what makes a capacitor sound better?

[tofu]

pardon my ignorance, but i've been doing some research on capacitors and this is one question that was always in the back of my mind. take note that i haven't listened to any of the caps i will be mentioning (yet)

howcome a 10% auricap supposedly sounds better than a 5% dayton? i thought 5% was a better value? wouldnt the cap with the tightest values logically sound better? howcome hovlands cost so much more than solens? they seem to have the same specs on paper, but their prices sure are different.

while surfing google, i came across this http://www.audioholics.com/forums/archive/index.php/t-4948.html

they did blind a/b tests with various expensive/cheap caps, and people couldn't really tell the difference.

i am neutral on the subject. i can't really form a legitimate opinion without hearing both for myself. i personally am going with more expensive caps to upgrade my crossovers (auricaps) just to play it safe, but i've heard a lot of conflicting reports.

i'd like to hear your unbiased opinions on the matter.

[djk]

What makes a cap sound good?

Low dielectric absorption (polypropylene)

Non-ferrous leads (copper or silver)

Soldered or welded leads (most cheap caps just mash the leads into the end with some tin paste)

Foil and film (cheaper caps are metalized film)

Old paper in oil caps sounded good as they tended to have foil construction and soldered or welded leads. Jensen is the only current company making them like this AFAIK.

Current oil caps from GE and ASC are metalized film with less than the best lead attachments and as a result do not sound the same as the old stuff.

Klipsch used WWII surplus PIO until stocks ran out in the early 70s, switched to mashed end film, and then back to custom made PIO in the later 70s. About 1985 they went to film again, using some better quality parts this time around.

[Good2BHome]

I think I undersand everything except the abreviation PIO.

Mike

[D-MAN]

It's means "Paper-In-Oil" capacitors. Klipsch originally used metal-can AC motor "kicker" caps, I believe for washing machines or something, weren't they?

DM

[Deang]

i'd like to hear your unbiased opinions on the matter

I don't have an unbiased opinion. Do you still want to know what I think?

[boom3]

Klipsch used motor-run paper in oil caps, which are designed for continous duty, as opposed to motor-start caps, which are for just that purpose. Back in the day,there was nothing better than paper caps for crossovers and the motor-run types were (and are) stable and rugged.

I don't want to get into the whole Capacitor of The Month issue, which has gotten me entirely too upset over the years. To judge by the online chatter and most (but not all) press reviews, there seems to be a direct correlation between amount of money paid for capacitors or any audio component and the perceived sound quality. That is about the only accepted "correlation" as audio tweakery goes further and further away from science into metaphysics.

[Good2BHome]

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

On 12/31/2004 2:44:19 AM DeanG wrote:

i'd like to hear your unbiased opinions on the matter

I don't have an unbiased opinion. Do you still want to know what I think?

http://forums.klipsch.com/idealbb/images/smilies/1.gif">

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

Yes I would!!

[JohnA]

Oh Man! You didn't REALLY ask that?!

[yaffstone]

Assuming there must be a 25 cent lecture on this somewhere in this forum, but not having found it yet, here goes: Is there any reason to replace the entire cross-over network, or can original performance be had by just replacing the caps with modern high quality? Does DeanG build entire new units to be better than original, or is it to match the IC values of the original circuit using available values of new components? For instantce, my Belles have 13 uF caps in them but I can only find 12 uF in new caps. I could add a 1 uF in parallel or go with a different inductor. Worse probably is the 2uF in the original with the closest new value of 2.2 uF, an intrinsic 10% problem. How sensitive are the absolute values of these components given that the originals were in the 10% tolerance range?

Instruction now please:

[Frzninvt]

Do you have to replace the whole network? NO, you can just update the components and keep the original autoformer.

You will not find the indentical values you need without some digging, they can be had though. If you do not want to invest in all new networks, I believe that DeanG can get you the 13uf and 2.2uf caps that are needed for the "AA" network. I would replace the old coils and zener diodes as well and use better quality wire. I piece updated my AK-2's in my '88 Khorn's to AK-3 functionality but I really need to replace the 13uf Aerovox caps in the midrange circuit and the three 2.0uf's in the tweeter circuit.

I have one of Dean's stunning Jensen PIO "AA" networks in my center La Scala the difference over the original was stunning! I would gladly purchase another set after hearing there performance first had.

Upgrading the aging crossover components in your old Heritage is some of the best things that you can do to further improve their already stunning performance.

[Deang]

Assuming there must be a 25 cent lecture on this somewhere in this forum...

Yeah, probably about 2 dozen threads, and I sure wish I had bookmarked them all.

Is there any reason to replace the entire cross-over network...

No, not really. However, in the old Heritage networks, one can get a little lift by replacing the T2A with the UT 3619 autotransformer. If one chooses to do this -- they might as well rebuild from scratch.

...or can original performance be had by just replacing the caps with modern high quality?

Yes.

Does DeanG build entire new units to be better than original, or is it to match the IC values of the original circuit using available values of new components?

Yes and yes. More related to this below.

For instance, my Belles have 13 uF caps in them but I can only find 12 uF in new caps. I could add a 1 uF in parallel or go with a different inductor. Worse probably is the 2uF in the original with the closest new value of 2.2 uF, an intrinsic 10% problem. How sensitive are the absolute values of these components given that the originals were in the 10% tolerance range?

Sounds like you have the Type A, but I don't know because you didn't say. My personal belief is that values should be as near dead on as possible. Some manufacturers/vendors give a 10% tolerance on their capacitors, but my experience with measuring them is that they are typically within 3% to 5%. I have on occasion run across the odd ball type that I've simply set aside and sent back for an exchange. The exception here was with Jensen -- and I had to be careful with their 2uF's.

I believe some other things too, and some of it has to be taken with a grain of salt. I am NOT an EE or filter designer. I am an audiophile that slings solder. So, I rely heavily on those who understand the math, or have the proper instruments for testing, and/or have been more involved in more extensive comparative listening than me. I place trust in those who know more than I do, and simply cling to the hem of their garments! Of course, as an "audiophile", I also reserve the right to reject any numbers or testing results that don't always line up with real world experiences. Now, doesn't THAT just give you the warm fuzzies?

I also live in the archives and at DIY sites, and if there in one thing I've learned, it is that though there is some agreement -- there is a hell of a lot of disagreement! So, I converge on where there is agreement, and then eat the meat and spit out the bones. Don't ask me where the dividing line is between the math and voodoo -- because I don't know!!

So, what's a guy to do? I start by putting trust in the original circuit. Hard to go completely wrong there, ya think? People make this very easy for me too. The number one thing I hear is this -- "I really like the way my speakers sound, what's going to "change"? Early on, I would tell them what I thought, but I also really wanted to know what they thought -- and so the work was more like a partnership in experimentation. Sometimes it's still like that. Nothing carries more weight than personal testimony and the confirmation that comes with it. After trading emails with everyone a few dozen times, I saw patterns developing that told me I was on the right track with some of the things I thought.

The best part isn't necessarily the "best" part.

The old Heritage networks utilized PIO motor runs, which by their very nature were somewhat "lossy". IOW's, PK "voiced" his speakers with capacitors that had higher than normal (by today's standards) resistance (ESR). The losses in the circuit contributed to the voicing. Is this important? Yeah, to a degree it is.

All passive components possess measurable amounts of resistance, inductance, and capacitance.

Capacitors have resistance and inductance

Inductors have capacitance and resistance

Resistors have capacitance and inductance

Rather simplistic, but it helps to remember it when yanking and replacing parts out of your own, or someone else's network.

I believe the perfect replacement capacitor for the old networks are the Jensens. Hans Jensen rolls the Heritage values for me, but I must order at least 10 pieces at a time. There is also shipping and duty. I originally built my networks with these, but my inability to handle the out of pocket expenses, and the fact that most couldn't afford them anyways -- forced me to seek an alternative. So, I decided to go with the Audience Auricap. However, before selling any, I built up a set of networks with the Jensens and sent them to Craig so he could compare them to the Auricap Type A's I had built him earlier in the year. His opinion what that there simply wasn't enough of a discernable difference between the two networks to warrant the additional cost of the Jensens. Bob Crites had also bought some 13uF and 2uF Jensens from me about the same time, and after extensive testing and comparison to his GE motor runs -- agreed. Even though I did have customers that emailed me to say there was indeed a difference between the GE cans and the Jensens, in the end I felt it was a mute point -- most simply couldn't afford them. So, the Auricap was in.

Why the Auricap? Why not Hovland?

This is where the best part isn't necessarily the "best" part comes in. The Hovland is a film and foil type, almost completely loss-less. ESR's are very, very low on these. I just don't believe they are the "best" part to use in these old networks. Yes, they are very clean, but they'll definitely bring the treble to bear. The number one complaint about these on the DIY forums is that they are "BRIGHT". Obviously it's circuit dependant, but for the old networks -- I think they are a general no-no if one wants to maintain the balanced sound of their speakers. The ESR's of the Auricaps fall between the PIO and the film and foil types. They lower insertion losses without shoving the horn into your ear.

Why not the Solens?

I've discussed this at length with Al. Basically, I refuse to use them because they're not yellow. I also believe thick filmed metallized polypropylenes with superior lead terminations sound better than el'cheapo caps from France. An Auricap of the same value and voltage rating is 50% larger. I will use them in LF circuits if I can't find a Dayton Audio equivalent however. The "thick film" advantage may be bordering on voodoo. However, I don't think so. Metallized polypropylenes are "self healing". One would think this is a good thing. However, the question that should be asked is what are they being healed from, and if they aren't "healing", what in the hell are they doing in the meantime? http://ldsg.snippets.org/appdx-ec.php3#CAPS

Why not GE motor run cans like BEC uses?

Because loudspeakers don't have motors in them. Well they do, sort of, I guess. Did I mention that they're not yellow?

[Deang]

I just went out to percyauido, partsexpress, and partsconnexion -- and was surprised to see that none of these places carry a high quality 13uF metallized polypropylene. However, I can get them direct from Audience if you want to try them. Just shoot me an email at dgwescott@sbcglobal.net All three of those places had 2uF's (partsexpress carries the Kimbers, very similiar to the Auricap, and I use them as well).

Another alternative is to run a 6uF and 6.8uF in parallel. The caps typically run just a wee bit to the high side, so this would bring you very close to 13uF.

[scott0527]

Dean,

I painted my Solen caps yellow and the sound improved immensely. Higher highs, lower lows

[Good2BHome]

Now, that wasn't so bad, was it!!!

But, I think you held back something. Dean, tell us how you REALLY feel.

[3dzapper]

Definately a new low Scott.

Rick

[BEC]

When you rebuild a lot of sets of Klipsch crossovers, you will find a variety of brands and types of caps used. Sometimes even when working on a pair of crossovers that seemingly have been produced together, you find caps that are different between the two boards. You may find a 2 uF cap made by Aerovox on one and in the same position on the other board a cap made by Plastic Capacitor Corp. On the type AA, you will sometimes find a single 13 uF cap. Other times a single 12.5 uF cap and pretty frequently you will find 3 4 uF caps in place of the 13 uF. The original caps cover a bewildering range of voltage ratings. Most seem to be rated at 100 VDC. Some are rated 100 Volts 60 hz. Some are rated as high as 600 VDC. I have seen one type rated at 140 VAC 60 hz. None of the ratings really sound like motor run caps to me. Seems the voltages ratings would be higher for any AC motor that I can think of.

Anyway, I said all that to say this. It looks to me like Paul Klipsch spent very little time obsessing about the type, brand, or pedigree of the capacitors he used in his networks. What does seem to be common between all the types and brands he used is that they all were sturdy industrial types and built to last a good many years.

Bob Crites

[Deang]

I have expected to see a big yellow can.

[BEC]

Couldn't find any yellow paint. I looked.

Bob

[LonestarBlues]

I ran across these new U.S.A. made oil capacitors at Audiocircle Marketplace-Commercial & on Audiogon. VH Audio V-Caps OIMP-oil

impregnated metallized polypropylene. You can check them out at www.v-cap.com/oilcaps.html.

The main website is www.vhaudio.

com. They also make a teflon capacitor. Read their info. What do y'all think. Is the info hyperbole marketing or does it sound like a potentially good capacitor? The website says they are expensive. How do they compare in price to Jensens?

Regards, Mike

[Deang]

"...combine the smooth musicality of paper in oil designs with the transparency and superior electrical properties of modern film capacitors."

It doesn't work that way.