Charge coulpled L96/3313B crossovers?

[briang]

After much pondering, I've decided to ask if anyone on this forums wouule be interested in helping me with creating a pair of charge coupled 4313B crossovers for one of my L96 pairs?

Who is up to the challange?

I did plenty of soldering in my undergraduate physics courses, so I think I have the skills to do some of the grunt work; however, layout of the components on a board and component choices are best left to the experienced.

Who wants to help?

[Zilch]

Link us to the schematic, please.

[I ain't huntin' it down.... :blink: ]

[sourceoneaudio]

Briang,
Here is a link to a value calculator this might help on your component/value search.

https://selectproducts.com/calculator_bandpass.html

J/S-S1A :D

[briang]

Link us to the schematic, please.

[I ain't huntin' it down.... :bouncy: ]

Thanks Zilch, I'll take that as a "yes, I'll help" :applaud:

In another post, Giskard created this amalgum of the L110, L110A, L96, 4313 and 4313B filters. I offer my thanks to Giskard for this photo and his other contributions.

[Zilch]

Thanks Zilch, I'll take that as a "yes, I'll help" :applaud:.Well, no, it was more like "Yes, I'll take a look." :p

You're wanting charge coupled 3113B, as I understand it.

That's not terribly tough, as there's only four capacitors, none of them inordinately large values.

Let me draw up the schematic, maybe tonight, then; there'll be some design choices in that, even.... :thmbsup:

[Robh3606]

Hello Briang

All you have to do to convert is to double the value and add a second cap of the 2X value. So to change a 1uf you would double it to 2uf and add a second 2uf cap. You then put the resistor at the junction of the 2 caps so you end up with a T on your schematic. The 2-3 meg resistor goes to the + on your 9 volt battery. The negative on the battery goes to ground. It's really easy to do. If you can read the schematic you can do this. Do a search and take a look at Giskards posted schematics. Solens are a good start and they are reasonably priced.

Rob:)

[briang]

Thanks to all those offering assistance.

I'm also concerned with how all the additional caps, resister and 9V will be accomodated from a real estate and organization/placement standpoint.

I'll likely use the existing L96 crossovers (if practical) and solder/desolder as needed to create the new filter circuits. If I need an additional bread board, I can do that too...

Thanks, I look forward to seeing these finished and hearing them...I blame all my fellow JBL afflicted friends here for my desire/disease to try new things and charge coupling...;)

[Zilch]

I'll likely use the existing L96 crossovers (if practical) and solder/desolder as needed to create the new filter circuits.Not likely.

Put up a pic of the factory crossover....

[briang]

Not likely.

Put up a pic of the factory crossover....

When I get home from work...;)

[Zilch]

Check me, please. Separate battery in LF is only required if you biamp; otherwise they can all run off one (through their individual 2.2M Rs).

You're gonna have $40+ per crossover in 33uF caps alone. Still want to go ahead with this?

Do you intend to biamp?

Link me to the discussion by Giskard. Let's see what he said about doing this....

[briang]

Check me, please. Separate battery in LF is only required if you biamp; otherwise they can all run off one (through their individual 2.2M Rs).

You're gonna have $40+ per crossover in 33uF caps alone. Still want to go ahead with this?

Do you intend to biamp?

Link me to the discussion by Giskard. Let's see what he said about doing this....

Go ahead = yes.
Biamp = nope.
Link= http://audioheritage.org/vbulletin/showthread.php?t=9854

[Zilch]

O.K., then. Do you have a micrometer or dial caliper to measure the wire diameter in L2, 3, and 4, or means to accurately measure their DCRs?

I can kinda guess from the pic once it's up, but it's better to know.... :thmbsup:

[4313B]

I have a bunch of the 3.5 mH 7.5 ohm coils since I used them all the time for custom installations of these systems. You will gain nothing by using anything else. The 0.30 mH has a DCR of ~ 0.3 ohms and the 0.75 mH has a DCR of ~ 0.6 ohms.

[Zilch]

Thank you, Giskard!

Brian, you should acquire a pair of the L3 inductors from Giskard.

I would advise against scavenging from your original crossovers until these are complete and you are satisfied with them. You may also want to be able to reinstall the originals at a later date.

I'll work up the rest of the parts list this weekend....

[X_X]

Hello Briang

All you have to do to convert is to double the value and add a second cap of the 2X value. So to change a 1uf you would double it to 2uf and add a second 2uf cap. You then put the resistor at the junction of the 2 caps so you end up with a T on your schematic. The 2-3 meg resistor goes to the + on your 9 volt battery. The negative on the battery goes to ground. It's really easy to do. If you can read the schematic you can do this. Do a search and take a look at Giskards posted schematics. Solens are a good start and they are reasonably priced.

Rob:)

Hi Rob,

The neg on the battery terminal goes to neg on the drivers, but NOT the neg on the main input. Correct? Just making sure. I'm doing 3145 CC networks and the phase is inverted, it doesn't make sense to tie all negs together. Yes or No?

Are there any advantages over using Solens? I'm guessing the charge coupling "effect" lowers the network's overall inductance through battery power and the mutuality of the caps, correct? Will "better" caps really be a marked improvement? Will different caps (barring Solens) need to be non-inductive?

Thank you,
Nathan.

[Robh3606]

Hello Nathan

That goes to ground, negative input ie schematic negative of the drivers, all the same point, not the individual driver negs which can be used in either polarity depending on the system design.

http://www.lansingheritage.org/html/jbl/specs/home-speakers/1993-k2-s5500.htm

JBL uses CC Solen networks in the Everest. There was quite a row about that and if the Solens were good enough. My point of view is the designer is happy with this. He knows what he's doing and I would make them a starting point as they are relatively inexpensive. Have a listen and if you feel the need try the more expensive audiophille caps. Send Ian a PM as he has tried different caps in his 4345's for an opinion. I have a pair of 4344 Ian built the networks for using Holovands?? and they are sweet!!! Bottom line is you really need to give it a go yourself to see which you like best.

Rob:)

[X_X]

Thanks Rob! :)


When you are building a cost no object speaker and Solens was the capacitor of choice (in an area where ANY big$ capacitor could have gone in there), it is testament to both the quality of the brand, and the subjective nature of audio. I already wired in Solens, so I'll know soon enough.

Is there any advantage to bypassing capacitors in a charge coupled network?

Nathan.

[Robh3606]

With the Solens the word is they sound better without when run in a CC network.

Rob:)

[Zilch]

I've posted an oscillograph of charge-coupling in action in these forums.

It moves the operating point to 9V, and signals with negative peaks less than that don't go through zero. Thus, there is no charge reversal or zero-crossing distortion.

In theory, it would work just as well around -9V if the battery is reversed, which I've done inadvertently on occasion. One side goes to "common," is all, assuming there is one in the circuit. Audio is AC.

Bypassing operates differently, but is generally regarded as unnecessary when biasing is used.

I quit bypassing when I switched to Solens, anyway....

[Zilch]

Here's the tentative parts list. Criticism welcome:

[I'm scanning a board layout. It's 9.5" x 6.0". Takes room to charge-couple.... :yes: ]

[Nightbrace]

Sorry I didn't notice this sooner, I will get pics of my nets and the parts list when I get home. I didn't know you wanted to get started so quickly.

I can't really comment on the actual building or design theory behind these networks, as I did little more than pay for them and plug and play. All I know is that the sound is fantastic. And I can't imagine why the L96 cannot benefit from the same networks as my 4313B's.

[Nightbrace]

With the Solens the word is they sound better without when run in a CC network.

Rob:)


Hmm, I'm pretty sure mine are Dayton :(, but they still sound good.

[Robh3606]

It moves the operating point to 9V, and signals with negative peaks less than that don't go through zero. Thus, there is no charge reversal or zero-crossing distortion.

Hello Zilch

With an AC signal why would the peaks effect the DC bias?? I always thought they were mutually exclusive and the AC would ride on the DC. The DC component of the voltage sums would remain constant.

Rob:)

[Nightbrace]

I'll likely use the existing L96 crossovers (if practical) and solder/desolder as needed to create the new filter circuits. If I need an additional bread board, I can do that too...

..

Your gonna have to buy all new components, I kept my 3113B networks in case I ever wanted to go back to the stock in the future or if I ever decide to sell them as I doubt I'd get more for a speaker with modded nets.

I wasn't terribly disappointed with the 3113B nets, but the CC nets are better in almost everyway and worth the effort. I gotta spend more of my time learning about network design as the improvements that can be made in a simple system with GREAT transducers like your L96's is nothing short of spectacular. I never really was "into" a contemporary network design till I swapped these in.

[Nightbrace]

Hello Zilch

With an AC signal why would the peaks effect the DC bias?? I always thought they were mutually exclusive and the AC would ride on the DC. The DC component of the voltage sums would remain constant.

Rob:)

They operate independently.

[Zilch]

Hello Zilch

With an AC signal why would the peaks effect the DC bias?? I always thought they were mutually exclusive and the AC would ride on the DC. The DC component of the voltage sums would remain constant.

Rob:)Yes, that's correct. With the AC riding on +9V DC, there's no zero crossing until the negative AC peak is -9V. Typically, HF alone never gets there:

http://audioheritage.org/vbulletin/showthread.php?p=130034&#post130034

It's also easy to see that big LF would carry the HF along with it through zero. Fortunately, the filters usually separate them.... ;)

[Nightbrace]

Yes, that's correct. With the AC riding on +9V DC, there's no zero crossing until the negative AC peak is -9V. Typically, HF alone never gets there:



What will happen should it get to -9V?

[briang]

Here's the tentative parts list. Criticism welcome:

[I'm scanning a board layout. It's 9.5" x 6.0". Takes room to charge-couple.... :yes: ]

I'll put the order in to PE. Thanks Zilch!

Now, what is your favorite beer?

[UreiCollector]

any thougts on raising the "bias" voltage to 18v or 27v? We need to stay within the ratings of the capacitor of course, but why not bias higher? I'm curious on thoughts regarding this. Discuss......:)

[Zilch]

The capacitors typically used are 200V minimum, so that's not an issue.

I believe Earl K and/or Tim G have tried other bias voltages. Check their posts on this subject for some info.

There was talk of using LOWER voltages, actually, as I recall.... :dont-know

[briang]

I guess it is time to get some LE5-9 to put in place of the LE5-12...or does it really matter? The LE5 library shows more efficiency from the flat wound wire in the LE5-9 Voicecoil.

Opinions, or experience is welcome...

[Zilch]

Similar factory L112 crossovers, one less capacitor, actually, and not biased.

We'll see what Brian comes up with, now.

["Chunky".... ;) ]

[Earl K]

I believe Earl K and/or Tim G have tried other bias voltages. Check their posts on this subject for some info.

There was talk of using LOWER voltages, actually, as I recall....
__________________


- I DC bias my polystyrene based PLLXO ( going into my Bryston 2BLP ) at @ 6 volts . This amp feeds signal to my Altec 288-8K drivers / going first through a HF contour circuit and a bit more HiPass filtering .

- My main ( passives ) 13.5 uF caps in the 2 pole HiPass are biased to around 3 volts / the 2.5uF HF bypass caps are biased at around 3 volts while the .09 uF Polypropylene & Polystyrene small value bypass caps in this circuit are DC biased to 9 volts.

- I ran these bypass caps ( .09 uF ) with about 4.5 volts ( but not attached to the battery ) for about 5 months ( these small value bypass caps don't share a common tie point with the main 2.5 uF ( HF bypass ) caps, ie ; they have their own independant feed resistor ).
- Attaching the battery for a constant DC load makes a bigggg difference in clarity and depth .
- In hind sight ; It's interesting how AC can kick the knickers out of any DC potential, in the absence of a constant DC source ( all very enlightening if one likes to tweak caps and circuits ) .

- My current "cap" recipe gives a wonderfully detailed & sweet sound with my gear . YMMV ... & FWIW ; my main caps are more or less, unobtainable ( since they are surplus ) so you'll have to find your own pleasing cap recipe .

- Yes, buy Solens ( one needs to start playing with DC biasing somewhere ) / just don't fall victim to believing they are the "be all end all" .

:)

[Aaron]

Sorry to drag up an old thread, but it answered nearly all of my questions...so why not keep all the answers in one place, eh?

The big question:
The N96 X-over has no cap for the LE10H... should I just leave it alone?

Robs instructions in post #6 are very clear on what to do if there ARE caps in the circuit.
But what if I'm faced with just an inductor?

[4313B]

The big question:
The N96 X-over has no cap for the LE10H... should I just leave it alone?It's cheaper.

Here's my charge coupled 3113B networks:

Biased 3113B networks (http://www.audioheritage.org/vbulletin/showthread.php?30553-Bionic-4313B-s&p=306983&viewfull=1#post306983)

[Aaron]

Is that a subtle hint that I should look into 3113B X-overs instead?
Will that X.O. play nice with an otherwise unmolested L96?
One thing I'm not ready to afford is buying 4313B mids and tweeters.

Assuming the 3113B isn't compatible with an L96:
Is there a way to C.C. the LE10 without demanding that somebody figure out a whole new X.O.?

Fantastic work on your X.O. by the way. =o]

[4313B]

Will that X.O. play nice with an otherwise unmolested L96?Yep. As you can see by the various network schematics I've posted of the stock systems. The L96, L112 and L150A used a less expensive version of the 4313B network (conjugates removed on the LE10/128 and capacitor value reduced on the LE5 from 16.5 uF to 13.5 uF).

[Aaron]

Welly well!
At the very least I'll take those crossovers off your hands and the LE5-9s just to see if they work too.

Thanks for everything, sir!