Jump to content

Rules for cap bypassing? Forte ii crossover


roadworn

Recommended Posts

1 hour ago, Deang said:

[JBL] now do DC biasing in their high end stuff.

 

Last I looked into it they're using diodes to derive the bias from the signal instead of using batteries.

 

As I recall, only the first cap got bias in any event, not the second...

  • Like 1
Link to comment
Share on other sites

4 hours ago, Jeffrey D. Medwin said:

Hello everyone, here is what I find, over the last 10-15 years of actively experimenting and listening to the results : 

 

No one film cap will play the whole musical envelope linearly by itself.  No single film cap, at any price,  exists to do that. 

 

To get the full frequency bandwidth, and most dynamic range. and the most complete musical expression takes several high quality caps in parallel, for EACH capacitor position,  be it in a speaker crossover, active circuit position, or in a power supply.  The choices of what value film cap to parallel, is always determined by listening, and by having a specialized knowledge, of what VALUE caps play where, in the frequency range.

 

So, this calls typically for , instead of a single cap, ................anywhere from three, to as many as eight, film caps connected in parallel,.......in one capacitor position in a circuit. 

 

The smaller  the value of the film cap, the higher up in plays in the frequency spectrum, that is the general relationship.

 

The quality of the film cap also enters into the capacitor selection process.  Want the best full frequency playback?..... open up your wallets. 

 

What values of caps plays what part of the musical spectrum, is proprietary information, I am not free to divulge.  However, I will say I used anywheres from 5 uF ( a GTO Pulse cap ) to as small as 0.01uF  as part of a multiple all-film cap bypass value.  I don't use electrolytics.   Additionally, I would estimate 90% of the values are 1 uF or less.  Contrary to what has been posted here, I have found this applies to speaker crossovers also. A single cap can sound nice, but if you apply multiple film cap bypassing properly, you would never be satisfied, by not doing so.

 

I am not talking theory here, below is my A7-8 crossover, going into 2020, just used to roll off my tweeter, the " C " of a 12 dB L/C :

 

1855319595_P1010039edit8.thumb.jpg.2c9d8c9a767340e6255e000a9d72a141.jpg

 

     

 

    

 

The above technique applies to all good speaker systems.

 

Recall, I suggested selection is always done by ear, listening.   If you notice a perceived drop out, you must make adjustments, very often adding an appropriate value quality cap in parallel with existing uF values that are in place.

 

I find implementing this technique really well,  is what separates average from exemplary. YMMV, that is fine with me.

 

Jeffrey Medwin

Oh boy...... :rolleyes:

 

 

Link to comment
Share on other sites

10 hours ago, glens said:

 

Last I looked into it they're using diodes to derive the bias from the signal instead of using batteries.

 

As I recall, only the first cap got bias in any event, not the second...

Okay, that's very cool. So, why are they doing these weird things with capacitors if there is no benefit?

Link to comment
Share on other sites

My take on it is that it's primarily for the market in Japan...

 

9 hours ago, Jeffrey D. Medwin said:

Its just not important at to be spot-on a computed C value.  THAT is not good audio design and implementation at all, IMHO.

 

And with that, as just an excerpt, this thread draws to close.  We can no longer use math, physics, or any kind of science whatsoever to prove a point in the face of such declaration.

  • Like 2
  • Haha 1
Link to comment
Share on other sites

9 hours ago, glens said:

........

 

And with that, as just an excerpt, this thread draws to close.  We can no longer use math, physics, or any kind of science whatsoever to prove a point in the face of such declaration.

 

Glens, I am afraid that sanity was rejected quite some time ago. Look at how many folks around here mate woofers with cabinets. Math, engineering, and physics would only get in the way.

Link to comment
Share on other sites

17 hours ago, glens said:

My take on it is that it's primarily for the market in Japan...

 

 

And with that, as just an excerpt, this thread draws to close.  We can no longer use math, physics, or any kind of science whatsoever to prove a point in the face of such declaration.

Kind of like todays politics.  Facts have no place.

Link to comment
Share on other sites

2 hours ago, Deang said:

What exactly do you mean by exactly?

I agree with your comment. (So, why are they doing these weird things with capacitors if there is no benefit mr. medwin? ) 🙂

Clearly Its not all voodoo opinions, There is some science and probably electronic rules behind adding different caps to caps to get a prefered voicing of a speaker. Which is why I started the thread. While I enjoy the Voodoo portion of all this ie: what I think sounds good vs other people. This fourm has an incredible amount of talented people that have already done many times what Iam doing for the first time... so Yes Voodoo good .. yes science good. I love it all, but I need to know what horrible mistakes to avoid. like adding two  .01uf  $35  caps that will in effect add nothing to my listening pleasure. And of course this is based on my humble system (odyssey Khartago amp, and Candela pre, and my soon to be rebuilt Forte ii's) Perhaps a $50k system would find those two little caps completely Mortifyingly horrible. 🙂

 

Link to comment
Share on other sites

On 2/8/2020 at 1:31 PM, Jeffrey D. Medwin said:

Its just not important at to be spot-on a computed C value.  THAT is not good audio design and implementation at all, IMHO. 

 

22 hours ago, glens said:

My take on it is that it's primarily for the market in Japan...

 

 

And with that, as just an excerpt, this thread draws to close.  We can no longer use math, physics, or any kind of science whatsoever to prove a point in the face of such declaration.

Hold on, almost every comment in this thread agrees that sticking to an exact value isn't critical, but when Jeff says so, all of a sudden we're all off to the dark ages? I mean, yeah some hyperbole, but his comment mostly makes sense. Look at the frequency response graph of any crossover; it is not perfect, nor are the "exact" values really exact. They are a combination of "close enough", available, and cost effective, and in the end it is the listening test that determines a successful design. Improving that design could definitely involve mixing type and variations in value. Response graphs will always come second to my ears.

  • Thanks 1
Link to comment
Share on other sites

[Edit: Wasn't this thread merged with 

yesterday when I posted this?  At any rate, this above is the thread of reference below.]

 

Now that this thread has been merged, the "proprietary" values for different frequency ranges can be found if they hadn't been edited out since.  I believe it went something like (paraphrasing) "if you detect (by ear) a small suckout at 2kHz, then you add a 1 uF cap to the bundle."  For different frequencies, different values are specified, some of them so small they can't possibly affect anything one hears.  I took him to task at the time about that, also asking if he decreased the size of the large ones to compensate.  Which is when I first saw him say that target "C values" are really immaterial.  At any rate, he's not saying "12 uF spec'd with 11.5 or 12.5 is cool," rather, he might wind up with 15 or 16, hell maybe even 20 by the time he's done, and that's okay because, well, you really should back up now 20 pages or whatever and see for yourself.  It's just crazy...

 

I surmise he's conflating value-by-frequency for high-impedance amplifier internals with those for low-impedance crossover networks.  The wild part is that it still "works" the same for him in the crossover as in the amp, facts be damned.

  • Like 2
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...