By-pass Capacitor?

[boom3]

(snip)

------------------------------------------

We measure many things in audio that don't seem to correlate to how it sounds, simply because we know how to measure them (the light is better).

An oscilloscope is a very poor device for measuring most things, you have a hard time even seeing 2% THD on one. But you can see a little something about caps.

Let's run a little 600hz through them (they should all be a straight line).

(1) Man, the tantalum looks pretty bad (imagine that, they sound bad too). (2) Aluminum looks a little better. (3) Mylar (polyester) looks even better. (4) Polypropylene, even more so. (5) Polystyrene looks like a straight line, too bad they don't make them bigger than about 0.1µF. (6) Paper in oil doesn't look too shabby, but it isn't perfect.

Now let's do some tweak stuff.

(snip)

7) A stock tantalum for reference. 8) 20V DC bias makes it look much better. 9) 100V DC bias makes it look very flat (wonder if JBL knew this when they designed their DC biased crossovers)! 10) Stock paper in oil for reference. 11) Paper in oil with a good bypass cap (looks pretty good, hmm). 12) Why we don't use ceramic caps in the signal path.

Note: the bypass cap in 11) was a polycarbonate, not available in larger values anymore. If you want to really do it up, use a small film-and-foil polypropylene and an even smaller polysteyrene for the bypass caps (like JBL does on their better stuff). Or you could go the DC bias route too (JBL does both on their top-of-the-line product).

Photos from Steve Bench.

I had seen this before...but this is measuring hysterersis at one frequency, right? I'd love to see this repeated, at a lower scope trace voltage setting (i.e. greater detail) with higher frequencies, with and without the "bypass". In terms of traditional capacitor specifications, I think "hysteresis" can be equated to "dissapation". Since we normally associate hysteresis with inductance, is this a measure of how inductive the caps are?

The DC bias pix are to me more siginificant...I had taken issue with some of the "mystic" copy writing of JBL's explanation for this some time ago, and I'm glad to see some evidence, however preliminary, that it can work. Now, can this have an effect on good quality film types? After all, nobody, certainly not JBL, is going to use a lousy tantalum in a crossover only to insert an expensive DC bias circuit.

[DrewBolce]

The Dayton's are an incredible value for about $1 each, other similar caps are so much money I won't use them.

Polystyrene is a problem. Rel-Cap is the only one I know making new high voltage polystyrene caps. Michael Percy sells them for about $4 each. New Old Stock (NOS) can sometimes be found for about half price. Make sure they are rated at at least 160V. Polystyrene can melt while soldering if you're not careful.

Ah, yes. Thou shall not touch the cap with the soldering iron. Cost me a fotune when I was putting my first computer together (1974 - yes, before the beginning).

I would not use Polystyrene in any place where voltage is a concern. I have been been using the Illinois polypropylene for some time now
(yellow axial from Antique electonics). 630v, priced well with size and
sound similar to most vintage caps.These caps have the same problem, one touch with the iron and poof, no cap.

If you are saying that you need 160v or better caps for speaker crossovers I would like to hear your reasoning. I have always felt that 100v and above caps were good, 50v could be used with wisdom.

[kg4guy]

Thanks I had my eye on those .Is there a proper way to wire a by-pass cap with the main cap. Thanks a bunch

[DrewBolce]

Thanks I had my eye on those .Is there a proper way to wire a by-pass cap with the main cap. Thanks a bunch

Not really, I used 0.22 and 0.47uf values. Try to get them about 10% of the big cap value and have them add up to the spec value. Paralled them with the big cap and set them up so you can change them. I would suggest auditioning them. Get the same values in the brand of big cap (example, Solen) so you are not listening to the value change.

Al has done some rethinking on the universals, I would suggest looking at his website and/or contacting him to get his latest insights.

[mark1101]

Yeah...for the benefit of the less experienced, we should note that paralleling caps to get a given value is perfectly legitimate and often necessary...not at all the same thing as the bypass cap notion.

What I did was call the supplier and have them measure the caps so I got the exact values I needed. Nevermind that an advertised Auricap does not come in the value you need. It does. Someone just needs to go through the bins and measue them......which is a very common service. One other complaint I heard was that their tolerance was only 10%. If you get someone to measure them......you get the exact value you need and no tolerance is involved. I built all my networks this way. In fact.....if you don't you are not going to have matching networks.

No offense, but if you buy cheap caps (Daytons for example)......and bypass them or not.......you are going to get cheap sound. Period. Test for yourself.......BY EAR. There is no way to buy a cheap cap and have the sound of the more expensive ones. If that was possible the cheap ones would cost a lot more.

If you think the cheap ones sound good......great......then we are done here. Go get them.

[boom3]

Yeah...for the benefit of the less experienced, we should note that paralleling caps to get a given value is perfectly legitimate and often necessary...not at all the same thing as the bypass cap notion.

What I did was call the supplier and have them measure the caps so I got the exact values I needed. Nevermind that an advertised Auricap does not come in the value you need. It does. Someone just needs to go through the bins and measue them......which is a very common service. One other complaint I heard was that their tolerance was only 10%. If you get someone to measure them......you get the exact value you need and no tolerance is involved. I built all my networks this way. In fact.....if you don't you are not going to have matching networks.

No offense, but if you buy cheap caps (Daytons for example)......and bypass them or not.......you are going to get cheap sound. Period. Test for yourself.......BY EAR. There is no way to buy a cheap cap and have the sound of the more expensive ones. If that was possible the cheap ones would cost a lot more.

If you think the cheap ones sound good......great......then we are done here. Go get them.

Thanks for the corroboration Mark. As you state, it is necessary, not just good practice, to measure components going into a circuit. LCR meters are pretty cheap these days...with accuracy acceptable for crossover work. I measure all my components and label them with actual values as I build my crossovers.

[dlp4341]

Hi

Do you know what current was used for the scope traces ?

Don

[djk]

26mA

http://members.aol.com/sbench102/caps1.html

http://members.aol.com/sbench102/caps.html

[dlp4341]

Thank you

Don

[DrWho]

There is no way to buy a cheap cap and have the sound of the more expensive ones. If that was possible the cheap ones would cost a lot more.

Why? What is different about the fabrication process that yields better performance? And then how is that fabrication process more expensive?

[djk]

"What is different about the fabrication process that yields better performance?"

Soldered lead-out wires rather than swedged.

"And then how is that fabrication process more expensive? "

Labor cost. Foil is also more expensive than metalized film.

[BEC]

"What is different about the fabrication process that yields better performance?"

Soldered lead-out wires rather than swedged.

"And then how is that fabrication process more expensive? "

Labor cost. Foil is also more expensive than metalized film.

And what do you think this added manufacturing cost might be? Think that perhaps 50 cents might cover the added material and labor cost? I expect a dollar would be way high.

Reminds me of the time perhaps 15 years ago when I was talking to a Cooper tire executive. He told me that the manfacturing cost (material and labor) difference between their most expensive tire of a certain size and their cheapest tire of that same size was under a dollar. Real cost difference was much higher because of (mainly) advertising cost on the more expensive tires.

Bob Crites

[kg4guy]

Is there any facts to back up that the more exspensive caps are better or is it a personal sound preference ?

[mark1101]

I don't know. I just know that in direct listening comparisons I made.........the budget caps sounded inferior to various degrees to the more expensive so-called "audiophile" metalized film, and film and foil caps that I tried. You might find a case where a cheap cap might sound more pleasing than more expensive ones. If you do, great. I never did though. In fact I found performance to be pretty much in line with pricing. Although it is subjective, I think it would be difficult to find a budget cap that is more pleasing than an Auricap for instance. I could easily hear the differences and was not satisfied with the budget sound. The performance justifies higher pricing IMHO. Albeit not as high as some cost. This all relates to the midrange and tweeter positions by the way. I used the budget caps (Solens) in the bass positions where I really could not hear any differences.

[DrewBolce]

"What is different about the fabrication process that yields better performance?"

Soldered lead-out wires rather than swedged.

"And then how is that fabrication process more expensive? "

Labor cost. Foil is also more expensive than metalized film.

So the Dayton Film and Foil are cheap how? The Auricaps mentioned are metalized, not film and foil. They are superior how?

The biggest myth in Audio today is that insanely more expensive is better. You can put your money into the sound or the bragging rights, your choice.

This is addressing the comments made before, not Don's quote. Essentially he is right. Materials, construction, quality control make better products not final sale price.

[mark1101]

Auricaps are very clearly far superior to Daytons in their sound at my home and in my system in tests I did acouple of years ago. Why? I have no idea. My comments have always been about the sound/performance. I really never cut one up to look at which is made better, or what their secret may be. It would be interesting to know what actually makes the sound better.

[tgw323]

I replaced the titanium tweeter cap on a JBL with a Dayton poly. It seemed to smooth it out some. I was quite pleased. As far as there being something better out there, I couldn't say because I didn't try it. But I would suspect that much more improvement would come at a substantial price.