Update 1977 Klipsch LaScala crossover.

[KT88]

1.5 years ago I described how pleased I am with polyester caps I traded in for sonicaps. It was only a test and I used standard values which did not exactly match the Klipsch requirements. I thought 10% tolerance was ok. I used 2x6.8 uF instead of the required 13uF. And for the tweeter of the AA crossover 2 caps each 2.2 uF instead of the required 2 uF per cap. I thought I was not that wrong, but somewhere Roy Delgado described that the uF values should be followed as closely as possible because the intended range of use of the 2 uF caps in series depends on the accuracy of the 13 uF. In the meantime, I have learned that the total ESR of caps in parallel is pretty much the same as the ESR of a single cap with the total capacitance.
So I tried it out, for 13uF I took 6.8 4.7 and 1.5 uF. For 2 uF I took 2x1 uF. I was able to match these polyester caps with each other so that I actually get 13.04 uF for both channels. 
I wouldn't have considered this change worth mentioning if it hadn't had such a positive effect. Ok, at the same time I removed some unnecessary connectors from the xover, of course the circuit remained the same, but these connectors are not responsible for the positive effect of the correct uF values. 
The result is that the LaScala sounds very "right" now. I couldn't get out of my chair because of the listening pleasure, although I just wanted to check that it works after soldering. The loud and sensitive horns register every small change for the better with great effect. Squaker and tweeter discolour much less and the added transparency and punch almost scared me.

Basically, this post is a plea for using JEM original Heritage vintage replacement caps. Only because of the high cost of shipping to Germany, customs and import VAT I did not order them from JEM. I can highly recommend it to every vintage Heritage owner to use JEM caps because only there the uF values are exactly matched to the Klipsch speakers and you will make the same positive experience. 

Below is a picture from today, I am not a master solderer but it works.

 

Don't believe anyone who says that polyester caps are not transparent or that violin strings sound like paper or any other nonsense. They are the caps that exactly fit the vintage Heritage models and Klipsch still use them today, as do other well-known manufacturers, e.g. my Stirling Broadcast LS 3/6 and LS 3/5a use the same type of caps. It's not about buying cheap parts but about the "right" cap in the context of its circuit and it's not about the "best" cap in its own

 

 

[pcbiz]

59 minutes ago, KT88 said:

1.5 years ago I described how pleased I am with polyester caps I traded in for sonicaps. It was only a test and I used standard values which did not exactly match the Klipsch requirements. I thought 10% tolerance was ok. I used 2x6.8 uF instead of the required 13uF. And for the tweeter of the AA crossover 2 caps each 2.2 uF instead of the required 2 uF per cap. I thought I was not that wrong, but somewhere Roy Delgado described that the uF values should be followed as closely as possible because the intended range of use of the 2 uF caps in series depends on the accuracy of the 13 uF. In the meantime, I have learned that the total ESR of caps in parallel is pretty much the same as the ESR of a single cap with the total capacitance.
So I tried it out, for 13uF I took 6.8 4.7 and 1.5 uF. For 2 uF I took 2x1 uF. I was able to match these polyester caps with each other so that I actually get 13.04 uF for both channels. 
I wouldn't have considered this change worth mentioning if it hadn't had such a positive effect. Ok, at the same time I removed some unnecessary connectors from the xover, of course the circuit remained the same, but these connectors are not responsible for the positive effect of the correct uF values. 
The result is that the LaScala sounds very "right" now. I couldn't get out of my chair because of the listening pleasure, although I just wanted to check that it works after soldering. The loud and sensitive horns register every small change for the better with great effect. Squaker and tweeter discolour much less and the added transparency and punch almost scared me.

Basically, this post is a plea for using JEM original Heritage vintage replacement caps. Only because of the high cost of shipping to Germany, customs and import VAT I did not order them from JEM. I can highly recommend it to every vintage Heritage owner to use JEM caps because only there the uF values are exactly matched to the Klipsch speakers and you will make the same positive experience. 

Below is a picture from today, I am not a master solderer but it works.

 

Don't believe anyone who says that polyester caps are not transparent or that violin strings sound like paper or any other nonsense. They are the caps that exactly fit the vintage Heritage models and Klipsch still use them today, as do other well-known manufacturers, e.g. my Stirling Broadcast LS 3/6 and LS 3/5a use the same type of caps. It's not about buying cheap parts but about the "right" cap in the context of its circuit and it's not about the "best" cap in its own

 

 

Nice! I have AA crossovers sitting here with new JEM caps. I might give them a try.

[KT88]

I ask the experts like @deang @captainbeefheart and of course all others about a phenomenon I observed (or rather heard) during the last listening sessions.
Some days ago I posted the photo above of my AA xover. I was spontaneously very pleased with the sound. A very precise and analytic sound, especially in the treble. I had never heard percussion cymbals shimmer and vibrate like that when they move back and forth in the air.
But there was another less good impression mixed into the sound experience more and more. It took two days before I was sure that something didn't sound right. It was an emotional problem. Especially in the midrange I had the impression that the sound was too thin. It sounded overdamped and not relaxed, as if the diaphragm of the mid horn was prevented from decaying naturally. It occurred to me that the new polyester caps might not be of the same quality as the ones I had used before - even though the capacitance values are now very accurate. I also thought that maybe it was a mistake to connect caps from different manufacturers in parallel (I did not get all three necessary values from one and the same manufacturer). Ok, there may be some truth in that but that was not the core issue.
Now I had an intuitive idea, at first it was not logical. I remembered that some time ago I had read the following article about cap resonances. Please, it is very unimportant which company published this. I am only interested in the principle of cap resonances:

 

https://www.humblehomemadehifi.com/download/INFO_ClarityCap_Technical_Report.pdf

 

If you like, please scroll down a bit to the heading "Mechanical resonances".
It is so that one reads everywhere that caps should be firmly mounted so that they do not absorb the vibrations of the speaker cabinet etc.. That may be true. Following this motto I had three cable ties tightly fastened around the caps as you can see in the first picture above in the first post. But this led to the negative sound effect. I only know this now because I have removed the outer cable ties on the left and right of each cap. See photo below.

 

Now my hypothesis. Caps have to deal with two types of vibrations. One is to protect them from the outside. But how do you deal with the mechanical resonances that they generate themselves when an AC voltage is applied? The white paper from Claritycap is about how this company minimises these self-generated mechanical resonances. But I was wondering how to deal with it when you have a cap that generates this kind of resonance anyway? My (unscientific) guess is that the cap has to vibrate. I have to protect it from vibrations from outside but it should vibrate "freely" which it generates itself. I was impressed by the comparison in the white paper of caps with electrostatic speakers. Now that I have removed the additional cable ties, it sounds natural again. You have to believe me, it's not a small difference. How to do it to get the best compromise would have to be narrowed down by experts…if I am right.

I have lost some of the detail and clarity that I had with three cable ties when trying to achieve a very pleasant and natural sound with only one cable tie.
I think this problem is easier to manage if you only have to fix one cap (JEM). Therefore, my post is an additional hint that one should not get into complications and better use a single matching cap from JEM. In addition, it could be a hint on how to fix caps optimally in general. Maybe you could embed the caps in foam before attaching them to allow the self-resonances but still reject the external resonances?

 

 

[Dave A]

Those nylon clips and cable ties will dent your caps over time. I use silicone caulk to fasten all caps to my boards now and there will be zero damage or stress this way.

[KT88]

16 minutes ago, Dave A said:

Those nylon clips and cable ties will dent your caps over time. I use silicone caulk to fasten all caps to my boards now and there will be zero damage or stress this way.

 

Thanks for the tip, Dave. I know the method some speaker manufacturers use to attach the caps with a hot glue gun. Maybe I use it because I have one of those.

[Deang]

¯\_(ツ)_/¯ 

[Dave A]

Hot glue is much quicker and I doubt the heat will hurt the cap.

[mboxler]

If you have the original crossover's T2A, it would be interesting to replace the 3654 with it.  I don't have a 3654, but it's supposed to be the same as the

3636 that I do have, less the extra taps.

 

Here are plots of the 3636 (orange) and T2A (blue).  I used a 13uf capacitor and tap 4 is connected to the K-55-M in my Khorn.

 

Even though tap 4 of the T2A is -3.35db vs the 3636's -3db, the voltage across the K-55-M is actually a little higher in the 400-600hz range.  This is due to the difference in inductance between the two autoformers.

 

Not sure I would hear the difference, but interesting nonetheless.

 

Mike

 

 

[KT88]

 

Mike, I've thought about that too. Somewhere in the basement is the rest of my original AA crossover. I unfortunately sinned 20 years ago and out of stupidity changed some parts of the original AA and lost a coil (the one for the bass) but the T2A should still be there. This is an interesting task to compare the sound with the original T2A Autoformer. Thank you very much!