Giskard Charge-coupled 3145 crossover network iron core inductor source

[tjm001]

I was kicking around the idea of building one of these for a 4345 DIY project. But where in the world would one find 5.4 and 4.8 mH iron core inductors?

http://www.audioheritage.org/vbulletin/attachment.php?attachmentid=36905&stc=1&d=1233438483

[JeffW]

Parts Express (http://www.parts-express.com/cat/solid-core-inductor-crossover-coils/299?N=21725+4294967118&Ne=10166&Nrs=collection%28%29%2Frecord%5Bendeca%3Amatches%2 8.%2C%22P_PortalID%22%2C%221%22%29+and+endeca%3Ama tches%28.%2C%22P_Searchable%22%2C%221%22%29%5D&PortalID=1&showMoreIds=21806) has 5.5 and 4.7 iron cores, just the first place I looked.

You could bi-amp them and do away with the larger coil (is it 5.5 or 6.4? I can't tell for sure).

Edit: I blew it up and that bigger one is 6.4mH

[JuniorJBL]

You can also get bigger and unwind it.

[tjm001]

Parts Express (http://www.parts-express.com/cat/solid-core-inductor-crossover-coils/299?N=21725+4294967118&Ne=10166&Nrs=collection%28%29%2Frecord%5Bendeca%3Amatches%2 8.%2C%22P_PortalID%22%2C%221%22%29+and+endeca%3Ama tches%28.%2C%22P_Searchable%22%2C%221%22%29%5D&PortalID=1&showMoreIds=21806) has 5.5 and 4.7 iron cores, just the first place I looked.

You could bi-amp them and do away with the larger coil (is it 5.5 or 6.4? I can't tell for sure).

Edit: I blew it up and that bigger one is 6.4mH

I should have been more clear in my original post. Parts Express was the first place I looked at before I posted and like you said they have 5.5 and 4.7 mH. But the schematic (it's a little fuzzy) states 5.4 and 4.8 mH. Close, but no cigar. Just thought with the many, many posts about this particular crossover someone hopefully knows where to get them. Thanks.

Tom

[Baron030]

If you can't find a coil with the required inductance value then order a coil with a slightly higher inductance value and then remove a few windings from it. But, it will require you to have an inductance meter.
Fortunately, the cost of Digital LCR Multimeters has come down a lot.

A quick source on amazon turned up the following for under $40.00:
"MASTECH MS8269 Handheld Digital Multimeter LCR Meter Resistance Capacitance Inductance & Temperature Tester"

And if you question the accurately of such a cheap LCR meter than you can always calibrate it by taking a reading of a coil before you remove any windings from it.

And if you have no long term need of a LCR meter then you can always sell it after the 3145 networks are built.
I would think that you could find a buyer for slightly used LCR meter right here on this site.

Baron030:)

[JeffW]

I should have been more clear in my original post. Parts Express was the first place I looked at before I posted and like you said they have 5.5 and 4.7 mH. But the schematic (it's a little fuzzy) states 5.4 and 4.8 mH. Close, but no cigar. Just thought with the many, many posts about this particular crossover someone hopefully knows where to get them. Thanks.

Tom

I still think it says 6.4, but that's just my opinion. And there's just 2% difference between 5.5 and 5.4, likely much closer than the tolerance of the capacitors.

Edit:

OK, here's his post with the list, I stand corrected, 5.4 it is:

http://www.audioheritage.org/vbulletin/showthread.php?32139-stock-4345-crossover-3145-schematic&p=324116&viewfull=1#post324116

[Earl K]

Here's a repost ( from Heather ) of Giskard's original BOM ( for the 3145 ).

http://www.audioheritage.org/vbulletin/attachment.php?attachmentid=26032&stc=1&d=1182738665

:)

[JeffW]

Looks like she got 5.4, as well. That schematic sure has a 6 looking 5 in it.

[Odd]

Better readable here?

75515

[JeffW]

Better readable here?

75515

I swear that still looks like a 6 to me, but the 5.1 ohm resistors are clearly 5s. Anyway, I'm pretty sure I was mistaken and 5.4mH is the correct value. I won't hazard a guess on why the parts numbers in the BOM above are 5.6mH, I ain't building them, anyway (and I'd still bi-amp that inductor out of the system, to boot).

Not 5.4mH, not iron core, not sure.

L125.6 http://solen.ca/products/inductors/litz-inductors-12awg-1-93mm/l125-6/

S125.6 http://solen.ca/products/inductors/standard-inductors-12awg-2-05mm/s125-6/

S145.6 http://solen.ca/products/inductors/standard-inductors-14awg-1-63mm/s145-6/ Has the same .63 DCR as the L125.6

[Earl K]

The first line of parts numbers ( in Giskards BOM ) are from Solen, see ; Solen S145-6 (http://solen.ca/products/inductors/standard-inductors-14awg-1-63mm/s145-6/).

That first L125-6 entry equates to Litz Wire, 12Ga, 5.6mH ( .63 DCR ) .

Both coil types would need to be unwound to get the correct 5.4mH value .

:)

PS; now you got it Jeff!

[tjm001]

I ain't building them, anyway (and I'd still bi-amp that inductor out of the system, to boot).



Thanks. That's a good point. I do plan to bi-amp. For some reason I was not thinking straight when I was glancing at the schematic. I was just thinking the top section was HF, the middle was MF and the bottom section was LF. :banghead: However, I sure wish I could hear from someone that actually built one of these as to the verified values, sources, etc. and any helpful tips and tricks. Thanks.

Tom

[tjm001]

OK I've been studying the old posts http://www.audioheritage.org/vbulletin/showthread.php?6633-Discussion-Thread-JBL-4343-to-4344-upgrade/page14 and http://www.audioheritage.org/vbulletin/showthread.php?1388-4343-crossover-modifications for several days now. These are excruciating detailed and drawn out communications over a long period of time back in 2005 between mainly a French guy in Canada (BandKMan) and others mostly (Ian Mackenzie) and others about crossovers for 434x systems with 4313B. I am enlightened and much more informed after reading these strings. I also think I now better understand 4313B's reaction to my post last year questioning Charge Coupled Crossover Networks. Sorry I stumbled into this about 10 or more years after the fact! But I am not a troll. I was and still am an innocent and honest bystander at the wrong place at the wrong time when I questioned Charge-coupled crossovers.

It seems obvious to me from reading all this is that the way to go on a 4345 DIY crossover is the bi-amp option. My problem is I'm from the old school of what electronics were back in the 1960's. I studied and was well versed with all the Capacitor/Inductor stuff back then, but my career path led me in an entirely different direction (Air Traffic Control and Law Enforcement). Other than the Nelson Pass crossovers I built for my L300s last year, I haven't fooled with this stuff since the late 1960s. I'm still into the Amp, pre-amp and passive crossover mode. So I'm a little shaky on the bi-amping.

I am confident I can build Cross-coupled crossovers for my 4345 project with minimal advice. My question at this point is, if I go bi-amp or even tri-amp as Greg Timbers recommends for the 2405, do I just follow the schematic eliminating the LF and UHF legs of the schematic? My instinct tells me there is more to it.
http://www.audioheritage.org/vbulletin/attachment.php?attachmentid=36905&stc=1&d=1233438483

[Earl K]

Hi ,

If you bi-amp; eliminate all the passive components within the section bounded by the dashes .

If you tri-amp ( as per GT's advice for adding an amp onto the UHF ) then do as above but also eliminate all the passive components in the UHF leg ( personally, I'd still include the variable Lpad for ease of component balance ) .

:)

[tjm001]

Hi ,

If you bi-amp; eliminate all the passive components within the section bounded by the dashes .

If you tri-amp ( as per GT's advice for adding an amp onto the UHF ) then do as above but also eliminate all the passive components in the UHF leg ( personally, I'd still include the variable Lpad for ease of component balance ) .

:)

Thanks. This helps. I'm beginning to get the picture. I need to bone up on the equipment that goes before the amps now. Any suggestions?

Tom

[Earl K]

Thanks. This helps. I'm beginning to get the picture. I need to bone up on the equipment that goes before the amps now. Any suggestions?

Tom

My suggestion is to read all the older threads ( spread throughout this great site ) about the 4345 ( & bookmark them all / if you haven't already ).
- They do offer a very good road-map in moving forward ( quality > listening wise, IMO ).

The standard crossovers ( mentioned in those threads ) are all quite acceptable electrically ( from this guys perspective ).
- The analogue types mentioned, were models from Bryston & Ashly ( & to a much lesser extent , the ancient JBL 5235 variants ).

:)

PS; If you decide to put a separate amp on the UHF, I believe GT was talking about a tiny classD GainClone of sorts ( preceded by its own dedicated HiPass ). This could easily be something from miniDSP ( out of China ).

[BMWCCA]

PS; If you decide to put a separate amp on the UHF, I believe GT was talking about a tiny classD GainClone of sorts ( preceded by its own dedicated HiPass ). This could easily be something from miniDSP ( out of China ).I'd considered doing this on the 4345, some day. I don't find any inadequacies in the current setup using the bi-amp version of the crossover design here, but figured it's a simple mod easily undone, after satisfying my curiosity. I've set aside a couple of Crown D45s for this purpose, figuring to also bi-amp the 030 system one-day—just for fun—with a pair of Crown VFX-2A crossover I've had sitting around for years. Just never enough time for all the plans. :dont-know:

[mech986]

Falcon Acoustics out of England also produces iron core, sintered ferrite core, and air core inductors for speaker crossovers. The favorable exchange rate is nice as is the custom specification of value, tolerance, and core type. Falcon have served the British speaker industry for 4 decades.

http://www.falconacoustics.co.uk/audio-inductors-ferrite-air-core-iron-dust.html?cat=60

http://www.falconacoustics.co.uk/audio-inductors-ferrite-air-core-iron-dust.html?cat=61

[Ian Mackenzie]

OK I've been studying the old posts http://www.audioheritage.org/vbulletin/showthread.php?6633-Discussion-Thread-JBL-4343-to-4344-upgrade/page14 and http://www.audioheritage.org/vbulletin/showthread.php?1388-4343-crossover-modifications for several days now. These are excruciating detailed and drawn out communications over a long period of time back in 2005 between mainly a French guy in Canada (BandKMan) and others mostly (Ian Mackenzie) and others about crossovers for 434x systems with 4313B. I am enlightened and much more informed after reading these strings. I also think I now better understand 4313B's reaction to my post last year questioning Charge Coupled Crossover Networks. Sorry I stumbled into this about 10 or more years after the fact! But I am not a troll. I was and still am an innocent and honest bystander at the wrong place at the wrong time when I questioned Charge-coupled crossovers.

It seems obvious to me from reading all this is that the way to go on a 4345 DIY crossover is the bi-amp option. My problem is I'm from the old school of what electronics were back in the 1960's. I studied and was well versed with all the Capacitor/Inductor stuff back then, but my career path led me in an entirely different direction (Air Traffic Control and Law Enforcement). Other than the Nelson Pass crossovers I built for my L300s last year, I haven't fooled with this stuff since the late 1960s. I'm still into the Amp, pre-amp and passive crossover mode. So I'm a little shaky on the bi-amping.

I am confident I can build Cross-coupled crossovers for my 4345 project with minimal advice. My question at this point is, if I go bi-amp or even tri-amp as Greg Timbers recommends for the 2405, do I just follow the schematic eliminating the LF and UHF legs of the schematic? My instinct tells me there is more to it.
http://www.audioheritage.org/vbulletin/attachment.php?attachmentid=36905&stc=1&d=1233438483

It's a learning curve and early on we were still discovering historically how these filters worked.

This was the equivalent 3145 schematic charge coupled

The differences are in the mid and uhf filters compared to the stock schematic and the original filter that GT designed.

The voltage drives in all cases are almost identical.

The most recent charge coupled filter used a diode to create the screen voltage from the signal instead of the battery.

Try and use the closest dcr values for the coils.

The filters should be separated by "air" and not in close proximity to each other or any ferris metal.

As mentioned just do biamp and feed the mid filter without the series capacitor and shunt inductor.

Be careful wiring up the crossover to observe the polarity of the drivers.

The factory crossover had a number of compromises and eliminating the biamp switch improved it a lot.

The adjustment of the Lpads requires care and l have covered this in a post elsewhere

These systems has a live character and I f you can get your hands on some of the acquaplas, dusted diaphragms are an improvement.

Links to most of these points can be found in the 4345 reference thread.

[tjm001]

It's a learning curve and early on we were still discovering historically how these filters worked.

This was the equivalent 3145 schematic charge coupled

The differences are in the mid and uhf filters compared to the stock schematic and the original filter that GT designed.

The voltage drives in all cases are almost identical.

The most recent charge coupled filter used a diode to create the screen voltage from the signal instead of the battery.

Try and use the closest dcr values for the coils.

The filters should be separated by "air" and not in close proximity to each other or any ferris metal.

As mentioned just do biamp and feed the mid filter without the series capacitor and shunt inductor.

Be careful wiring up the crossover to observe the polarity of the drivers.

The factory crossover had a number of compromises and eliminating the biamp switch improved it a lot.

The adjustment of the Lpads requires care and l have covered this in a post elsewhere

These systems has a live character and I f you can get your hands on some of the acquaplas, dusted diaphragms are an improvement.

Links to most of these points can be found in the 4345 reference thread.

Thanks! I've decided to do the triamp thing as GT described with the little amp for the 2405. So the crossover construction should be fairly straightforward since there will only be three air core inductors.

Tom

[fencki]

hi.
as i am in the same state with my JBL 4345 i have found another (different) schematic for the 4345 here in the threads.
it was mentioned it is an final schematic with slightly different values.

would anybody recommend this schematic over the "old" one?

http://www.audioheritage.org/vbulletin/showthread.php?10613-4345-Includes-Designer-s-Post
post #4 (PDF)

thanky a lot for your input here.

regards from vienna
slavko

[Ian Mackenzie]

The schematic on the right in the PDF will work fine.

[fencki]

can you please just explain to an non-pro, what difference the two schematics (differnence in values of the C´s, L´s and R´s) have?

[Ian Mackenzie]

No Lpads and standard component values


The end result is the same

[fencki]

No Lpads and standard component values


The end result is the same

R44+45+46 is the subsitute schematic for the L-pad. that´s how i understand it. (i will put L-pads instead of R44+45+46)
(also R35+36+37 and R29+30+31)

but i would like to know why the values changed slightly over time?

i am taking clarity sa caps instead of solen caps, because they are not so common and available here in austria.

thx

regards

[Ian Mackenzie]

Not wanting to damp your curiosity

You wanted a non pro response

The values are based on the load values in each specific schematic

An Lpad is not an ideal constant impedance attenuator, that is why the fixed network is used

Do not change anything

The voltage drives are set up specifically for those specific parts

If you want to use Lpads then use the other equivalent 3145 schematic

[quindecima]

Ian,

Do you have a link to the schematic with the diode instead of the battery? I looked for one but couldn't find it.

[Ian Mackenzie]

Thanks! I've decided to do the triamp thing as GT described with the little amp for the 2405. So the crossover construction should be fairly straightforward since there will only be three air core inductors.

Tom


Any progress on this yet?

[tjm001]

Any progress on this yet?

Thanks. Still trying to gather all the components. I just built the L300s last year so I'm still enjoying them. Not sure if I should leave them intact and sell them, or pirate the LE85s and 2405s to use in the 4345 project. Probably makes sense to leave the L300s intact and get different mid and high frequency drivers. I regret selling those short horns last year on eBay. :banghead: I'll probably start building the boxes after I get back from my Florida trip next month. I'm in no rush. I will start a new thread about the construction when it gets underway.

Tom

[Hoerninger]

Ian,

Do you have a link to the schematic with the diode instead of the battery? I looked for one but couldn't find it.
Look here :cool:
http://www.audioheritage.org/vbulletin/attachment.php?attachmentid=64799&d=1425751189

Enjoy :)
____________
Peter

[hpham]

Hello,
After few month I had collected all drivers for 4 way: 2235H - 2121H - LE85+H91(Blue) - 077. May I follow below schematic or need to modify something for 2121H and LE85. I think 077 can take 2405 cross of 4344 is fine. I will built cabinet 4344 because we have very detail drawing in this forum.

77274

[Ian Mackenzie]

I recommend that you bi amp your project

This will significantly reduce complexity and cost of the passive crossover and improve the dynamic performance of the loudspeaker

Concerning the schematic that particular version of the circuit is the 4344/4345.

This version has component values and voltage drives optimised for the 2122H mid cone.

I will investigate what revision of the original 3143 factory was applicable for the 2121 midrange driver.

The 2122H driver is now NLA and people are either making a driver using a 2123 or E110 basket with an after market recone kit.

[Ian Mackenzie]

Okay,

I spent a while looking at some old threads on the subject of the 4343 3143 network.

At the time (2003-2005) most people were interested in restoring old 4343 or upgrading to the 4344 model.

The 2121 driver used in the 4343 is known for some break up modes and at the time it was felt that most people would benefit from a new equivalent crossover to incorporate the improved 2122 driver as used in the 4344-4345.

The crossover schematic you posted is the charge coupled network for the 2122 as used on the 4344 and 4345.

An updated schematic for the 4343 was not investigated and those who decided to keep the 2121 driver simply updated or re built the 3143 network or they used the 4340 schematic which is bi amp only.

Suggest you check the condition of the 2121 foam surrounds. There is nothing wrong with the 2121 driver, the 2122 driver is a refinement.

The 4341 was the passive only version.

So you have some decisions:

Use the 2121 driver Y/N
If "Y" build a 3143 or 3140 network (can be charge coupled)

DO NOT attempt to cut and past parts of the original 3143 network with the updated 2122 network as you will find yourself in a huge mess based on recent feedback.. The voltage drives are not compatible.

If "N" investigate a 2122 driver but very hard to find or the 2123 (but requires a specific network)

See this post
http://www.audioheritage.org/vbulletin/showthread.php?1388-4343-crossover-modifications&p=66931&viewfull=1#post66931

[hpham]

Thank Ian,
here is my 2121H foam surrounds condition :crying: it not so nice but I will keep and re-foam therefore I chosen "Y" and go for 3143 cross as you advised. I don't seen charge coupler for 3143 yet and I will try to find.

77295

I'm wonder why 3143 with 52uF filter but 3140 wasn't. The inductor 2.9mH are necessary or not because it also not appearance on 3140.

77296

[Ian Mackenzie]

Thank Ian,
here is my 2121H foam surrounds condition :crying: it not so nice but I will keep and re-foam therefore I chosen "Y" and go for 3143 cross as you advised. I don't seen charge coupler for 3143 yet and I will try to find.

77295

I'm wonder why 3143 with 52uF filter but 3140 wasn't. The inductor 2.9mH are necessary or not because it also not appearance on 3140.

77296

All your questions are highly relevant and frequently whenever the 3143 network enquiry comes up.

Firstly, visit this thread on re foaming if your intend to attempt it yourself.

Otherwise I recommend a qualified technician to avoid damage to the cone and voice coil.

There is some debate on the authenticity of after market re cone kits in terms of obtaining the drivers original specifications.

Look carefully before you purchase your re foam OR re-cone kit.

It may be possible to buy the JBL 2121 recone kit.
Pm member (Edgewound to confirm this)

http://www.audioheritage.org/vbulletin/showthread.php?469-Resurround-Step-by-step

Secondly, There is no charge coupled network so I will post a schematic given you plan to use the 2121.

Can I confirm you intend to run your project full passive OR Biamp?


In relation to the 52 uf capacitor this question also comes up with the 3143 network.

For clarity the 3140 network was the bi-amp only version for the 4340 monitor.

The 3143 network was designed for switchable operation from passive to bi amp mode in the stock factory.

It was later found that this switch was at the request of the marketing department and is a departure from the most direct and pure signal path.

Because the biamp high pass filter for the 2121 is done actively in the then 5234A active crossover the 52 uf capacitor and then 2.9 mh inductor is not required.

To improve protection of the HF and UHF drivers from amplifier turn on thump the signal is taken after the 52 uf capacitor in passive mode. (It took a while to understand why JBL did that)

[Ian Mackenzie]

Simplified 3143 charge coupled schematic

I can provide you more detailed information on how to build this.

Send me a pm

[Ian Mackenzie]

Simplified 3143 charge coupled schematic (simulation Linearx LEAP Crossover Shop)

I can provide you more detailed information on how to build this.

This should be rather good in charge coupled configuration

I might even build this myself

Send me a pm

As a side note while I was looking through some old threads a member wrote he found some new old stock grey 4343s and set them up at home.

Apparently they absolutely killed his Klipschorns.

But they were stock 4343s so if he still around he might be interested in this.

Ian

[Doctor_Electron]

An often overlooked but critical factor in crossover network design & implementation is the DCR of the inductors, which directly determines th " Q ", and therefore the steepness of the slope of the pole in which it is used.
A crossover section, which is an electrical filter, is made up of a number of poles connected to each other in order to approximate the desired steepness of slope and passband characteristics.

So if your intent is to copy the performance of the original network based on its electrical design and its interaction with the connected drivers, you must know the original inductors' DCR and incorporate this resistance value into the clone inductors' total series DC resistance.

For example, I had some 3107 networks, that each had two 1.8 mH inductors which in conjunction with the capacitors used made up a 12 db/octave network which controlled the response of the 4350 Studio Montor's 2202 midbass and 2440 midrange drivers.

The 1.8 mH inductors each had a DCR of 3.5 ohms, far higher than you would expect in a typical OEM coil which tend to be of low resistance.

A high quality non-inductive resistor of the proper wattage can be added in series with an inductor whose DCR is less than that of the original unit, to achieve the needed total series resistance.

Some of the less mathematically challenged than I members here might elaborate on this basic premise.

I hope this helps in you project. Regards, - D _ E -

[Ian Mackenzie]

You are correct

That is why JBL published voltage drives into the standard test jig are useful and this has been the key in all the equivalent schematics to date.

What is interesting is the stock network is not unlike my own diy network l did back in 1979. That system has been legendary.

But there's always room for improvement so l plan to put in a charge coupled network

The John Eargle book Handbook of Recording Engineering has much of JBLs thinking on crossover network of that era and that was the basis of my original network


The posted simplified schematic is a primer.