking up QRM. Not a fan of tantalums in the AP (audio path).

 

[CJ]

Lab work in audio is usually confined to sine wave testing, I have never seen a spec for a composite waveform ie; music.

Marantz also used tants, input caps to the phono preamp and also on the tone board. I bet they did not want any leakage hitting the tone controls in the case of the tone board. Most people rip out the tants and install Wima's.

 

[JW]

Yeah, it does sound better. The polyester ones aren't much bigger than the tants, but the polypropylene caps sound even better in my opinion (that is based admittedly on listening tests I did a long time ago in the pultec high boost band)

 

[JohnRoberts]

As we veer off into the "sound" of different capacitor media, this is a very old well inspected topic. I wrote about capacitors in my circa 80's "Audio Mythology" column. There are many good articles written on the subject (beside mine) and discussed here over the years.

JR

 

[Tubetec]

Ive swapped out smaller value electrolytic coupling caps in a few stomp boxes , usually values around 0.47-1.5uf ,
I definately prefered the sound of the foils , I didnt bother digging too deep trying to find out the reasons , the lytics were probably 20 or more years old ,so probably past their best .

 

[rock soderstrom]

The capacitor sound is a shy creature. I wanted to believe in it, really. I made an effort and even built myself a comparison rig with which I could hear some capacitor types comparatively. Blind of course. It was a headphone amplifier on OP-amp basis.

The result was more than disappointing. I could hear (almost) no differences. What I could guess was not reproducible. My favorite was a pair of pink electrolytic capacitors removed from a power supply of an Italian organ after 40 years of use. These had entered the race as underdogs against high-end capacitors and expensive NOS vintage caps.

My conclusion was, if NFB is in the game the capacitor sound does not matter. Perhaps in extremely reduced environments such as in a tube microphone or preamp. Maybe. But otherwise I choose (besides the technical values) capacitors by color.

PS: The subject of audiophools was raised here. Here is my absolute reading tip, I recently discovered. It is a review of 5654/6AK5w tubes. Unfortunately it is only in German, really too bad. A linguistic gem.

https://www.audionist.de/2019/01/05/5654-tube-rolling/

 

[JohnRoberts]

Ive swapped out smaller value electrolytic coupling caps in a few stomp boxes , usually values around 0.47-1.5uf ,
I definately prefered the sound of the foils , I didnt bother digging too deep trying to find out the reasons , the lytics were probably 20 or more years old ,so probably past their best .

Nobody claims electrolytic caps sound good, but replacing a very old one that may be suffering from low capacitance can sound better just because of the benefit from correct capacitance.

The capacitor sound is shy creature. I wanted to believe in it, really. I made an effort and even built myself a comparison rig with which I could hear some capacitor types comparatively. Blind of course. It was a headphone amplifier on OP-amp basis.

The result was more than disappointing. I could hear (almost) no differences. What I could guess was not reproducible. My favorite was a pair of pink electrolytic capacitors removed from a power supply of an Italian organ after 40 years of use. These had entered the race as underdogs against high-end capacitors and expensive NOS vintage caps.

My conclusion was, if NFB is in the game the capacitor sound does not matter. Perhaps in extremely reduced environments such as in a tube microphone or preamp. Maybe. But otherwise I choose (besides the technical values) capacitors by color.

PS: The subject of audiophools was raised here. Here is my absolute reading tip, I recently discovered. It is a review of 5654/6AK5w tubes. Unfortunately it is only in German, really too bad. A linguistic gem.

https://www.audionist.de/2019/01/05/5654-tube-rolling/

Indeed NF generally forces response to follow the transfer function as defined by the negative feedback network, but of course capacitors used in NF paths can and will influence that transfer function. With electrolytic capacitors any non-ideal behavior is typically a function of terminal voltage so the pole frequency of the RC can impact the audibility. In other words try not to use electrolytic caps for RIAA EQ poles occurring in the midrange.

JR

 

[CJ]

In industry if your manager says "do you hear enough difference between the film and the lytic to justify the cost increase" then he or she is probably going to want to see some proof in the form of scope shots or distortion numbers, like my old manager who he would always say " I m from the show me state, where is the evidence"

I am sure that the human ear can process information in ways that test instruments have yet to measure.

 

[Newmarket]

Lab work in audio is usually confined to sine wave testing, I have never seen a spec for a composite waveform ie; music.

Marantz also used tants, input caps to the phono preamp and also on the tone board. I bet they did not want any leakage hitting the tone controls in the case of the tone board. Most people rip out the tants and install Wima's.

Square Waves are the obvious option. There's a whole thread on that here.

 

[Ricardus]

I am sure that the human ear can process information in ways that test instruments have yet to measure.

Well, you'll get a Nobel or something when you can demonstrate this, but until that day, this is magical thinking.

 

[CJ]

Yes I remember, along with all the different types of white noise, pink noise, weighted this way and the other.

The brain can also dub in perceived differences, I can not remember all the times I thought that part A sounded better than part B, only to find out that the solder joint had broke and thus I was hearing things. In fact, even the stomp boxes I build seem to sound better after I put them in a nice shiny box. Which reminds me. What ever happened to that Shiny Box audio company?

 

[Tubetec]

Remember what they did to the people who first suggested the world was round Ric ?

What about Arthur.C.Clarks science fiction that turned into science fact years later ,
The military industrial complex systematically discredited him , but quietly used his ideas for all they were worth on the QT.
https://lakdiva.org/clarke/1945ww/

 

[Tubetec]

Wernher Von Brauns V-2 rocket , re-badged with a few extra winglets added became the Hermes A-1 , the UK and the Soviets also took a piece of that pie to kickstart their own ICBM program .

 

[Mr. Moscode]

I've not seen anyone disclose their test setup for listening for component effects.

When you listen for the sound quality of a component you're not listening to just that part but the whole chain.

You may not hear 1 film cap in a system loaded with 'lytics. Subtle errors compound and add up to obscure.

Of course it depends in how critical to the beholder the sound is.

 

[JohnRoberts]

I've not seen anyone disclose their test setup for listening for component effects.

When you listen for the sound quality of a component you're not listening to just that part but the whole chain.

You may not hear 1 film cap in a system loaded with 'lytics. Subtle errors compound and add up to obscure.

Of course it depends in how critical to the beholder the sound is.

Mine is simple, I don't design by ear.... (except perhaps for EFX and I don't that anymore).

I also don't argue with people on the internets about things they claim to hear. It can't be proved to any statistical certainty without rigourous time consuming double blind testing.

JR

 

[abbey road d enfer]

What about possible antenna effects of physically larger caps?It depends very much on the node's impedance

Susceptibility to interference depends very much on the node's impedance and its nominal operating level, and the level at the source of disturbance.

For example, I've not done any tests, but I replaced some tiny .47uF tantalums with big 2X1 inch MKP Wimas. I get nervous sometimes that they might be picking up something . . . .

What is there to pick up, that's the question. How far, how strong is the possible source of disturbance...

 

[Newmarket]

I've not seen anyone disclose their test setup for listening for component effects.

When you listen for the sound quality of a component you're not listening to just that part but the whole chain.

You may not hear 1 film cap in a system loaded with 'lytics. Subtle errors compound and add up to obscure.

Of course it depends in how critical to the beholder the sound is.

Well it's pretty well established that distortion wrt electrolytics is highly dependent on the voltage across them. So for a typical pro audio dual rail scenario that is going to be only mVs. More if interfacing to a 9V ported fx pedal.
Fairly obvious to avoid them in filter / EQ circuits etc.
And stick to film or class 1 ceramics to minimise voltage (and temp) coefficients.

 

[57sputnik]

Panasonic is always a fine choice, as is United, Wurth, and Kemet.

I've used a lot of Panasonics back in the old days recapping MCI consoles.

Never bought United Chemi-Con.

Any experience with Visahy/BCC?