Electrolytic Capacitor Characteristics

[steve]

Below is the graph showing the characteristics in electrolytic capacitors. Several points to be had, which
you already know, but I thought some proof would be helpful. (The graphs are from the article "Picking Capacitors"
by Walter Jung and Richard Marsh.)

The first characteristics are the low frequencies at which various (ufd) electrolytic capacitors become
non linear, and inductive. With newer, high frequency electrolytic capacitors, I am sure the non linearities will occur
at higher frequencies, although the quality will still be lower than films.

The second is that quality varies vs manufacturer.

Third, the bigger brand names does not necessarily mean better specifications. Any input is welcome.

Fourth, notice the difference between electrolytic capacitors and film capacitors by comparing the different graphs.
Compare the frequencies of the two graphs (on the X-axis).

Cheers
steve

[Folsom]

The only real way to do it is by ear.

[HAL]

You might look at organic polymer electrolytic caps.  Designed many years after the Jung/Marsh article from 1980.

Low ESR for power supplies and extended frequency response as they are used in switching power supplies.

[steve]

The only real way to do it is by ear.

I agree. Since most individuals don't have anyway of testing, listening is about the only way. I guess one could
check via the discharge/high impedance meter method, but the DA is pretty much known.

I think it is helpful that in the graph, one of the big names, Mallory, faired worst than lesser known brands.

As far as I know most or all of their manufacturing equipment was sold to Antique/CE distribution. Don't know
about the materials.

cheers

steve

[steve]

You might look at organic polymer electrolytic caps.  Designed many years after the Jung/Marsh article from 1980.

Low ESR for power supplies and extended frequency response as they are used in switching power supplies.

I agree Hal. The graph does not cover the higher frequency types. Those are still a long way from films though imo.

I think it will help customers/diyers to understand that typical electrolytic capacitors affect the sound, especially as the capacitance (ufd) increases. Solid state amps tend to use huge electrolytic capacitors so both DF/esr and DA are negative factors.

cheers

steve

[P.I.]

The only real way to do it is by ear.

Experience and the primary biological auditory transducer is the deal, for sure.   

[steve]

So how does one check the electrolytic capacitor by ear, for the best sound, when the circuit design is flawed?
If the design is flawed, the best we can hope for is a flawed capacitor to "patch" the flawed circuit as best we can.

Interesting problem.

cheers

steve

[Folsom]

Not a problem.

Define flawed.

Your description about patching sounds like coloration and aberration expectations from the capacitor.

[steve]

Not a problem.

Define flawed.

Your description about patching sounds like coloration and aberration expectations from the capacitor.

Not accurate in absolute terms but artificially flavoring the sound.

Even most poly caps are not close to accurate, so how can an electrolytic cap be accurate when its
characteristics, such as DA, DF/ESR, and Inductance are many if not hundreds of times higher than any
polypropylene capacitor?

cheers

steve

[Folsom]

Are you saying the circuit is flawed because you're using the electrolytic capacitor to begin with?

[steve]

Are you saying the circuit is flawed because you're using the electrolytic capacitor to begin with?

I use all polypropylenes, separate power transformers, in my monoblock amps 2 power supplies. The
3rd power supply, the final output tube high voltage power supply, a polypropylene with electrolytics. There are
other techniques of getting around problems, at least to some degree.
   
My preamplifiers 6 stage filering circuit, final stages all polypropylenes with voltage regulation.

cheers

steve

[Folsom]

That's easy with tubes. But aren't tubes inherently flawed? 

[steve]

That's easy with tubes. But aren't tubes inherently flawed? 

Evidently not as I was able to run some sophisticated listening tests, tweaked, and determined my
PP monoblocks output sounded the same as the input, kinda like first generation tapes. Can't say that for
SETs or SS designs. NOS tubes are horribly expensive for the results obtained.

Distortion can be a problem with tubes. SETs are yesterday's news, high distortion; one can design much better
with PP, wired triodes. Wish tube manufacturers would manufacture high power idht cathode type triodes. Not that
difficult.

cheers

steve

[Folsom]

I have no issues with SETs.

One has to ask how much they care about distortions figures that are less than 1% and how important the memory distortion is...

[steve]

I have no issues with SETs.

One has to ask how much they care about distortions figures that are less than 1% and how important the memory distortion is...

If that is your cup of tea, enjoy it. I kinda wanted to keep this string on capacitors if possible. Maybe start a new string?

And 1% at what watt output? Most competently designed PP amp will have less distortion than any dht amp for a given wattage output, and better frequency response. Below are distortion figures for a 300b. Please note, the distortion only becomes very low with very low wattage output. Of course, one could add negative feedback to lower it. Many do with PP designs and OTL designs as well. One must also addressed the high distortion of the driver/input tubes.

The other graph is the intermodulation products at maximum power output of a 3/2 tube characteristic curve (IMD).
(Courtesy Eimac.) Notice the 3rd, 5th and 7th orders.

-20db means approximately 10% distortion, -40db equates to 1%.

Other tubes such a 45s and 2A3s have so little power output. 845s etc have problems with power supplies and driver/signal stage distortions.

cheers

steve