what brand of caps?

[djk]

0.1µF foil-and-film bypass caps , about $1 each.

http://www.parts-express.com/pe/pshowdetl.cfm?PartNumber=027-452

A 2µF Solen is $2.95, a 2µF Dayton + 0.1µF F&F bypass is $2.50

Older Klipsch networks (type A, AA, B, B2, E) run the signal through the midrange cap before going through the tweeter cap(s), so the midrange cap will benefit from a bypass cap (the newer networks do not do this).

A 12µF+ 0.47µF F&F bypass is within 4% of the nominal 13µF value, if you are a real anal type use two 6.2µF + 0.47µF F&F bypass (within 1%).

[djk]

Mouser is OK, but the Bennic caps at Parts Express are less money.

[colterphoto1]

is a 6.2 + 6.8 equally good?

[jhoak]

If I was re-doing AAs I would use Solen 12.0uF and Jantzen CrossCap 1.0uF. Under $10 per network for the 13.0uF.

I don't have a good reason why but I always try to get 1 big cap as close as possible to the value I need and make out the difference with a small one.

It probably doesn't make a single bit of difference in the end so long as you're at 13.0uF total.

[Panacea Engineering]

I totally agree...One big one small and yes, it probably make NO difference what so ever.....!

[colterphoto1]

"and still measure good"

They don't, the ESR has started to climb.

But you don't need a meter, you can hear it.

I recently heard a pair of corner horns using a Type A crossover with the original bathtub caps. I sold the guy the crossovers over 30 years ago and I know everything is untouched on them. These are some of the sweetest sounding speakers I've heard lately. Shortly after I heard these there was a discussion on the forum here concerning ESR on ageing crossover caps, and the point was made (correctly) that high ESR would shift the crossover frequency and attenuate the signal going to the driver on that crossover section. I don't own an ESR meter but I do have a signal generator, a 200 mHz 'scope, freq counter, and DVMs so I ran sine waves and measured the crossover's output. The crossover points were correct and there was no loss through the caps.

So if the ESR has increased since the caps were new, doesn't it appear that the increase is not significant enough to detrimentally affect performance? Or am I going wrong somewhere?

Don, how will I know one type from the other? Are they both the tin can style? And any networks with the black rectangle caps, are those ok or do they degrade also?

[Don Richard]

They are the ones in the picture of the crossover in this thread:

http://community.klipsch.com/forums/t/151173.aspx

[colterphoto1]

I think I have some of those Don. Will check. I recall having some old AA that I replaced and hardly a soul could tell the difference. This could be the reason.

Ok, I"VE GOT A BUNCH OF CAPS HERE, now what? Start ripping old stuff out, solder new ones in place and use the little hold down plates/ty raps. Anything particular you guys would do? Maybe check the batch and try to match them up in pairs as best I can? Hope that wasn't a mistake getting 6.2 and 6.8, that was the best way I found to come up with a value of exactly 13 uf as spec's in the schematic.

[Johnny's Jill]

Hope that wasn't a mistake getting 6.2 and 6.8, that was the best way I found to come up with a value of exactly 13 uf as spec's in the schematic.

Its going to be close enough.

Craig

[tigerwoodKhorns]

Hope that wasn't a mistake getting 6.2 and 6.8, that was the best way I found to come up with a value of exactly 13 uf as spec's in the schematic.

Its going to be close enough.

Craig

Mike, the 6.2 and 6.8 will work just fine and you will not tell a difference. I restored a very nice set of B&W's from the mid 80's and they used 2 or 3 caps, close in value, to reach the spec. (i.e. like using a 6.2 and a 6.8). They did not use one big and one small and these were high end speakers. I have done it both ways. FWIW, the B&W's sounded the same so the original caps were probably still good.

Chill out and have a beer. In engineering we used to call this "analysis paralasis." [:|]

You have a ton of speakers (or at least used to). Try to find some of those really old caps and have them tested. Chuch the rest and buy some inexpensive but quality caps to replace with.

Here is a tidbit. I also restored a pair of JBL L100T's, using Jensen's bypassed with some expensive little buggers (forget the brand, "theta's" maybe). Didn't really notice a difference, but the speakers were popular when I sold them on ebay with them fancy crossovers.

Now, the JBL 100's are not the most accurate speakers. But I used to have a pair of B&W Nautilus 801's, and they used Bennic caps in them. They were made around 2000, cost $11K and used Bennic caps. I also had the B&W Signature 800's. These were $20,000 speakers and here is a big change (circa 2002):

"While the other Nautilus speakers use film as well as film-bypassed electrolytic capacitors, and air-cored as well as iron-dust-cored inductors, the S800 uses film capacitors and air-core coils exclusively, even for the low-frequency bass filter"

So just 10 years ago high end speakers were using some bypassed electrolytic caps is speakers that cost up to $11K. And none of them were using "designer caps." And I can tell you, having owned these speakers, they were outstanding, even with the Benic caps[:$]. The N801's (with the Bennic caps and electrolytic caps) and the S800's (with all film caps and air core inductors) sounded exactly the same. I had both of them here for a while and no difference in sound, the S800's were just prettier.

Now, it seems like all high end speakers lift their skirts to show you their designer caps. Of course one can argue that they are progressing and have now learned that better caps make the speakers better, but come on, did it take until around 2005 to figure this out, or are they just responding to market demand? It reminds me of when Klipsch used Monster Cable on the outside of the K Horn and the same old wire on the inside that they had always used.

The moral here is that if you are using a decent film cap, Solen, Clarity PX series, Jensen, etc., you are using the same stuff that is used in the high end stuff. So don't sweat it.

One last thing, B&W's engineers will take your calls. When I was restoring the older speakers from the 80's I asked if I should use silver solder. He laughed. (read the label, it is like 4% silver). He also told me that putting in some new caps that are decent quality should make the speakers better "by bringing them back to spec." This all helped break me of my voodo ways. If you are using decent parts, (meaning decent film and foil caps) the differences that you hear are from your changes equalizing the sound (rolling off the top, boosting bass regions, etc.).

I hope this post does not make me too unpopular but this has been my journey. Enjoy your restoration.

[colterphoto1]

Exactly what I needed to hear. Thanks!

[Quiet_Hollow]

This all helped break me of my voodo ways.

...and buying an ESR meter helped me break mine. It has come in handy for a myriad of other household electronic repairs too. []

[MikeFord]

When the original crossovers were designed, the engineers knew about ESR and compensated for it. Replacing an original cap with a new cap that has lower ESR is removing resistence from the original circuit, which changes frequency response.

Classic example of "different" sounding good regardless of merit.

OTOH I think a lower ESR cap is usually a better choice, but it may require more than a simple parts swap to keep frequency response accurate.

[djk]

"When the original crossovers were designed, the engineers knew about ESR
and compensated for it. Replacing an original cap with a new cap that
has lower ESR is removing resistance from the original circuit, which
changes frequency response.'

Hey, that sounds like you might know what you're talking about, but it's not well thought out..

The ESR on the caps used in the Heritage series when new measured so low as to be totally ignored from a design standpoint. Now an electrolytic is a different subject, but were not used in Heritage Klipsch until the series II version of some of the smaller models came out. Even then they were not in series but were used in parallel with the woofers. The PTC devices on some of the tweeters were much more of an issue for changing the sound, the post-trip recovered resistance is much higher than the initial resistance.

[DrWho]

Extra ESR also exagerrates the voltage non-linearities of the capacitor as well...

Btw Dennis, are you talking about the thermal hysteresis or is the actual PTC value changing when it gets back to room temp?

[djk]

"the actual PTC value changing when it gets back to room temp?"

Yes.

The RXE 050 used by Klipsch is 0.50Ω cold, and 1.17Ω post-trip.

That resistance is far exceeds the ESR of a fresh cap of any sort.

Bennic Electrolytics DISSIPATION FACTOR Max 3% @1KHZ ( When @1KHZ
Capacitance under 6.8μ (smaller values will be less). This implies an ESR of about 0.175Ω at 4Khz for a 6.8µF (the highest of the values rated at 3% MAX).

Experience in replacing large values (~40µF) of a lower quality electrolytic capacitor in the midrange crossover with a film type required adding series resistance to restore the frequency balance. The new capacitor sounded much cleaner, clearer, and more detailed, especially during quieter passages in the music (lower dielectric absorption).

[Axz Hout]

A

[MikeFord]

"When the original crossovers were designed, the engineers knew about ESR
and compensated for it. Replacing an original cap with a new cap that
has lower ESR is removing resistance from the original circuit, which
changes frequency response.'

Hey, that sounds like you might know what you're talking about, but it's not well thought out..

The ESR on the caps used in the Heritage series when new measured so low as to be totally ignored from a design standpoint. Now an electrolytic is a different subject, but were not used in Heritage Klipsch until the series II version of some of the smaller models came out. Even then they were not in series but were used in parallel with the woofers. The PTC devices on some of the tweeters were much more of an issue for changing the sound, the post-trip recovered resistance is much higher than the initial resistance.

I'm not sure I know what I am talking about, but I have a vast appreciation for what I don't know. Point being fiddling with anybody else's design has the inherent problem of not knowing fully why they did what they did. That said, have fun, you own them, try what you want. I started swapping caps 35 years ago with a set of DQ10's, and a crossover doesn't get much more twisted than those (35 components if I remember correctly). Putting 220 uf film caps into my Quad ESL63's made a huge difference in the sound I'm still not fully certain the technical extent of.

I've got a set of film caps waiting right now to put into my Heresy's, but my preference is to go with a comprehensive update of the whole crossover that has been well researched and tested, subjectively and objectively for its changes.

[colterphoto1]

Ok, got all my parts here for the LSI AA re-cap project. Any particular solder/ construction notes you guys can pass along? I have the little Ty-rap tie-downs. Just strap the 6.2 and 6.8 together and run in parallel instead of the 13uf, right? Got photos?

[MikeFord]

Ok, got all my parts here for the LSI AA re-cap project. Any particular solder/ construction notes you guys can pass along? I have the little Ty-rap tie-downs. Just strap the 6.2 and 6.8 together and run in parallel instead of the 13uf, right? Got photos?

I am solidly in the crimp camp, and solder with good quality electronic silver solder, 4% if I recall, only when crimping isn't practical.