Have a look at schematic section below from a mid-80s CD player (ADC, model CD-100x).

Note C530 and 540. 2.2uF/50 v polarized electros. These are very low-cost caps if you go for electro, polarized. TTBOMK, they are used to block DC. But I'm not sure why the 50v rating. Perhaps, that's all that's avail. in this low-uF value range.

I want to improve performance. I have some high-quality Wima MKPs, in same uF, and voltage, but they are not polarized.
I also have some $$$, high-quality 35v Vishay TANTALEX
https://www.vishay.com/docs/40015/150d.pdf
If a one of these tant's is overvoltaged/abused, will it fail as an open (that's preferable!) or will it short (not good)?

BOTTOM LINE: What might go wrong, in a worst case scenario, if one does not follow the OEM schematic (below)? E.g., don't want to blow anything up that the CD player is connected to, like a preamp or receiver input section. Or, at least, minimize noise and hum.

Thanks!

https://i.postimg.cc/SR9YZCh2/Screenshot-at-2024-03-22-19-51-17.png

 

The Wima caps are good parts. The Ele caps must be put in, in the right direction, the Wima do not care.

 

The Wima caps are good parts. The Ele caps must be put in, in the right direction, the Wima do not care.

You didn't answer the questions.
Why did the manuf. use an electro polarized? The manuf. could have used a NP/BP electro., yes? Why or why not?
And why that "high" voltage rating for a line level cd player output?

 

The polarized capacitors were used because there is DC present at that point. Not a whole lot, but some. and in some instances it is wise to not use the bare minimum value of voltage rating.
I CERTAINLY would not choose to risk having DC present on the output of a pre-amplifier that might be used to drive a direct coupled amplifier. Of course, I would never choose to own a direct coupled audio amplifier. In addition, there may be some other characteristic that is not so obvious. And why choose a non-polarized cap when there is some DC present. AND it may also be that there were a lot of that device on hand and it was the required value. Besides all of that, the designer may have known something that we don't know.

 

The polarized capacitors were used because there is DC present at that point. Not a whole lot, but some. and in some instances it is wise to not use the bare minimum value of voltage rating.

The schematic below is from another CD player of the same vintage. Opamp-based output, and the manuf (NAD/Toshiba/etc) chose to create an output NP capacitance using two polarized electro caps back to back. Note the combined value and voltage.

 

Never use Tantalum capacitors on audio - noisy old things.
Nothing wrong with a polarised electrolytic on audio, as long as it has a DC voltage across it, and it is not used as a filter (i.e. 1/(2πRC) is not within the audio range.
The first question to answer is whether the op-amp that is driving the output is contributing more or less to the output noise and distortion as the capacitors might.

 

I recall a time when it was usually correct to replace the "RUBYCON" brand electrolytic capacitors in Sony tape decks because they were electrically leaky. This was often in warranty repairs, and it seemed that perhaps there was a serious quality issue, ignored by the purchasing people. Really, there is sometimes more to design decisions than just the circuit requirements.

 

You will not notice an improvement in performance by replacing the output capacitors unless they are defective (dried out from old age?).

 

C530 and C540 at the low voltage audio are not critical. 10μ/16V aluminium electrolytics will be fine.
Don't use tantalum as already stated.

 

Undoubtedly there are audiophiles who can hear the distortion caused by the two polarized caps in series, while most engineers are unable to measure it or even detect that added distortion.

 

A single non-polarised electrolytic works better than two polarised ones back to back.
It's all in here:
https://linearaudio.net/cyril-batemans-capacitor-sound-articles
Cyril Bateman was the engineer that could measure it! And @MisterBill2 , you're quite correct, most engineers would be unable to measure it.

 

The ESR and Vloss of those orig. 1985-made electro caps (2.2uF, 50v) is ....

... and what I've replaced them was is ...

( ESR and loss is much lower on the yellow Visahy polys) ...

BTW, none of y'all's 'dvice wuz useful. Typical, confuzed AAC feedback.

The clues to answering the queries posed in the OP -- about why that 50v rating voltage?, why polarized electro in that location?, and why an electro (not poly)? -- can be gleaned from the schematic wisdom.
I'll let y'all try to guess ... and will supply the ANSWERS in the back of the book in due time.

 

The output coupling capacitor is 2.2uF. It is a highpass filter with the input impedance of the input of the amplifier.
If the amplifier input is very low at 10k ohms then harmonic distortion of extremely low frequencies below 7.3Hz are increased a little because the highpass filter attenuates those very low fundamental frequencies.
Editted

 

Interesting. CX20152 is a multiplexed 16-bit DAC. That means that there is a lot of phase shift at high frequencies due to the multiplexing process. The right-channel samples are delayed by 11us, which is 80 degrees phase shift at 20kHz.
The Philips TDA1540 although only 14-bit was always thought to sound better because it had two separate DACs and therefore no phase shift at high frequencies.

 

Interesting. CX20152 is a multiplexed 16-bit DAC. That means that there is a lot of phase shift at high frequencies due to the multiplexing process. The right-channel samples are delayed by 11us, which is 80 degrees phase shift at 20kHz.
The Philips TDA1540 although only 14-bit was always thought to sound better because it had two separate DACs and therefore no phase shift at high frequencies.

The 4053 multiplexer takes care of much of the phase issues if used correctly.



The followup to the TDA1540, the coveted tda1541 , could be used both in dual and multiplex modes. It's a great dac in either case.
One thing that dual dac chips introduce is phase distortion because of that dual use: after decoder or oversampler, the signal now has two (dual) complex paths with trace-length issues, etc.

BTW: I have heard some of latest-n-greatest dac technologies, including ESS and AKM delta-sigma, and discrete ladder R2R. All held equal, none sound as good as 1st-4th gen devices from the 80s (pre-1-bit era).

EDIT:
The CX20152 can be used in dual mono. One per channel, like TDA1540/TDA1541.
You'll need two CX20152s. See the official Sony datasheet:
https://manuals.lddb.com/DACs/SONY/CX20152.pdf

 

The ESR and Vloss of those orig. 1985-made electro caps (2.2uF, 50v) is ....

... and what I've replaced them was is ...

( ESR and loss is much lower on the yellow Visahy polys) ...

BTW, none of y'all's 'dvice wuz useful. Typical, confuzed AAC feedback.

The clues to answering the queries posed in the OP -- about why that 50v rating voltage?, why polarized electro in that location?, and why an electro (not poly)? -- can be gleaned from the schematic wisdom.
I'll let y'all try to guess ... and will supply the ANSWERS in the back of the book in due time.

Ialready suggested that the polarized was because there was a DC bias voltage on the one side.

 

Ialready suggested that the polarized was because there was a DC bias voltage on the one side.

For the new poly caps , I measured 0.05vdc at output jack and 0.5vdc before cap.
The polarized electrolytic at that 50v value is what is cheaply available for 2.2uf. The opamp rails are +-8v.
Those cheap electros are also small and fit nicely onto that budgeted pcb area.