AA networks...

[Chief bonehead]

5 hours ago, BEC said:

 

Let me tell you how I did this.  I was rebuilding lots of AA crossovers around 10 to 12 years ago.  I was also testing the old caps to see how they had changed.  Most of them had a pretty high ESR, but just occasionally, I would find one that read very good on ESR and capacitance.  I mean an ESR of less than 0.1 ohms.  Pretty soon I had enough of those old caps that read good to use them together in a new AA crossover.  I ran curves for that crossover on the spectrum analyzer.  Then compared that to curves I got with the new caps I was using.  Almost a perfect match.  Just to see the difference I did that comparison again with some old caps that all tested about 0.5 ohms ESR.  Could see some difference there.

 

So, I think, as long as ESR is low enough, doesn't make any difference if the dielectric is paper, mica or whisky bottles.  That last sentence paraphrases a statement of PWK in a letter written in 1984.

 

Bob Crites

thats the way to do it.  curious, did you do voltage transfer curves?

 

[BEC]

Chief,

I use a spectrum analyzer (Stanford Research Model SR1+) that sweeps 0 to 25K at 1 volt into the crossover input.  All outputs are terminated into the expected impedance with resistors and I run a trace of woofer, squawker and tweeter all plotted on top of each other.  I can see the electrical crossover frequencies and voltage output of each section.  I do that for every crossover I build or rebuild.

[mustang guy]

Roy, for us newbs, could you point us in the direction to learn about voltage transfer curves with regards to crossovers? I did some Googling, and the closest thing I could find was info on constant voltage crossovers or amp (CMOS) design with lots of math. Nothing of any educational benefit comes up when I type in voltage transfer curve. Is what Bob did checking the VTC?

 

Could you explain it to me like I'm 5?

[Deang]

The "curve" is the output. The process is called "voltage transfer function of a filter". 

 

You're funny Roy. When you were handing out those schematics for the Jubilee, (which call out Mylars), no restrictions were issued regarding capacitor types. I specifically asked you about this at the first Jubfest in 2007, and you shrugged the whole thing off, and referred to different capacitor types as "salt and pepper" - but NOW we have to run a Transfer Function analysis on our four and six element filter builds!

 

I sure hope Marion EQ'd his filters back to where they're supposed to be after spending $3K on those Solen air cores and Mundorfs - because those networks sounded really terrible. 

[Marvel]

So... is Klipsch palnning to sell electronic crossovers and amps as a package for the La Scalas and Klipschorns, so the time alignment issues can be dealt with. Even aligning my tweeter on my LaScalas made a huge and very noticeable improvement in smoothness, detail and imaging.

 

It's just a thought, although I am quite content with my passive crossovers.

 

Bruce

[mustang guy]

1 hour ago, Deang said:

The "curve" is the output. The process is called "voltage transfer function of a filter". ... 

Now I'm sorry I asked. I found this: http://www.aast.edu/pheed/staffadminview/pdf_retreive.php?url=352_32115_EC434_2012_4__3_1_Chapter_03.pdf&stafftype=staffcourses

 

   

[Chief bonehead]

2 hours ago, Deang said:

The "curve" is the output. The process is called "voltage transfer function of a filter". 

 

You're funny Roy. When you were handing out those schematics for the Jubilee, (which call out Mylars), no restrictions were issued regarding capacitor types. I specifically asked you about this at the first Jubfest in 2007, and you shrugged the whole thing off, and referred to different capacitor types as "salt and pepper" - but NOW we have to run a Transfer Function analysis on our four and six element filter builds!

 

I sure hope Marion EQ'd his filters back to where they're supposed to be after spending $3K on those Solen air cores and Mundorfs - because those networks sounded really terrible. 

and you are funny looking!

 

and i need paul harvey in here to tell the rest of the story...obviously you dont remember well.....or is it selective memory...

[Chief bonehead]

12 hours ago, BEC said:

Chief,

I use a spectrum analyzer (Stanford Research Model SR1+) that sweeps 0 to 25K at 1 volt into the crossover input.  All outputs are terminated into the expected impedance with resistors and I run a trace of woofer, squawker and tweeter all plotted on top of each other.  I can see the electrical crossover frequencies and voltage output of each section.  I do that for every crossover I build or rebuild.

Mr bob....thats the way to do it.  do you have reference curves of the original klipsch nets to compare to?  

[Chief bonehead]

39 minutes ago, Marvel said:

So... is Klipsch palnning to sell electronic crossovers and amps as a package for the La Scalas and Klipschorns, so the time alignment issues can be dealt with. Even aligning my tweeter on my LaScalas made a huge and very noticeable improvement in smoothness, detail and imaging.

 

It's just a thought, although I am quite content with my passive crossovers.

 

Bruce

ummm no.  we could maybe if time permits but it would only be settings....just like the settings for the jub variants.

[Chief bonehead]

19 hours ago, mark1101 said:

 

Bob, Thanks for the explanation.  You were doing great until the last sentence.  You have always done me a great service but I'll be honest as I can be, Sonicaps sound lousy to me.  They are dry and brittle sounding and don't seem to change after many hours of listening.  I have been patient with them and given them the chance they deserve.  I have tried several different caps in various networks and every single one sounded at least a little different to me.  Some more than others.  In total I have listened to very few compared to the very large number available.

 

That said...........in all your testing have you found any other brand that matches the Klipsch FR signature, or did you stop at Sonicaps.  I have to believe that in 2016 another review is in order because Sonicaps just don't cut it, and the choices are astounding.

 

What would be great is if you could identify some of these that would provide the correct signature and offer some choices, or provide a running list for the forum as you find these out.

 

I hope you don't believe every cap that has the same ESR sounds exactly the same.  If you do, never mind.  

 

Please don't take any of this the wrong way.  This is about a constructive discussion that could go a long way for many of us.

sonicapphobic....:)

[Marvel]

7 minutes ago, Chief bonehead said:

ummm no.  we could maybe if time permits but it would only be settings....just like the settings for the jub variants.

 

Thanks for that, Roy. It was a serious question, although I know most folks probably just want to connect to their receiver or amplifier and be done with it.

 

Bruce

[mark1101]

18 minutes ago, Chief bonehead said:

sonicapphobic....:)

 

I know I am, and no hope for me probably.  I still buy the expensive Auricaps.  

 

But.............I have always used the EQ to adjust the FR.  I'm probably screwing something else up (so please don't tell me)........because I am happy.

[JL Sargent]

When I went from the old oil filled caps to SONICAPS in my Belles with type A crossover I lost something. I went from a pair of very smooth and NOT bright sounding speakers to a pair of speakers that needed additional work to sound good. Yes, the Sonicaps were much lower in ESR, I measured both the old and new caps in this regard. The speakers were much smoother with just the right amount of brightness with those old wornout caps. There was something special with that original oil filled setup. I can't even put my finger on it. PWK magic if you will.

[jimjimbo]

Things must be slow at the plant....

[Chris A]

3 minutes ago, JL Sargent said:

There was something special with that original oil filled setup.

There are more direct/better ways of controlling loudspeaker EQ than using high ESR (Equivalent Series Resistance) capacitors.  You could simply insert high-wattage resistors in series with the midrange and tweeter to drop the output, or use something like an L-Pad, or, even better, use digital EQ on your input to correct the in-room response. 

 

Chris

[Deang]

Sure don't remember things like I used, but I remember that.  If there is a "rest of the story", you might share it. Same with the resistor used to attenuate - you said to try different values "and see which one you like". 

 

Since you won't address comments or objections with any useful information, I think this is pointless. Baiting Bob is kind of sad too, as if the man actually has some original Klipsch AA curves stashed away in his shop. 

[Deang]

1 hour ago, wdecho said:

Bob I believe if you offered the option of adding L-pads on the mids and highs on the AA's you would have more happy customers. 

 

I would be really surprised if Bob jumps back in at this point. 

 

Bob won't use resistors or l-pads. Like PK, he doesn't believe these type of changes should be made at the loudspeaker, but with the use of tone controls.

[Chief bonehead]

3 minutes ago, Deang said:

Sure don't remember things like I used, but I remember that.  If there is a "rest of the story", you might share it. Same with the resistor used to attenuate - you said to try different values "and see which one you like". 

 

Since you won't address comments or objections with any useful information, I think this is pointless. Baiting Bob is kind of sad too, as if the man actually has some original Klipsch AA curves stashed away in his shop. 

dude i am going to call you 50 cent...you get most of your stories with me about 50% out but forget the other 50% i said....when my dad had his shop and replaced components for anything electrical he always had one to compare with to make sure that it worked as it should.  

[Chief bonehead]

2 hours ago, wdecho said:

Bob I believe if you offered the option of adding L-pads on the mids and highs on the AA's you would have more happy customers. There is no way to predict what a person's room is going to sound like.  They could turn the L-pads off or add as much or little attenuation they like. Just IMHO. 

L pads can make worse than better.  if the impedance is consistent in the band you are trying attenuate, then adding resistance like an L pad will cause the entire band to get louder or softer.  if the impedance is not consistent in the band, then what will happen is that some parts of the band will get louder than the other parts or softer.  in other words it will not get louder or softer linearly.  when i did the jub networks, i was able to insert a resistor and suggest changing it to get it louder or softer (salt or pepper) because it was linear.  got lucky i guess....

 

if one is sure of the impedance, having an autoformer that has 1 db taps is the next best thing....but then there are other problems....

 

oh oh.  i am sorry.  you asked mr bob and not me.  sorry.

 

[Deang]

3 hours ago, mustang guy said:

Now I'm sorry I asked. I found this: http://www.aast.edu/pheed/staffadminview/pdf_retreive.php?url=352_32115_EC434_2012_4__3_1_Chapter_03.pdf&stafftype=staffcourses

 

   

 

There is test equipment/software you can buy that does this. It runs a sweep, and then generates the output. The Bode 100 by Omicron Labs is nice - it runs about $6K.