4344 MKII MF Caps

[jim henderson]

The 4344 MKII network schematic calls out 72uF caps on the MF (2123H) section. What kind of caps are these and where can I find them?

[Earl K]

Hi Jim

I would experiment with Charge-Coupling large value 250 volt, MPP Solens in mutiples in that midbass area. My "first' impulse might be to purchase 68uf mpp caps from Solen and then look to "tame/fix" the Solens with other high-grade ( see,,, expensive) caps. I personally haven't done this - plus I don't know if Solens have a sonic signature that actually "needs fixing" when used on a relatively slow cone of the 10" speaker. I'd check that out first with a 3.9 uf Solen - paralleled to the main 68 uf cap - you can always "bypass these two with another .1 uf high-grade audiophile cap. Charge-Coupling will take care of any "crossover" distortion concerns because of the class "A" bias topology .

Are you biamping or going the full passive charge-coupled route ?

regards ,> Earl K:)

[jim henderson]

I am biamping, so the MF network can forgo the high-pass section.

[4313B]

I'm not too sure what the type of dieletric is in that cap but I do believe it is still available from Pro for ~$70 each.

The complete high pass and low pass sections of the network are still available and in stock.

[Ian Mackenzie]

Jim,

If this helps, I have had very favourable results with the AEON range of metalised polypropolyene capacitors in the midrange high pass section of my 4345 network, bypassed with a AEON fim foil 0.1 uf.

Ian:smthsail:

[4313B]

You mean AXON Capacitors?

Here's something that may be interesting -

Appendix E - Capacitors (http://ldsg.snippets.org/appdx-ec.php3#CAPRECS)

[4313B]

Hey Jim,

It looks like Michael Percy Audio (http://www.percyaudio.com/) has some 70 uF AudioCaps (PPMF) for $46.50 each, considerably cheaper than the $67.02 each for the stock JBL caps.

[jim henderson]

Thanks for the AudioCaps PPMF suggestion. I'm also considering the following two options that are about half the price of the AudioCaps PPMF:

Option A: Dayton PPM 40 uF + 30 uF +2 uF, bypassed with an AudioCap PPT 0.01 uF

Option B: Axon PPM 68 uF + 4 uF, bypassed with an AudioCap PPT 0.01 uF

Opinions?

[Earl K]

Hi Jim

It's a bit of a struggle to offer any meaningfull advice for your situation since I feel you could be approaching this question from the wrong direction.

All capacitors do have sonic signatures that are quite audible when used inline ( in series) on himid devices - especially compression drivers.

If I were you, I'd develop an opinion about which capacitors you like the sound of by buying $100.00 worth of different values and listening to them on compression drivers. This is a very direct route to a realistic opinion. More so than than spending $100.00 on 4 capacitors that are hidden in the second pole / 2nd element position of a low pass circuit.

I've never seen anyone express an opinion of how much that second pole position influences the overall sound quality of the whole filter but I'd offer my opinion that it's only about 20% . Based on that, I'd actually be asking questions about the sonic qualities of various inductors placed before the low mid transducer.

For about $93.00, you can buy from North Creek Music (http://www.northcreekmusic.com/Zen.html) (4) 1.0uf "Harmony"caps, (4) 2.0uf "Zen" caps, (4) 3.0uf , (4) 6.0uf "Zen" caps, & (4) 10.0uf "Zen" caps . These caps will give you a nice kit to work from for building up different crossovers & compensation networks for JBLs' 1inch compression drivers. The S3100 & S3100 mkII networks make a lot of use of 3uf , 6uf & 12uf values or multiples of them. I'd recommend building the basic HF portion of the S3100 ( not charge-coupled ) first if you are using some sort of CD horn - leave out the LCR portions of the HF network . You'll need to buy a 2.4 mh inductor for that network. Building this will give you a real good idea of just what these capacitors sound like and what they offer in the way of timbral balance. Charge-coupling your effort will then indicate just how "clear" the whole thing can sound.

I would also buy 4 lowly 70uf NP electrolytics ( these are cheap - about $2.00 a piece ) and experiment with charge-coupling them . I'd also place these in front of your compresson drivers as DC blockers to form an opinion about just how do they sound. Then I'd bypass them with those (4) 1.0uf Harmony caps that claim to fix everything - including giving us world peace - and see just what it is they can do . There should be some surprises in this exercise. After you have chosen an inductor for your midbass area, I'd move these charge-coupled bypassed electrolytics into the second pole position and take a listen . Eventually, if I was convinced that Zen caps were my cup of tea , I'd buy (4) 75uf from Northcreek ( 24.95 at this time ) - charge-couple them , maybe bypass them, replace the electrolytics and call it a day . (75uf vs 72uf is a small difference in a charge-coupled network - divide the value by 4 to see what I mean ). If your are still curious about AuriCaps / AudioCaps / Hovlands / etc. ( and what might have been )- well then you can buy some to bypass ( fix ) the modest investment you made in the Zens.

regards ,> Earl K :) ( & thanks to Ian Mac for the North Creek link )

[4313B]

"I've never seen anyone express an opinion of how much that second pole position influences the overall sound quality of the whole filter but I'd offer my opinion that it's only about 20% ."

G.T. has stated on a few occasions that he feels it is 50%. After many years of messing around with this crap I would have to agree with him. His suggestion is to use the best possible poly caps you can find.

[Earl K]

Hi Giskard

G.T. has stated on a few occasions that he feels it is 50%. After many years of messing around with this crap I would have to agree with him.

Thanks for articulating that . It's a valuable insight .

regards <. Earl K :)

[Tom Loizeaux]

Your post called "Opinions on ( your $100 budget ) ?" was great!
Maybe I'm just lazy, but if you, or someone, would try your capacitor test and report the findings, it would be of great intrest and help to many on this forum. Yes, it is subjective, but we are all working toward the finer points of music reproduction and I would value opinions from most of the members here.

Tom

[Robh3606]

Hello Tom

For what it's worth what I did was take JBL's standard network mix Mylar, bypass them with polypropylene and then Charge Couple them. The standard mix is really not all that expensive and more importantly the end result is a Charge Coupled network which is you goal. I got my Mylars from Madisound. You could always parralel up to get your values. I don't know how these caps are quality wise as far as Mylars go but they are very reasonable. That said using them Charge Coupled they make a very nice sounding end result. Could the expensive polpropylene sound better?? Yeah maybe but I am sticking with these for now.


Rob:)

[4313B]

If you have a decent 2-way system or a minimal 3-way system such as a 4312C you can test different dielectrics and charge-coupled capacitors pretty easily. I think the caps in a 4312C are 3.0 uF and 10.0 uF so it isn't too big of a hit in the pocketbook to try several brands of caps.

As Rob has stated, the stock JBL mix since the introduction of the L110A has been mylar bypassed with polypropylene and that is definitely a great place to start. You might want to take that combination one step further by adding in an additional polystyrene bypass capacitor like JBL did in such systems as the L250, 250Ti, and XPL200A. First charge couple the base mylars and note the results. Then add in the bypass caps and determine if their use is warranted.

"Could the expensive polpropylene sound better?? Yeah maybe but I am sticking with these for now."
It all depends on how far someone wants to go. Some guys have to have $1,000 interconnects. Some guys do just fine with the interconnects packaged with their components. Early on when this topic first started up several forums ago, some guys were advocating wholesale network replacement using the absolute best parts one could find and those solutions could cost thousands of dollars in a complex system such as a 250Ti or 4345.

[4313B]

I guess my only concern with your suggestion Earl would be that listening to just a CD/Horn, especially one that requires EQ and leaving that EQ out, doesn't yield a fully functional balanced system for test purposes. I personally find listening to just a CD/Horn obnoxious in and of itself.

[Earl K]

Hi Giskard


I guess my only concern with your suggestion Earl would be that listening to just a CD/Horn, especially one that requires EQ and leaving that EQ out, doesn't yield a fully functional balanced system for test purposes. I personally find listening to just a CD/Horn obnoxious in and of itself.


Yes, that's a very reasonable concern. Uncompensated CD horns are even worse sounding than listening to straight exponentials . :p :rotfl: :duck:

I mentioned the S3100/S3100mkII because I'm under the impression that the CD horn compensation is built into both those circuits. Yes/No ??? I don't have the modelling software so I have to actually build these things to see what's happening - but - that CD compensation seems to be there with the 3uf cap followed by the 2.4mh coil(shunt) followed by the 6uf cap that is part of a combo RC/pad network. This arrangement is of course reversed for the mkII that includes more padding selectons and two much smaller shunt coils ( .2mh & a .6mh after the LCR anti-resonance traps ).

Both of these circuits have piqued my interest and as a result I'll soon be following my own advice and buying a bunch of North Creek caps / Solen coils, so I can build and then audition portions of the HF circuits . I'm particulary interested in the older S3100. Reports on these speakers consistently rates this driver/horn-circuit combo very favourably. That says a lot considering the use of the commonplace 2426 titanium diaphragmed driver.

regards ,> Earl K :)

[4313B]

Hi Earl,

I just looked at the original S3100 circuit and the 1.0 mH coil across the CD has no DCR value listed. That DCR value has to be known in order to build the network correctly. If I use 7.5 ohms as the DCR the voltage drive looks "ok". Leaving the DCR out or making it very small results in an utterly disasterous notch around 2.5 kHz. It would be highly beneficial to get someone with S3100's to measure the DCR values of the coils. :)

[Earl K]

Hi Giskard

I agree that the DCR of that 1mh coil needs to be known to successfully buildup the whole circuit. As you point out, varying the R gives different "notch" amounts. I didn't mention it ( I should have ) - but I view that portion of the circuit as a LCR segment for notch filtering. I did mention not to include LCRs in the buildup.

FWIW; The published response curve for the K2, M9500 seems to show deliberate "notches" put into the HF response. Their location makes sense for my listening preferences .

regards ,> Earl K :)

[Earl K]

Hi RobH

what I did was take JBL's standard network mix Mylar, bypass them with polypropylene and then Charge Couple them.

Hmmm, I was under the impression you bought surplus GE polypropylenes ( large value ) for your charge-coupling project. I thought your base capacitor was now a GE MPP type instead of the stock Mylar. So it was it GE Mylar ( or something ) that you bought ? I'm :confused:

regards <> Earl K :D

[Robh3606]

Hello Earl

It was Carli Mylars 4.7uF a GE surplus .033 Polypropylene and a Parts Express .001 Audio Grade Polypropylene. So the GE's are in there as part of the "Bypass" to get me to my 5uf nominal. Want to swap out the .001 for a polystrene/foil when I get up off my butt and order some. That will stop me on the cap quest for a while.

Regards Rob:)

[Earl K]

Hi Rob

Thanks for the clarification.. Nice to know that the Charge-Coupling also did the trick with mylars .

regards <> Earl K :cheers:

[4313B]

"Nice to know that the Charge-Coupling also did the trick with mylars."

Biasing works with any kind of capacitor. :)

[Earl K]

Hi Giskard

Do you know whether or not JBL uses biased/bipassed electrolytics in any of their K2 type speakers ?

( I was hoping this question might have answered itself by now with a picture of that S9500 crossover network from Niklas in Sweden ). :wave:

Oh well ! :eek:

<. Earl K :)

[4313B]

I couldn't imagine them NOT using electrolytics but I went in search of a picture anyway and came up with this -
INTRODUCING PROJECT K2 S9800 FROM JBL (http://www.jbl.com/home/k2_story/k2_intro.asp)

[Earl K]

Nice Informative Picture !

Thanks !

<. Earl K:)

[jim henderson]

Whoa, that's a charge-coupled, three-way, FOURTH-ORDER crossover. I can't wait to see that schematic.

[JohnH]

Where is the other side of the 9V battery connected?

[jim henderson]

Black - Negative - Ground.