Cornwall Crossover Capactitor Experiments

[Seadog]

I decided to consider recapping my 26-year-old Cornwall crossovers to improve the high frequency detail to more like what I get from my CW IIs. The CW IIs are quite a bit brighter that the CWs and I thought this might be due in part to the aging caps in the Cornwalls. Overall, I like the CWs better than the IIs, but I think my CWs could benefit from a little more resolution of detail (for lack of better terminology).

So my first experiment was to replace the old caps with new GE motor run caps (cheap experiment, $18 for all four caps). I changed the caps on one speaker and compared to the other with the old caps, and to my surprise, there is no appreciable difference. I suppose I thought that the new caps might have a lower ESR and give me more high end detail, but the old caps are apparently fine.

Ive read the threads on the subject and I know some like the original style caps, some like metalized film caps (Im concerned about hearing the healing process), and some like foil. BTW, I am using P-traps to calm down the squawkers.

Now the question, what will something like Hovlands do for me? Since my old caps appear to be o.k,, will the $100 for four good film and foils be worth the cost (not a cheap experiment)? Or should I give up and be happy with the original crossover components since the K-77 is no match for the K-79 in the detail category?

Edit: I just noticed the typo in the subject title

[Klipschguy]

My impression is it's probably a waste of money and you won't hear much of a difference (after giving it a month to cancel out the Baranek effect).

Bob Crites (another forum memeber) has tested some GE motor run caps and they demonstrated some pretty nice numbers - for any cap. He also informed me personally that most of the old Cornwall caps he has tested do Okay, whereas old La Scala/Khorn caps are usually more out of spec (who know why??).

(Bob, if I've misquoted your work, my apologies.)

Andy

[BEC]

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

On 1/24/2005 7:21:57 PM Klipschguy wrote:

My impression is it's probably a waste of money and you won't hear much of a difference (after giving it a month to cancel out the Baranek effect).

Bob Crites (another forum memeber) has tested some GE motor run caps and they demonstrated some pretty nice numbers - for any cap. He also informed me personally that most of the old Cornwall caps he has tested do Okay, whereas old La Scala/Khorn caps are usually more out of spec (who know why??).

(Bob, if I've misquoted your work, my apologies.)

Andy

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

Andy,

That is an accurate quote. I can't say with 100 percent certainty the reason, but it does seem that the Cornwalls and Heresys take better care of their crossover caps than do the Khorns, Lascalas and Belles. A theory I can't prove would be that the interiors of the Cornwalls and Heresys keep the crossover caps from seeing as wide temperature swings as the types that have the crossovers just sitting in the back. What I can say for sure is that the typical 2 uF cap from a Type A or AA crossover has an ESR of around 0.7 ohms. I think the worst 2 uF cap I have seen from a Type B or Type E crossover had an ESR of around 0.3 ohms. What we are talking about here are caps that were seemingly identical when installed by Klipsch and that are of similar age.

Also, it seems to me, I start to hear the difference in caps when the ESR gets up around 0.5 ohms. Most of the caps we use in crossovers now seem to start out life with ESR of around 0.02 ohms. I don't know where the ones Klipsch used started but I expect that when new, they may have had ESR numbers similar to the new caps available now.

The Klipsch type AA crossovers are effected the most by this increase in cap ESR. With the typical original 2 uF caps having each increased in ESR to around 0.7 ohms and with two of those in series with the tweeter, the tweeter will get around 17 percent less voltage than it should.

Bob Crites

[Deang]

...the K-77 is no match for the K-79 in the detail category...

Interesting comment considering the K-79 is 3db less sensitive compared to the K-77. The additional "detail" from the Cornwall II is probably due to the different network and the way the drivers are integrated. Maybe just voiced differently?

Numbers, numbers, numbers.

[Klipschguy]

Bob,

Very interesting (stroking my chin with the thumb and forefinger). Is there any inherent problem with using a cap with a little higher ESR if one likes the way the speaker sounds?

Andy

[BEC]

Not a problem that I can think of. I would want to have the highs back where they were when the speaker was new but all that is a matter of personal taste and opinion.

Bob

[Seadog]

"Interesting comment considering the K-79 is 3db less sensitive compared to the K-77. The additional "detail" from the Cornwall II is probably due to the different network and the way the drivers are integrated. Maybe just voiced differently?"

Dean, this must be the case because the IIs are certainly brighter in the high end. The networks must be putting more energy to the K-79. BTW, the IIs measure about 2 or 3 dB louder than the CWs based on my cheap Rat Shack SPL meter.

Based on the responses here, I think I will just stay with the stock caps on the CW. I'm satisfied now that they must sound pretty much like they did when new. I just did not want to be giving up anything due to old caps.

Actually, I've never been dissatisfied with the sound of my CWs, but when I did the A/B with the CW IIs, I began to have doubts, and it was nagging me. I think I'm better now .

[Deang]

"Based on the responses here, I think I will just stay with the stock caps on the CW. I'm satisfied now that they must sound pretty much like they did when new."

Are you using the .1mH and 3uF P-trap? If so, the 3uF value should have the lowest insertion loss possible. That cap should be a high quality film and foil type (Hovland or AudioCap PPT Theta). The rest of the network can be built with thick filmed metallized types with soldered lead attachments. The T2A's should be replaced with UT 3619's, and all connections should be soldered using low lead, silver content solder. If you do these things, I'm convinced you'll hear a noticeable improvement. The improvement isn't subjective, because I have yet to be told by anyone that they preferred the sound of their old networks over the rebuilt ones I sent them. Incidently, my Type B rebuilds out-number my A or AA rebuilds 3 to 1. IOW's, the majority of feedback I receive after a sale are from Cornwallers. Certainly this begs the question: How come they hear a difference and you don't?

[Seadog]

I built my P-traps using AudioCap PPT Theta caps and ERSE perfect layer inductors, so they should be fine.

I have never heard CWs with crossovers built from premium components. That was the point of asking for opinions here before I bought expensive capacitors to play with.

Thanks to all who responded.

Maybe Ill just paint the motor run caps yellow and be done with it!

Dean, you have a PM.