Which is better, Bipolar or biased unidirectional caps?

Svart

Well-known member
Joined
Jun 4, 2004
Messages
5,134
Location
Atlanta GA USA
Ok so I have a choice..

using Bipolar caps or using two polarized caps facing each other and then biased with voltage.

Which ones and why?

Price is not considered here, Plastic caps are not considered either. Just electrolytics.

Panasonic VS series bipolars are one thought.

Nichicon UD is the other.

They are roughly the same impedence ratings.
 

Svart

Well-known member
Joined
Jun 4, 2004
Messages
5,134
Location
Atlanta GA USA
Sorry, I should have been more clear. These will be interstage couplers. Something similar to the 4k buss schemo floating around.
 

Svart

Well-known member
Joined
Jun 4, 2004
Messages
5,134
Location
Atlanta GA USA
this isn't related to caps but you wouldn't know how the 9k does the bussing do you? I guess it is related because I want to know if they attempted their "dc coupled" schtuff here too..
 

Svart

Well-known member
Joined
Jun 4, 2004
Messages
5,134
Location
Atlanta GA USA
makes you wonder.. if noise can get into the signal through the bias networks, then why not bias the other input to the opamp through a bias network that sets bias halfway between the rails.. essentially "ground". this should cancel at least some of the common noise..

Ok I'll try bipolars as they are not *that* much more expensive in the quantities I'll be using.
 

bcarso

Well-known member
Joined
Feb 20, 2005
Messages
4,055
Location
San Fernando Valley, Los Angeles
I wish it were easy to link to some of the Electronics World articles that Bateman wrote about lytics, including a lot of experimental data. I found it very much counter to my intuitions and prejudices---but the guy has been involved with capacitors all his life evidently, and he's not a loony as far as I can tell.

If I can locate the issues with the articles I could at least paraphrase the results.

It may be some kind of piezoelectric effect that balances when it is a true bipolar, as opposed to merely placing them back-to-back.

The Jensen Transformers note is buried in one of their white papers and has been mentioned by others in here before. They even name a specific brand/type that they measured and found to be the best. IIRC they offer no theory as to why they should work so well.
 

Gus

Well-known member
Joined
Jun 3, 2004
Messages
4,779
Location
n
I have a question about this.

I thought bipolars were two back to back polar caps in one case.

I have not cut one open to check. Never cared enought before.

Is there something different with the construction of the bipolar caps that are liked?
 

Svart

Well-known member
Joined
Jun 4, 2004
Messages
5,134
Location
Atlanta GA USA
just looking over the ssl4k schematics again and I still wonder why they spent money and realestate on 2 polar caps and 2 resistors instead of using 1 bipolar cap. You would think the total cost of the extra parts AND the cost of making more pads/through holes for the extra parts would be more than just buying a single bipolar..

OR

Did they do it for another reason?
 

rodabod

Well-known member
Joined
May 12, 2005
Messages
2,896
Location
London
Well, I recently bought lots of bipolars, and they are more than double the cost of polarised ones in general, so I think using two polarised caps would be cheaper. Also, the CapXon ones which I am using are really big - probably about 5 times the size of some polarised ones which I used previously.

Some bipolar caps are just two polarised ones back-to-back, others are not.
 

bcarso

Well-known member
Joined
Feb 20, 2005
Messages
4,055
Location
San Fernando Valley, Los Angeles
IIRC there's a way that a common element in the case of the "true" bipolar 'lytic is involved in achieving the bipolarity. I'll see if I can find that description.

It would still be modeled as back-to-back caps, but one could imagine that the common element (unless I merely dreamt this :? ) might contribute to some sort of "balanced" behavior, maybe revealed by the spectrum of the distortion.

More generally:

As I believe I mentioned before, according to a friend's account, Peter Qvortrup of Audio Note (audiophoolery alert advisory for those of you that tend to be allergic, and I'm getting a bit of a rash as I type myself) believes that the secret to good-sounding caps is damping the piezoelectric activity. He suggests feeding an a.c. signal to a prospective cap (wound platic films for example) and listening to their acoustic emissions. I've done this for some and you can definitely hear sound coming out. But I have not gone beyond that to attempt to correlate the amount and character of the sound to how the cap measures, let alone "sounds" in the more conventional in-circuit sense.

I will say that Peter's own caps of choice, these massive paper-copper foil-oil quarter-sticks of dynamite affairs he sells that will set you back an appendage or two, are indeed dead quiet in the acoustic emission test. They also leak like crazy (ionic contaminants in the oil??), and I quipped they may be the only caps that require the use of a d.c. servo :razz:
 

bcarso

Well-known member
Joined
Feb 20, 2005
Messages
4,055
Location
San Fernando Valley, Los Angeles
[quote author="rodabod"]Well, I recently bought lots of bipolars, and they are more than double the cost of polarised ones in general, so I think using two polarised caps would be cheaper. Also, the CapXon ones which I am using are really big - probably about 5 times the size of some polarised ones which I used previously.

Some bipolar caps are just two polarised ones back-to-back, others are not.[/quote]

Some lower-loss bipolar lytics recommended for voice-coil-level crossovers are a good deal larger than other bipolars, I've noticed.
 

Gus

Well-known member
Joined
Jun 3, 2004
Messages
4,779
Location
n
From reading the voltage difference is what forms the oxide layers on the Al plates. Now this needs the potential different to to be the right way on the plates.

Yes a polar Al elecro cap has inductance, DA, resistance and a diode effect this has been in books for decades IIRC.

What I am not understanding is are the "new" bipolars a different chemisty and build with only two plates or are they two caps with four plates inside one housing?

Or could it be three foils one not connected to the outside world the middle one and the outer two connected to the leads? Is this how they are made?

If they are one cap I would guess they would need to be closer to a super cap type chemisty/constuction.

Whats the good new bipolars number letters and brand? I think I would like to buy some and cut some open and compare to older bipolars for the construction.
 

bcarso

Well-known member
Joined
Feb 20, 2005
Messages
4,055
Location
San Fernando Valley, Los Angeles
OK I managed to find one of the Bateman articles---not the one with his distortion results but a general overview. It is in the December 1997 EW.

On pg. 999 he writes:

"With aluminium electrolytics, the aluminium base anode foil has the dielectric oxide, i.e. Al2O3 , pre-formed electrolytically on its surface prior to assembly. The anode foil acts as one conducting electrode.

This electrolytically formed dielectric's thickness is self-regulating, attaining some 14Å for each volt applied in manufacture.....The true second electrode is the electrolyte material with which the separating tissue paper is impregnated.

Assuming a polarized capacitor, connection to the electrolyte is made using a second, usually thinner, aluminium foil or cathode. While this cathode is not deliberately formed, it inevitably possesses a much thinner naturally occurring aluminium oxide, electrically equivalent to a few volts. In this way, a pair of back-to-back capacitors is created. One has the desired capacitance and voltage capability while the other has a much greater capacitance but only a few volts capability, applied in reverse.

[NB:]

With a non-polarized, or bi-polar capacitor, this unformed cathode is replaced by a second deliberately formed anode foil. This results in two capacitors back-to-back, usually having the same capacitance and voltage capability."


So there it is.
 

Michael Tibes

Well-known member
Joined
Jun 5, 2004
Messages
759
Location
Berlin, Germany
What would be a good type or brand of bipolars? I remember that there were no real longlife / low impedance types available when I checked things a few a years ago. I always like to use the best available quality (about some thousand hours lifespan at 105 degrees C) because I hate the idea of having to redo my work in some years if the device was used a lot and the caps start to give up.
And who does sell bipolars in europe? Those days it seemed only possible to buy like 10.000 directly from the manufacturer in Japan or so...

Michael
 

rodabod

Well-known member
Joined
May 12, 2005
Messages
2,896
Location
London
Radiospares and Farnell sell them in Europe in CapXon and Nitai brands.

Search RS for part number 521-1659

and

http://uk.farnell.com/jsp/endecaSearch/partDetail.jsp?SKU=3174815&N=0
 

Svart

Well-known member
Joined
Jun 4, 2004
Messages
5,134
Location
Atlanta GA USA
Bipolars are easy to find, but I am trying to find SMD bipolars.. makes things a little harder.. :shock:

I guess I'm going to have to break down and get throughhole parts.

I'm looking hard at the Nichicon ES parts. best price i can find is around .62$ each in quantities of 50.
 

bcarso

Well-known member
Joined
Feb 20, 2005
Messages
4,055
Location
San Fernando Valley, Los Angeles
[quote author="Svart"]Bipolars are easy to find, but I am trying to find SMD bipolars.. makes things a little harder.. :shock:

I guess I'm going to have to break down and get throughhole parts.

I'm looking hard at the Nichicon ES parts. best price i can find is around .62$ each in quantities of 50.[/quote]

SMD lytics are bastard stepchildren of the move to SMD everywhere. I'm surprised to know they exist at all.
 

Svart

Well-known member
Joined
Jun 4, 2004
Messages
5,134
Location
Atlanta GA USA
You can find them, however most are less than 47uf. Many cap companies list higher capacities but finding stock anywhere is proving very difficult unless you buy high quantities from the manufacturer.
 

Svart

Well-known member
Joined
Jun 4, 2004
Messages
5,134
Location
Atlanta GA USA
Awesome work! I've already given up using biased unipolar caps in favor of bipolars but you have cemented that firmly.

thanks!
 
Top