How about a capacitor discussion?

[Alex SE]

Polarized electrolytics like to be polarized. Very often however we see input coupling capacitors feeding an opamp or diff pair, and both sides of the electrolytic at zero volts DC. When audio signals, which are AC, are passing through the cap, (lets assume the + lead is towards the input) the negative part of the audio signal actually reverse biases the electrolytic! Reversing polarity on an electrolytic breaks down the electrolyte insulator inside, and decreases the capacitance. An audio grade capacitor is designed to act more like a non-polar electrolytic in these situations.

That said means leave it as it is? If there was polarized from beginning, replace it with polarized and stay on the safe side. Modern caps should not be worse than those used 30 years ago.
I haven't placed an order yet but will go thru the list and make some changes, UFG instead of UES.
Just had too much other things to do both at work and privately, and even worst, now I have emergency situation and have to fly tonight. Hope I'll be back in one month, so don't think that I've just left discussion I've started.

 

[vince666]

That said means leave it as it is?

i guess Nakdoc meant that, about the coupling caps (the ones subjected to zero volts DC), it's better to go with bipolars, which is also how I do it myself.
Since these caps are only subjected to the signal (which is AC), a bipolar cannot be reverse biased by the negative parts of the signal waveforms because it can work both ways.

 

[Alex SE]

i guess Nakdoc meant that, about the coupling caps (the ones subjected to zero volts DC), it's better to go with bipolars, which is also how I do it myself.
Since these caps are only subjected to the signal (which is AC), a bipolar cannot be reverse biased by the negative parts of the signal waveforms because it can work both ways.

I was thinking, as Nakdoc said "An audio grade capacitor is designed to act more like a non-polar electrolytic in these situations", why go for bipolars in a signal path then? How I understand is, if cap is in a signal path, connected in series, where 0V presented, the one can go for bipolar but:
For example, there is a capacitor is series (coupling) right after the input, so input signal goes thru capacitor in question. Feeding signal is say 500mV. Would you consider that as a 0V since that it is less than 1V?

 

[8991XJ]

Hafler DH-200 input cap is bi-polar.

I talked to a Nichicon engineer about their 'audio grade' caps. He said that there are differences which were trade secrets, materials of the electrolyte, maybe, the way the foil is laser etched and the insulation used between layers are my guesses. The data sheets indicate the leads are tinned copper and not 'lead free wire as used in other caps. The other cap leads can be picked up by magnets, the audio caps have copper guts seen when they are cut.

 

[Gepetto]

Try and use film caps where ever possible for audio coupling purposes. PP is the best film audio dielectric but may not be possible given size restrictions. Mylar (polyester) will pack more C into a smaller footprint and is a good second choice.

Polystyrenes are great but scarce as hen's teeth and are only generally available in very small capacitances.

 

[vince666]

I was thinking, as Nakdoc said "An audio grade capacitor is designed to act more like a non-polar electrolytic in these situations", why go for bipolars in a signal path then? How I understand is, if cap is in a signal path, connected in series, where 0V presented, the one can go for bipolar but:
For example, there is a capacitor is series (coupling) right after the input, so input signal goes thru capacitor in question. Feeding signal is say 500mV. Would you consider that as a 0V since that it is less than 1V?

when we talk about biased capacitors we mean DC biasing, as Nakdoc wrote.
when a cap is subjected only to signal, that's AC which goes positive/negative with the signal.

If a polarized audio grade cap is designed to act more like a non-polar one, it's just obvious that a true non-polar one will act like that even better, because it won't only act "more like a non-polar one" but it will act exactly as a non-polar one, doesn't it?

a good reason to use non polar ones in the signal path?
this is what Alex/A.N.T. advices and, for me, this is just a good reason in itself.

 

[WOPL Sniffer]

when we talk about biased capacitors we mean DC biasing, as Nakdoc wrote.
when a cap is subjected only to signal, that's AC which goes positive/negative with the signal.

If a polarized audio grade cap is designed to act more like a non-polar one, it's just obvious that a true non-polar one will act like that even better, because it won't only act "more like a non-polar one" but it will act exactly as a non-polar one, doesn't it?

a good reason to use non polar ones in the signal path?
this is what Alex/A.N.T. advices and, for me, this is just a good reason in itself.

HUH?????????????????????????

 

[Gepetto]

when we talk about biased capacitors we mean DC biasing, as Nakdoc wrote.
when a cap is subjected only to signal, that's AC which goes positive/negative with the signal.

If a polarized audio grade cap is designed to act more like a non-polar one, it's just obvious that a true non-polar one will act like that even better, because it won't only act "more like a non-polar one" but it will act exactly as a non-polar one, doesn't it?

a good reason to use non polar ones in the signal path?
this is what Alex/A.N.T. advices and, for me, this is just a good reason in itself.

Non polars are just 2 back to back polar electrolytics. You suffer the series impact of 2 caps instead of one.

 

[vince666]

Non polars are just 2 back to back polar electrolytics. You suffer the series impact of 2 caps instead of one.

As you might easily imagine, I am not expert in electronics then I simply try to read and apply what I read from those who are more expert than myself, then any hints/explainations are very welcome.

Anyways, the modifies in the AIWA AD-F880 suggested by Alex do imply non polarized caps as coupling ones in the PB amp (the deck had originally two polarized caps in anti-series at that place).
Same goes for modifies he suggested to improve the Technics RS-B565 by putting non polarized caps in the signal path (the deck originally had only polarized ones).
In both cases, the sound of the decks got improved.

So, are you trying to explain me that the above wasn't a good solution?

Of course, I am curious to learn some more about this... even because in the near future I'll need to recap a few decks then I'll be in need of choosing what to put in such decks.

 

[Skratch]

You have to be very careful when using flux capacitors

 

[vince666]

like this one?

 

[vince666]

Non polars are just 2 back to back polar electrolytics. You suffer the series impact of 2 caps instead of one.

about this point, i've tried googling a bit about non polars into audio path and tried various points of view on a few audio forums.

the most common points i've read around could be resumed with bipolars being better than using two polarized back to back and that they get lower distortion than a single polarized and also than two polarized back to back.

if with series impact you mean ESR, might it be that a proper bipolar is better than two polarized back to back where you simply add their own ESR?

of course, I am meaning good quality bipolars like Nichicon Muse ES (aren't these just caps designed for audio?) or also Panasonic SU bipolars which Alex surely likes (in fact, i knew about these just from a few of his posts).

 

[20tajk7]

Some Sony ES decks have several back to back polarized capacitors in the signal path, changing them to a single Nichicon Muse makes a really good sound upgrade.

 

[Alex SE]

Seems that capacitors are kind a sci-fi story for itself. Most of us want to push our gear to the limits which in many cases can give a headache just regarding caps.
I have recaped 3 of my decks, one of them 965 which have very big schematics. Replaced "some of" polarized by bipolars. Correctly? I hope so. It works and sounds nice and no smoke from it att all. Now when Im looking at SM, I'm not 100% sure if I should placed polarized audio caps on 3 or 4 places. Think I'll do so on the other 965. I mean, those were polarized from the factory, It can't be wrong to go for the same.

 

[BlueCrab]

Just follow Joe's advice and use film capacitors in the signal path. Most of the polarized electrolytic caps in your equipment are there to bypass the power supply, but I understand that this requires looking at the schematic and being able to discern which are which. Using a bipolar electrolytic in place of polarized one will generally work as long as capacitance and working voltage are adhered to, but I wouldn't want to do that where the cap is being used to bypass the power supply.

I do think too much is made of this. There are good reasons capacitors are made in various ways and it is important to use match the right type to the application. But it also easy to get wrapped around the axle worrying about squeezing the last drop of performance out of your equipment when moving your head 6 inches in your listening room has a far greater effect.

My career was spent working on airborne radar. Radar returns, once down converted from the receiver, end up as signals in the audio range - DC to about 100kHz - before being digitized for signal processing. No leaf is unturned to extract every little bit of performance. Cost is not a factor because the production run is small (usually). Top quality is used throughout. But there are no audio grade capacitors. No cryogenically cooled conductors or cables. The signal processor is much more sensitive than you ears (but maybe not my dogs).

If I spend $700 on a golf driver and $70 on another, guess which one I hit better? Now, it may be that the $700 driver is really better. Or it maybe it's all just between my ears.

 

[vince666]

Just follow Joe's advice and use film capacitors in the signal path. Most of the polarized electrolytic caps in your equipment are there to bypass the power supply, but I understand that this requires looking at the schematic and being able to discern which are which. Using a bipolar electrolytic in place of polarized one will generally work as long as capacitance and working voltage are adhered to, but I wouldn't want to do that where the cap is being used to bypass the power supply.

the "problem" about film caps is that if you need, say, 10uF then a film cap would be way too large size and also very expensive, especially if we choose the polypropylene ones... actually, also a 1uF WIMA polypropylene cap would be quite a bit large.

In general, when going to recap a deck, if anything, I try to carefully look at the whole schematic and then put bipolars to replace the ones which are clearly working as coupling caps, i.e. connected in series through the signal path so that the signal goes from one leg to the other leg of the cap so that it's there to let AC signal pass through it while also blocking any DC which might be there.

But, if a cap has one leg connected to the ground (and on the other leg there is some DC voltage), it's clearly there to filter out AC ripple and it's always DC biased and, in this case, I go with a good low impedance polarized like Panasonic FM or similar quality stuff.

Last, those caps which aren't exactly working as coupling caps but they aren't either working as supply bypass caps between some DC voltage and ground... i.e. like some of those near the dolby chips or in the negative feedback areas of preamps/op-amps, then I tend to go with low leaking ones like Nichicon UKL... also, if I happen to find any old tantalums then I replace them with Nichicon UKL.

Of course, I go with Aluminium Electrolytic caps because their capacitance values would be a bit too large for film ones and, btw, the deck just had electrolytics at that place.

Till now, the few decks I've fully recapped by following the above "rules" are working really nice.

 

[vince666]

about sizes... below we have 3 different Nichicon 1uF bipolars and 1uF WIMA PP cap from one of the few series which, on the specsheet, they mention they are suitable for high quality audio... so, for sure that WIMA would be a great choice but it's just HUGE in comparison and it just won't fit in place... and it's only 1uF... I don't even know if they make the 10uF one (if so, it would be a lot bigger than that), a value which is often found in audio signal paths, i.e. as output coupling capacitor at line out.

 

[20tajk7]

The Aurex PC-X80AD uses a lot of film capacitors in the signal path, the 10uf are as big as a 1000uf electrolytic :
http://knisi2001.web.fc2.com/pc-x80ad.html

 

[vince666]

The Aurex PC-X80AD uses a lot of film capacitors in the signal path, the 10uf are as big as a 1000uf electrolytic :
http://knisi2001.web.fc2.com/pc-x80ad.html

if they are polyester film, they might be slightly smaller than polypropylene ones of the same values but, for audio path coupling, polypropylene is preferable as I seemed to get it.

 

[Alex SE]

Wima is probably suitable for DIY projects, but for recap... as you guys use to say: HUH?