ALKing with Hoggy

[Deang]

Just like any other network, don't stray too far from what was modeled. Depending on what part of the circuit -- small changes can have dramatic, and not so wonderful results. You would need eight 10uF Hovlands for both boards, spend $360.00 to get them, and might end up with a treble that pins you to the wall. It might not be too much fun either trying to solder all those leads together. Too much heat for too long and you can kiss your $360.00 goodbye. Keep it simple -- just get a good quality 40uF metallized polypropylene, or two good quality 20uF's of the same and and parallel them. There is no need to go nutty here. Hell, that much money and you're within striking distance of a set of Al's ESN's or Klipsch AK-4's!

Just go with 6.8uF Auricap or Kimber for the other.

Auricaps can be purchased through www.percyaudio.com or http://www.partsconnexion.com/catalog/CapacitorsFilm.html The Kimbers, which can be bought from www.partsexpress.com are nice too. I use them as well.

As far as "conditioning caps" goes, it might be BS, or it might not be. However, does doing it hurt anything? No -- so why not? It only takes a few minutes. I use a bank of 9V batteries in series to do it.

http://ldsg.snippets.org/appdx-ec.php3#CAPS

http://members.aol.com/sbench102/caps.html

[tigerwoodKhorns]

The Partsexpress site lists the following for the Auricaps:

* These are SIDEREAL Capspredecessors to AURICAPindentical, except lead material is solid-core, not stranded copper.

Is this accurate?

The otehr sites don't list the Auricaps in values that I will need. Are the sidreals teh same as teh Auricaps?

Thanks,

Chris

[Al Klappenberger]

Guys,

Get this capacitor "conditioning" idea out of you heads! It makes no differende at all! If you think you hear a difference you are simply proving the fact that the human brain hears what it wants to hear. This is the norm, not an exception! Putting two 20 uF caps where a 40 uF cap is called for WILL make a difference. This will reduce losses, but I am not sure you could actually hear that difference though.

Al K.

[Chris Robinson]

Al,

Can you help us understand WHY cap conditioning makes no difference?

Merci.

[Al Klappenberger]

Guys,

Here's some evidence to support my claim that the 6.2 uF + 1 uF

capacitors in the tweeter filter of my network is the least important. The

attachment is a text file that is the output format of the power analysis

segment of my filter design program (PCFILT). You may have to view this with

a text editor. It won't look right with a fancy word processor. It shows a

representation of the tweeter filter with each section or component shown as

a box. Around each box is the voltage across it and the current through it.

The actual loss (in watts) is shown inside the box. This is loss due to

component "Q". The "branch" numbers appear at the far left. Branch 3 is the

.20 mHy inductor, for example.

The input level was adjusted to provide 2V RMS from a 1/10 Ohm

source. The output (Termination) is 8 Ohms. The frequency is 6000 Hz, which

is the crossover frequency. It the crossover, half the power goes to the

tweeter and the other half to the squawker. That is, 3 dB loss to the

tweeter. The output voltage is 1.41V. So: 20 * log10(1.41 / 2.0) = -3.04 db.

All of the capacitors are assumed to have a equal "Q" of 1000.

Now compare the voltage across the 3 parts in the tweeter filter:

6.2+1uF has .649V. The .2 mHy inductor has 1.553V across it. The 2.2 uF

Hovland has 2.55V across it.

Now compare the actual loss within each part. The biggest loss is

through the .2 mHy inductor. It is wasting 8 mW (Milliwatt). This is to be

expected. Inductors are ALWAYS the weakest link. The loss in the 6.2+1uF cap

is to low to display and shows as 0. The 2.2 uF Hovland, on the other hand,

is wasting 1 mW to spite that it is assumed to have the same "Q" as the 6.2+1

uF combination.

The 39+1 uF capacitor is also wasting 0 W. This may not be real

because the "Q" is probably NOT 1000 at 6000 Hz. This is why I think

upgrading the 39+1 combination to two 20 uF caps of higher quality may be

worth doing.

Al K.

Anl.txt

[Al Klappenberger]

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

On 1/18/2005 7:58:27 AM Chris Robinson wrote:

Al,

Can you help us understand WHY cap conditioning makes no difference?

Merci.

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

First off, this is like UFO close encounters. The advocates of conditioning should prove that it does make a difference rather than me prove that it does not!

BUT: I tried it and found it makes no measurable difference and have oscilloscope screen photographs to prove it.

Also: a capacitor stores electrons in one side, then the other. What difference does it make if it's doing it at 20 cyles per second or 20 seconds per cycle! Every half cycle it completely forgets what it was doing the last half cycle. "Conditioning" is just running it at 20 seconds (or so) per cycle with a square wave. Nothing more.

Al K.

[tigerwoodKhorns]

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

On 1/18/2005 9:05:46 AM Al Klappenberger wrote:

Now compare the actual loss within each part. The biggest loss is

through the .2 mHy inductor. It is wasting 8 mW (Milliwatt). This is to be

expected. Inductors are ALWAYS the weakest link.

upgrading the 39+1 combination to two 20 uF caps of higher quality may be

worth doing.

Al K.

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

AL,

I am going to try the two Auricap 20's in place of the 39 + 1. I will use a 6.8 Auricap in the other spot. Shoudld I use a bypass (maybe a 0.1 mHy)?

A quick inductor question:

The Solen solid 2.4 mHy inductor is $19.14.

The Solen Litz 2.4 mHy inductor is $23.87.

Is it worth trying the Litz? Does the litz have to be mounted on its side to cool? (I am moiunting teh crossovers outside of the speaker so this will not be a problem).

I will also be custom fabricating the junctions out of brass. My other choices are stainless or aluminum. If necessary, I can get copper but it will be difficult. What is your reccomendation.

Thanks,

Chris

[Parrot]

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

On 1/18/2005 9:13:06 AM Al Klappenberger wrote:

Every half cycle it completely forgets what it was doing the last half cycle.

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

This happens to the best of us as we get older.

[Al Klappenberger]

Chris,

Don't worry about the 1 uF bypass caps with the Auricap substitute. Put a 7 uFd in place of the 6.2+1 and 20+20 auricaps in place of the 39+1. I am not convinced the bypass cap does anyhting significant when its connected around a polypropylene cap. I use them just to "make sure" around the 39 and to get up to 7 uF around the 6.2.

There is no advantage to Litz wire at woofer frequency. It's a waste of money. There is also no heating problem with any of this. The only thing that may get hot is the woofer voice coil!

BTW: If you are willing return one of those modified networks to me along with the paperwork, I'll measure it and see if I can see any difference with instruments.

Al k.

[tigerwoodKhorns]

[Al Klappenberger]

Chris,

Interesting post!

Al K.

[tigerwoodKhorns]

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

On 1/18/2005 3:34:24 PM Al Klappenberger wrote:

Chris,

Interesting post!

Al K.

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

Al,

I am a man of few words (if any!!).

For some reason the rich text does not post from my computer.

After pulling ip a periodic table, I have decided to use copper plates to fabricate the connectors instead of brass.

As far as heat goes, is it a concertn at all (specifically, can the crossovers be enclosed in a case)?

Thanks,

Chris

[Al Klappenberger]

Chris,

What "connectors" are you talking about? The binding "L shaped" posts? You can get them from Mouser electronics.

Yes, you can but them inside an enclosed space. The only part that might get warm is the 10 W 10 Ohm resistors and that's only if you are trying to annoy the neighbors!

Al K.

[NOSValves]

Al,

Man I haven't seen you post this much in a long time ! Good to see you posting.

Craig

[tigerwoodKhorns]

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

On 1/18/2005 4:54:18 PM NOSValves wrote:

Al,

Man I haven't seen you post this much in a long time ! Good to see you posting.

Craig

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

And the value of the information is outstanding!!

Al,

I am going to fabricate the small parts at a friend's shop. Strictly eyewash but I am extremely anal retentive (can I still say that here?).

I want to buy everything from Parts Connexion. What transformer will I need or should I stay with the parts from Texas?

Craig,

Please see my post in 2 channel entitled

"Dean G / Bob Crites; Type A question for my HT Setup"

I want to build Type A's for my HT system using Hovlands and Auricaps. I see that you had a set of these and wanted some info.

Thanks,

Chris

[Al Klappenberger]

Chris,

As to the transformer, you best stick with the Texas source.

What do you mean by build a type A? My type A or the Klipsch type A?

I think I asked for this confusion by calling it the "Universal type A replacment"!

Al K

[Deang]

I will also be custom fabricating the junctions out of brass. My other choices are stainless or aluminum. If necessary, I can get copper but it will be difficult. What is your recomendation.

18K gold.

[tigerwoodKhorns]

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

On 1/18/2005 7:21:34 PM Al Klappenberger wrote:

Chris,

As to the transformer, you best stick with the Texas source.

What do you mean by build a type A? My type A or the Klipsch type A?

I think I asked for this confusion by calling it the "Universal type A replacment"!

Al K

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

AL,

I am using your design in my 2 channel setup with K-horns. I also want to put new crossovers in my Belles that are part of my Home Theater system. I thought that a Klipsch Type A using Auricaps and Hovlands woudld be really cheap to build for that setup, if it sounds good.

However, I read that the Type A network is designed for Paper in Oil and does not like the reduced resistence of the Hovlands and Auricaps. If I have to go Paper in Oil, then the ALK's look like the way to go as the price will be about the same and I wil get the adjustable squaker and constant impedence benefits.

Speaking of the transformer, I am planning on buying new ones for the K-horns. How do I test the ones in my Belles and Heresys? If they are good, I can avoid having to buy 6 additional transformers.

Thanks,

Chris

[Deang]

However, I read that the Type A network is designed for Paper in Oil and does not like the reduced resistence of the Hovlands and Auricaps.

I didn't say the speakers wouldn't "like" the reduced resistance of the Auricaps. However, I did say the ears might not like the near zero loss of the Hovlands in those older networks. However, many run them and seem to like them just fine. I like the thick film metallized types because their ESR's fall between the PIO and the Film and Foils. They give you clarity without going sterile.

Speaking of the transformer, I am planning on buying new ones for the K-horns. How do I test the ones in my Belles and Heresys? If they are good, I can avoid having to buy 6 additional transformers.

If you are going for as good as it gets -- then buy the six UT 3619's. However, the stock T2A's you currently have are fine. BTW, if the speakers work, then the autotransformers are doing their job.

[Deang]

Wrong file -- ignore.