Selecting Capacitors to Minimize Distortion in Audio Applications

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
The one which makes the music have more pulling power, lowers listener fatigue and makes listening more pleasant overall. It’s art, I suppose
Maybe figure out what that is, then figure out how to make more of it. ;)

Those don't sound like objective metrics. While lots of sounds are unpleasant, and some distortion is associated with listener fatigue. But you did say it is art which is considered subjective.

JR
 
If a customer or artist, who knows and cares nothing about capacitor dielectric, hears a difference and picks the one that measures worse, are they incorrect?
 
Dramatic failures is a bit of an under-statement. Fire hazard liability risk is more accurate.

Really nobody is using regular tantalum capacitors for modern designs these days, except in very niche applications. If you use them in a commercial design, you'd better make damn well sure you understand their failure modes in the context of your application and you cover it thoroughly during product validation. Otherwise, make damn well sure you've got good lawyers on hand!


Frightening stuff. It's shocking to think we might actually have been taking our lives in our own hands every time we turn the blasted things on. And some folks have the nerve to talk about suffering for their art? Sheesh!

I'm off to check the latest govt. stats regarding hazardous fires caused by flame throwing tantalums in Neves...


P.S. ;)
 
Last edited:
They don't Brian. These were/are used as audio coupling caps.

I've seen far less failure rate of the tants in Neve than I have of the electrolytic caps.
 
Dramatic failures is a bit of an under-statement. Fire hazard liability risk is more accurate.
more like spectacular failures with light show, but mainly when used across a power rail without current limiting.
Really nobody is using regular tantalum capacitors for modern designs these days, except in very niche applications. If you use them in a commercial design, you'd better make damn well sure you understand their failure modes in the context of your application and you cover it thoroughly during product validation. Otherwise, make damn well sure you've got good lawyers on hand!
While not these days, and I've probably shared this story too many times, but the last time I designed in a tantalum cap was back in the 80s, when Popular Electronics offered me a cover story for a kit article if I became a CX licensee and cranked out a kit design overnight. CX was CBS's ill fated noise reduction system for vinyl. I used exactly one tantalum capacitor in the playback expander side chain time constant circuit, because I observed that the encoder/compressor (designed by Urie) used a tantalum in their side chain. One notable characteristic of tantalum is lousy dielectric absorption, which mainly matters in sample and hold circuits, or side chain time constants.

It doesn't matter all that much for side chains, but I tend to over engineer stuff...

JR
 
Maybe figure out what that is, then figure out how to make more of it. ;)

Those don't sound like objective metrics. While lots of sounds are unpleasant, and some distortion is associated with listener fatigue. But you did say it is art which is considered subjective.

JR

Well, when it comes to intelligibility, sound quality etc. people like the engineers at Neumann, I have been told, spent a lot of time listening.

How do you quantify the 'right amount' of presence peak in a capsule, or the 'right amount' of directional high frequency response?

Assuming that every time you try and improve (remove) distortion and noise figures, there is a tradeoff and a change in the overall sound, it's always a (personal) choice as to what is the best compromise, IMO.
 
The SSL9000 preamp board that Gustav sells uses a Tant in the meter driver circuit. I always wondered why.
Could be for timing functions: A property of the tantalum capacitors is that they are relatively stable over time regarding capacitance, so they can be used in timing functions that shouldn't change too much over time. This also why they have a place in e.g. the GSSL sidechain

/Jakob E.
 
Well, when it comes to intelligibility, sound quality etc. people like the engineers at Neumann, I have been told, spent a lot of time listening.

How do you quantify the 'right amount' of presence peak in a capsule, or the 'right amount' of directional high frequency response?
market research?
Assuming that every time you try and improve (remove) distortion and noise figures, there is a tradeoff and a change in the overall sound, it's always a (personal) choice as to what is the best compromise, IMO.
less noise and less distortion is generally considered an audio path improvement.

There have been psycho-acoustic research findings that the presence of HF noise can be perceived as higher HF response than actual.

The use of added distortion as a vocal enhancement (exciter) effect is pretty old news. I recall decades ago when these were first used and overused, that I would try to tune in my FM radio stations better, to clean up the audio signal.

JR
 
The graphs don't tell the whole story. The TI INA1650 high-CMRR dual line receiver evaluation board uses 10uF 35V X7R 1206 ceramic caps on its inputs with excellent results. High impedance and low cutoff frequency can allow the use of high-K ceramic caps while still providing low distortion.
 
I know Sphere used Tants as decoupling caps on their SPA-62 op amps. When they went, they went hard. AT least they had current limiting bulbs to save the supply...usually.
 
The graphs don't tell the whole story. The TI INA1650 high-CMRR dual line receiver evaluation board uses 10uF 35V X7R 1206 ceramic caps on its inputs with excellent results. High impedance and low cutoff frequency can allow the use of high-K ceramic caps while still providing low distortion.
Indeed.
The choice of making the LF cut-off frequency at 0.03Hz results in more than 50dB attenuation for the voltage across the cap.
Even at max input voltage (+/-18V p-to-p), the voltage across the cap at 20Hz is less than 60mV p-to-p.
My RoT for electrolytics is factor 100, which usually results in LF cut off at 0.2Hz.
Here, the choice of a factor almost one order of magnitude higher is dictated by the higher non-linarities of high-K ceramics. These non-linearities result in drastic decrease of nominal value with increasing instantaneous voltage, which results in increased voltage across, for even more distortion.
 
The use of added distortion as a vocal enhancement (exciter) effect is pretty old news. I recall decades ago when these were first used and overused, that I would try to tune in my FM radio stations better, to clean up the audio signal.
Today, the use of dirtifiers seems to be an obligatory step for compensating the lack of substance of many tracks. Such substance should be built in the studio, but how many recordings are actually done in a studio today?
Except for electric guitar (with many restrictions), distortion is not an acquired taste for me; whenever I hear it I think of it as something that's been added and stands out as being artificial.
 
Well, when it comes to intelligibility, sound quality etc. people like the engineers at Neumann, I have been told, spent a lot of time listening.

How do you quantify the 'right amount' of presence peak in a capsule, or the 'right amount' of directional high frequency response?
These are "linear distortions", they don't add harmonics to the spectrum. Just like using the tone control on a phonograph, it's perfectly legitimate to "tune" these effects.
Assuming that every time you try and improve (remove) distortion and noise figures, there is a tradeoff and a change in the overall sound, it's always a (personal) choice as to what is the best compromise, IMO.
Indeed. The German designers submitted their products to the likes of Karajan, Furtwangler, Klemperer for final tuning, even if it was not perfectly aligned with their objective measurements.
Neumann did not sell mics to their employees, but to institutions that had to please musicians.
 
Back
Top