Capaicitors

acorn · Post by **acorn** » Mon Dec 16, 2013 1:57 pm

Would there be any advantage in replacing a standard electrolytic
capacitor in a cathode bias output stage with non polarised electrolytic type.

Acorn

IslandPink · Post by **IslandPink** » Mon Dec 16, 2013 6:46 pm

Not unless you can get a Black Gate NH !
You might try an ansar PP if you can get the value you need in a reasonable size.

simon · Post by **simon** » Mon Dec 16, 2013 7:31 pm

There was a thread on DIY Audio about just this subject a few weeks ago. Some were adamant they could hear a difference, others maintained that a difference can only be heard if the design wasn't a good one. And so the argument went around. And around. And around.

I've never tried the same values of lytics and PP as cathode bypasses so don't know. And if I did hear a difference, my design would be inadequate anyway

.

Post by **Nick** » Mon Dec 16, 2013 8:20 pm

I've never tried the same values of lytics and PP as cathode bypasses so don't know. And if I did hear a difference, my design would be inadequate anyway

Given the potential tolerance of electrolytics, without measuring and matching, I don't think you can say if they were the same value anyway.

IslandPink · Post by **IslandPink** » Mon Dec 16, 2013 8:55 pm

The problem as I see it is the quality of caps available as non-polarised is doubtful anyway - I'm not aware of any above 'base level' that are currently available.
However, am I right that you can construct a non-polar cap from two ordinary electrolytics ?
- In which case it would be well worth trying - to see if there were any phase issues related to inductance that might be cancelled .... like I found at various stages in the DHT heater project. ( and the Audio Note advice on black gate usage for cathode by-pass .. ) Maybe we are missing an obvious cheap tweak

If some people on DIY Audio heard a difference then we might discount the 'average' conclusion from there, as it will be heavily weighted by the typical bozos who inhabit that place !

Mike H · Post by **Mike H** » Tue Dec 17, 2013 10:53 am

Only one way to find out, get a bunch of different brands / types and try them

Just to confuse the issue further, I did mention elsewhere that my stereo 6V6 SE amp sounded shite compared to the original AM wireless amp (identical design it was based on), only main difference as far as I can see is the 'lytics. The wireless amp uses 30 year old non-polarised 'lytics from Maplin (originally) for cathode bypass (both stages). The stereo version had Panasonic FC and the treble is way too OTT.

jack · Post by **jack** » Tue Dec 17, 2013 11:30 am

IslandPink wrote:However, am I right that you can construct a non-polar cap from two ordinary electrolytics ?

Absolutely - two ordinary electrolytic capacitors in series + to + or - to - will do the trick nicely - obviously the rules regarding caps in series come into play - the total capacitance will be halved (assuming both caps are the same value), the ESR will be doubled (sum of both caps) as will the parasitic inductance which can lead to "interesting" side effects, and leakage will be (possibly much) greater. Each cap must be rated at at least the full voltage swing.

IslandPink · Post by **IslandPink** » Tue Dec 17, 2013 1:05 pm

Thanks Nick .
Now, on the inductance, is it possible if you put them phycially close to each other - or eg. with their top faces touching, you could partially cancel the inductance ? That might be worth trying. After all, assuming the 'goop' inside doesn't have much magnetic effect, they seem a bit like an air-cored inductor (?) .

acorn · Post by **acorn** » Tue Dec 17, 2013 1:31 pm

I have just tried two non - poarised 100uF Rubycons and they sound quite good but as stated it is hard to tell, I see Maplin sell this type as well.

Acorn

al newall · Post by **al newall** » Wed Dec 18, 2013 8:05 am

These are worth trying, they sounded different in a crossover but a bit too toppy for me.

http://www.hificollective.co.uk/compone ... polar.html

But back to back caps were a success.

pre65 · Post by **pre65** » Wed Dec 18, 2013 9:43 am

Looking at the data sheet for those, it says service life is 3,000 hours.

That does not seem a lot, just over a year at 8 hours a day.

IslandPink · Post by **IslandPink** » Wed Dec 18, 2013 11:06 am

If you open up the .pdf file, you'll see they have a life of 5000 hours at +85°C , that's what I expected - usually these 'working life' numbers are quoted for elevated temperature . You should get several years at 30-40°C like most electrolytics, I would think .

jack · Post by **jack** » Wed Dec 18, 2013 11:34 am

IslandPink wrote:Thanks Nick .
Now, on the inductance, is it possible if you put them phycially close to each other - or eg. with their top faces touching, you could partially cancel the inductance ? That might be worth trying. After all, assuming the 'goop' inside doesn't have much magnetic effect, they seem a bit like an air-cored inductor (?) .

Not sure that you'll be able to lower the total inductance by simply repositioning the caps - parasitic inductance is internal to the caps and there is not much (if any) stray external field - you'll just have to live with it, like the ESR and leakage - making a non-polarised cap this way is a compromise, not a direct replacement.

Toppsy · Post by **Toppsy** » Wed Dec 18, 2013 1:05 pm

Al wrote

These are worth trying, they sounded different in a crossover but a bit too toppy for me.

What you saying they sound too much like me!

Not sure I have been insulted or complimented

al newall · Post by **al newall** » Wed Dec 18, 2013 6:59 pm

audio-talk

Capaicitors

#1 Capaicitors

#2

#3

#4

#5

#6

#7

#8

#9 Capaicitors

#10

#11

#12

#13

#14

#15