Best type of capacitors for recapping Type A Xovers?

[Deang]

"...bypassing with a tiny value like that makes absoultely no sense in any way. "

You had a great explanation for that too. It does appear that using it in crossovers is probably a waste of time and money, but I haven't done any critical listening doing it. I guess it's still commonly used in tube amp power supplies with the big value electrolytics -- but I don't know why exactly. Craig?

Well, there's enough in this thread along the quality/price end of things that has me wanting to re-evaluate the way I'm building, but when you have something that sounds great and is working good -- you hate to mess with it. Sonicaps in place of Kimbers and Auricaps definitely looks like a good bet though. I know Jeff knows caps, and the guy spent a lot of time on the phone with me teaching me stuff when I first started doing this -- which ironically wasn't even with Klipsch speakers -- but Dahlquist DQ-10's. I hadn't really thought about it, but Jeff is probably the main reason I'm such a cap bigot. He also turned me on to the Thetas, and unless something has changed I believe it's still his favorite cap.

I'm too hard on the GE cans -- they sounded good when I tried them. It's probably more paranoia than anything else. I do think they make things a bit incisive when the volume gets going, but there's probably only a half dozen or so around here that enjoy music at the SPL's I'm prone to engaging in from time to time anyways.

Craig, in the smaller values that you use, you might be interested in knowing that the Thetas are double wound.

JC, the Solen F/F are good caps (white ones). I've used them often and wouldn't have a bit of reservation substituting them in place of Thetas.

Tofu, still want those parts?

JC, I use the Erse air cores you're talking about. Al wants me to quit using them and go back to litz. Bob says that in the Klipsch networks it's a waste of money. Al says in the Klipsch networks anything is a waste of money. If you are building ALKs, you should use what Al has spec'd out. If you are building a Klipsch network, you should use what Klipsch spec'd out and get the inductors from Bob. I'm straddling the fence with coils, and my *** is starting to get sore. My personal experience with this stuff tells me the AC component of an inductor needs to be accounted for. Litz has more resistance, and what I've decided is I like the results in the Super AA design using 14AWG Erse Air cores. It might just be because the filter has a lower part count than the ALK so I can cheat a bit.

Al, you could use Auricaps or Kimbers -- but they ain't as cool looking.

[tofu]

Now I got a question: I got a pile of base boards made on a CNC machine for my Universal network that have a hole cut for a terminal post just in the right place for the lead length of the Hovland 2.2 uF cap. What reasonably priced caps have stranded wire leads of the same or longer length as the Hovlands? I don't want to drill holes and add another terminal post just to change cap brands! If I just moved that one terminal post the leads of the 6.2 + 1.0 wouldn't be long enough.

Al K.

well, al, i'm sure you get discounts for buying in bulk, but from what i priced out, the hovlands cost ~$80 four, versus using four fastcaps at $6.88. is drilling holes REALLY that bad? []

[tofu]

Tofu, still want those parts?

i sure do! but due to this thread, i've had to make a bit of a revision to my list. i think this is sufficient to make a good ALK universal network.

2 2.40 mHy inductor wound of AWG #12 solid $63.18

2 .30 mHy inductor wound of AWG #14 Litz wire $18.38

2 .20 mHy inductor wound of AWG #14 Litz wire $15.14

4 20 uFd. +-5% Solen "Fast Cap" $22.44

2 6.8 uFd. +-5% Solen "Fast cap" $5.28

4 2.2 uFd. +-5% Solen "Fast caps" $6.88

comes out to about $131 + shipping/handling. let me know if i made any errors pricing out the caps/inductors, or if you suggest any changes.

would a Money Order be all right with you? my spending limit on paypal is down to around 40 bucks.

[Al Klappenberger]

Tofu,

Your parts list will definitly make a good crossover. The important parts to get right are the .20 and .30 Litz inductors. The pricing depends on where you get them but they won't be as expensive as Hovlands by any sight!

Al K.

[BEC]

Man, we are getting too close to agreement on this cap thing. Guess we need to get started on inductors.

Bob Crites

[Al Klappenberger]

Guys,

Here's another factor to argue over: Value tollerance. One thing I like about the Hovlands and the Solen FastCaps is that they are +-5% tolerance and they really do fall within that range. I test every one I use. I just looked at Auricaps and Kimber Kaps. The are both speced at +-10%. That would make making matched set more difficult. All my Universal networks are build with the caps matched to within 1% of each other between pairs. If they caps I use are +- 10% that's going to make the spread wider. I match them using an lcr meter connected to a computer program via an RS232 port. The program just sorts them low to high and I divide them into groups of two. If any one is beyond +-5% the computer flags it.

Al K.

[jwc]

Dude!

2 2.40 mHy inductor wound of AWG #12 solid $63.18

2 .30 mHy inductor wound of AWG #14 Litz wire $18.38

2 .20 mHy inductor wound of AWG #14 Litz wire $15.14

That is a fabulous quote. Is the solid one the S122.4?

Damn. If you won't buy em, I will.

[tofu]

Dude!

2 2.40 mHy inductor wound of AWG #12 solid $63.18

2 .30 mHy inductor wound of AWG #14 Litz wire $18.38

2 .20 mHy inductor wound of AWG #14 Litz wire $15.14

That is a fabulous quote. Is the solid one the S122.4?

Damn. If you won't buy em, I will.

yep, those are the prices from the solen site. al's schematic calls for a awg#14 for the woofer, but dean recommended i buy the #12 or #10 if possible. #10 was a bit too rich for my blood at $50 per inductor.

[tofu]

Tofu,

Your parts list will definitly make a good crossover. The important parts to get right are the .20 and .30 Litz inductors. The pricing depends on where you get them but they won't be as expensive as Hovlands by any sight!

Al K.

thanks for the words of confidence. makes things a lot easier knowing i can use 100% solen parts instead of having to shop around northcreek and hovland.

[Al Klappenberger]

Tofo,

I get the inductors for even less! (Nah-nah nah na-na!) []

Al K.

[tofu]

pff []

[NOSValves]

Guys,

Here's another factor to argue over: Value tollerance. One thing I like about the Hovlands and the Solen FastCaps is that they are +-5% tolerance and they really do fall within that range. I test every one I use. I just looked at Auricaps and Kimber Kaps. The are both speced at +-10%. That would make making matched set more difficult. All my Universal networks are build with the caps matched to within 1% of each other between pairs. If they caps I use are +- 10% that's going to make the spread wider. I match them using an lcr meter connected to a computer program via an RS232 port. The program just sorts them low to high and I divide them into groups of two. If any one is beyond +-5% the computer flags it.

Al K.

Al,

Do yourself a favor and get ahold of Jeff at http://www.soniccraft.com/ or Danny at http://www.GR-research.com/ talk to either one about the quanity you will be purchasing and they will work out a great deal for you. The Sonicap is a 5% cap and IMHO every bit as good as a Auricap the only difference is it uses a solid tinned copper lead. I have no clue why anyone would prefer those pain in the butt stranded leads. The leads are pretty long also but you would have to ask them about the values your going to need.

Craig

[NOSValves]

Man, we are getting too close to agreement on this cap thing. Guess we need to get started on inductors. Bob Crites

That is only because I have been down working all day[] I'm still not in agreement at all. I think caps make a noticable difference. Can I prove it with some test equipment?..... Nope. But I can and do hear the difference just about everyday. If its my imagination... then I like my imagination and god I hope it stays.

Craig

[Deang]

I don't agree with anyone, especially Craig.

[NOSValves]

Why thank you Dean that makes me feel great!!

[BEC]

Dean,

Did you talk to the crossover customer I referred to you? Or are you not checking email again.

Bob

[jwc]

I will agree that this is one of the best damn threads I've seen here in weeks. Weird how such a simple thread topic set it off.

Listenen to "Crowbar" really loud on my Cornscalas (sbb) and there is no fatigue even though this is some really thrashy fatboy heavy metal. Now why is this important? Well, I have some Dean G'ed ALK B's in these and there is no fatigue at all. On this setup, it is all Solen parts. Two of the capacitors are F/F. I am mentioning this because I think design plays a role as well as stereo components. If there was some yellow caps in there, I bet It would make little difference.

On the other hand, let's consider some of the basic first order networks which is somewhat how this thread was started. Let's say I stuck a type "B" network in there. One with fast caps and the other one with Hovlands. I bet there would be a noticeable difference between the two. Unlike the Hovlands added to the ALK "B" network.

Could it be plausible that the more "complexity" of the network may make it less sensitive to lower cost capacitors. Or is there a better word than "complex" that would make my arguement believable.

I'm gettin ready to build some ALK Universal A's, Super AA, and Type B's vs B-3. It would be great to get away with "fast caps" from Solen.

jc

[BEC]

jc,

I think that if more complex means more caps in the signal path, it is more important to have good caps in the more complex crossover. Again, a good cap is one with capacitance the right value, ESR very low (in the hundredths of an ohm) and adequate working voltage.

Bob Crites

[Erik Mandaville]

Bob:

"Again, a good cap is one with capacitance the right value, ESR very low (in the hundredths of an ohm) and adequate working voltage"

I agree completely. My point through this entire thread had to do with the idea that cost, regardless of the kind or brand of capacitor, really has nothing to do with what a person prefers. I brought up the possibility of using poly/oil caps because I think they are excellent capacitors for a reasonable cost. I also notice no difference in their behavior at different volume levels. If an owner is interested in a crossover upgrade that is more in keeping with what the speakers originally came with, the oil cans are of course a great way of doing that. Klipsch evidently thought well enough of their perforamance in the Heritage line to make extensive use of them.

I also appreciate what Al mentioned about this. I have always thought Solens to be extremely good capacitors, and there's a very broad range of values and working voltages available.

Dean: Let me explain the autoformer variable L-pad issue again. For some reason it got snarled up in this thread about capacitors. The reason the variable control for the squawker came into existence had to do with the fact that Craig wanted a way to manually turn down the level on the mid-horn. You probably remember a million posts ago where he said his ears were being 'ripped off' and he wanted to know what was going on. A few days later he arrived at our front door, carrying the entire top-hat of one of his horns.[] In addition to telling me my system was also missing some important LF info. (he's right about that), we found that the squawkers in my klipschorns sounded very similar to his. In any event, he wanted to reduce the output, but his networks (if I remember this right) had fixed inputs and outputs on the autoformer. So, my solution was the variable L-pad. In terms of performance, the two are the same to me. My argument about it's use had to do only with ease of use compared to the fixed input/output connections on the autoformer. Do I think they are 'better' in terms of performance? No. All they do is allow the user to turn the volume up or down until the balance seems right (also a subjective judgement). That's the only thing I did to his networks. The variable control also maintains a constant impedance in relation the driver it's used with, and that's a good thing.

Caps and coils: As Al also mentioned, all the books I have on crossover design stress the importance of good inductors. Capacitors, as long as they have the characteristics Bob outlined above, are almost taken for granted -- a decent mylar or other plastic capacitor on the tweeter, and that's about it. Some even specify non-polarized electrolytics, which can be perfectly suitable for woofer use. I think motor-run oil caps are also fine, and I have them in my own networks -- which BTW, will soon have brand new autoformers. Since it hasn't been done yet, I want to work out a switching device to enable switching between the settings.

Erik

[Erik Mandaville]

Al:

You might try these. I've copied a 2uf Kimber Cap value from Audio Electronic Supply just as an idea. These are still expensive, but not as costly as Hovlands.

http://www.audioelectronicsupply.com/cgi-bin/audioelectronicsupply.com/view_services.cgi?request=detail&dept_id=13&aisle_id=137&prod_num=CAP_KK_2_200

Erik