4343 crossover modifications

[B&KMan]

But, generally , JBL uses a lot of 100 volt caps in its crossover .

<>

Thanks Earl for tips

Other question in this regards,

My small yellow caps is 200 V but readable info on caps is 200 V =>

this is a V DC or VAC rating???

The 200 VDC is Eqivalent to half ac power no ?? so maybe 100 VAC is good read ??

Anybody is ok whit that ??



Of course the Signal is AC but many CIE is put DC RATED no ???.

[Earl K]

Hi Jean

My small yellow caps is 200 V but readable info on caps is 200 V =>

this is a V DC or VAC rating???

- It could be a DC rating. It's hard to know if it doesn't say .
- When I mention a rating I'm talking DC .
- If I was you, I'd buy Solens 250V caps. That's more than enough in a rating / plus I like the length to diameter ratio of these caps more than the higher voltage types .
- Remember 2 , 250V caps in series ( end to end ) = a 500 volt cap .

The 200 VDC is Eqivalent to half ac power no ?? so maybe 100 VAC is good read ??

- Yes, 200VDC does at least equal 100VAC maybe even 130 VAC .

<. Earl K

[Earl K]

Hi Jean ( again )

Well, this pict is a gold information... (I assume is relevant of the 3143 spec.)

After print in big scale, and check my all mesures and build hand curve by comparison of report data. the verdict is this pict.

(I adapt my scale x and y for good corrolation in this graph.)

- Nice work on overlaying your voltage drives ontop of Guidos , it certainly helps make useful comparisons .

The weird curve of 15 inch 2231 is
below at start ???
double slope is expression of bad caps no ???

- maybe not

What the f(...) with 2420. Bad soldering ??? bad caps ??? or

nightmare a war of clone !!!

- see below

It is possible to thrd part is expose this problem of maybe the other remplacement diaghram is not exact good response ???

- Well, aluminum diaphragms do wear out by getting hard ( usually around the suspension ) / but that usually limits the lower frequency response / you have just the opposite happening .

Here's my thought for today ;

- Guidos' pic of the voltage drive of the 4343 crossover uses a 8 ohm resistive load. ( I believe )

- To make a meaningful comparison to his pic / you should rerun your tests with 8 ohm load resistors in place of the drivers .

- Then you can compare the theoretical crossover ( Guido ) to its' practical implementation ( you ). Right now your drivers act as loads that are much more complex than Guidos' simulated voltage drive. Your drivers ( loads' ) have inductive and capacitive reactance that isn't accounted for with Guidos voltage drives which is only a simple resistive load .

So I will go at my start question what the life of CAPS
(according of the type) ???

- Since you live in or near Solens' hometown ( I think ) you can easily get Solen stuff relatively on the cheap ( sans shipping ) .

- To start, I'd just buy a bunch of the appropriate value caps ( no DC biasing for now ) / install them / and rerun your voltage tests . Thats the KISS method at work .

[Robh3606]

I would consider building the 4344/45 network. Both the 4340 and 4343 have a peak in the midrange drivers passband. The 4344/45 network does a good job of supressing this peak in the drivers response, in this case the 2122H. I would also not keep the 52uf in series with the High and UHF drivers in passive mode but wire them like the biamp set-up. In the 4344/4345 schematic they are all in parrallel. The drivers used in the 4343 and 4344 are for the most part identicle for the High and UHF drivers so underlying issues such as powerhandling should not be a problem. Attached is a 2 meter in room measurement on my RTA 1 second averaging. This is raw with no EQ.


http://www.lansingheritage.org/image...4340/page2.jpg

http://www.lansingheritage.org/image...4343/page3.jpg

http://www.lansingheritage.org/image...345/page03.jpg

http://www.jblproservice.com/pdf/Net...%20Network.pdf



Rob

[B&KMan]

Attached is a 2 meter in room measurement on my RTA 1 second averaging. This is raw with no EQ.
Rob

Thanks ver much for many hint but before analyse that lease let me know what is Y scale of the RTA jpg ??? what db value for each pixel ???




Jean.

[B&KMan]

Hi Jean ( again )
- Nice work on overlaying your voltage drives ontop of Guidos , it certainly helps make useful comparisons .

Well thanks for limitation and hint of this test.

In according , the driver is less than 8 Ohms and react on signal on other hand the random noise is not contain transient or hight peak so the rms is realy good...

Last night I try to understand the debate on guildo and giskars on 52 UF caps is serial or parralle on hot signal in regards of HF and UHF.

Actually my alanlyse Is demonstrate the 52 UF is parralle in regard of positif signal HF and UHF but the 4 UF caps is in serial for HF AND UHF !!!

It is a wrong build or what ???

I dismentale the other speakrs tonight for determine is wrong montage at specific or the schematic is little wrong...

But for now thanks very much (and thanks for your comprehension of my "wack" english


jean.

[B&KMan]

Attached is a 2 meter in room measurement on my RTA 1 second averaging. This is raw with no EQ.

Rob

Please let me 2 others questions...

1---

1 seconds is Half of the time SLOW in slm or standart ISO...

Well Incridebile beautyful line !!!

2---
What is the second pict on paper B&K ???

Thanks

Jean

[B&KMan]

- It could be a DC rating. It's hard to know if it doesn't say .
- When I mention a rating I'm talking DC .
- If I was you, I'd buy Solens 250V caps. That's more than enough in a rating / plus I like the length to diameter ratio of these caps more than the higher voltage types .
- Remember 2 , 250V caps in series ( end to end ) = a 500 volt cap .

- Yes, 200VDC does at least equal 100VAC maybe even 130 VAC .

<. Earl K

Well Let me ask a last question for now...

The time rate it is in comparison of the capacity volatage ???

More voltage more slow time transient response no ??? more energy keep by caps ???

Thanks

Jean

[Robh3606]

Hello Jean

That is a Behringer 8024 display. It is set for 1db per pixel in the Y scale.

The second picture is the raw 2122H response curve to show the rise in response the networks flatens out. I have ran the 2122H in an all active network and you need to notch out the response above 800Hz before your crossover point at 1.2K.

Rob

[B&KMan]

Hello Jean

That is a Behringer 8024 display. It is set for 1db per pixel in the Y scale.

Rob

So full actual scale is 80 dB Right ?? and all "line" horizontal is 10 dB ...

I observe you have a kill echoe option it is work good ??

VERY Impressive curve...

[Robh3606]

If I cross reference that level at 2 meters with my SPL meter the average is about 73dB so at 1 meter it's close to 80db. Full scale is another 30db above the average. It's a strange scale it runs 64dB full range and 8db for each of the 8 major divisions. If I was running .5 db per pixel it would be 4db per major division 32 full scale.

Rob

[Guido]

Here are some more informations about the 2121H

http://audioheritage.csdco.com/vbull...highlight=2121

[4313B]

I've been working on a simplified schematic for the 4344/4345 Rob. Would you be willing to try it out and measure it?

[Earl K]

Hi Jean

Well Let me ask a last question for now...

The time rate it is in comparison of the capacity volatage ???

More voltage more slow time transient response no ??? more energy keep by caps ???

I interpret your question to be asking the following ( correct me if this is wrong ;

Is there a direct correlation between the "Slew Rate" of a capacitor and its' voltage capacity ?
__________________________________or
_________________ ( which by extension might be rewritten as ):

____________ Does a thicker film effect a capacitors "Slew Rate" ?

- Simple answer / I don't know .

- I'll have to go read some specification sheets for industrial grade capacitors.

- I'll get back to you with what I find out.

[4313B]

I would consider building the 4344/45 network.

Yes. I've harped on this incessantly. Anyone who owns a 4340, 4341, or 4343 has the golden opportunity to update their system to a 4344. Recone the 2121's as 2122's and build the 4344 networks. If you're going to spend alot of money on network components you might as well put that money into the better version.