New Equivalent 3145 crossover Network

[Ian Mackenzie]

New Equivalent 3145 Crossover
I will have at last completed the new equivalent crossover boards tonight.

The intention will be to first run some JBL test jig measurements and then post some voltage drives and driver response graphs.

Below are images of the crossover boards and spinning of the wooden coil formers to exact dimensions on a Lathe. This was necessary to arrive at the precise inductance and dc resistance required for this design courtesy of "Giskard".

The resistors are a temporary item until I can arrange some Mills resisters which are not available locally. Perhaps a member can offer to help out here..no doubt they are superior to sand cast types.

The capacitors used in this instance are Claritycaps , film and foil 630 volt 5% tolerance. The boards are polypropolyne and were actually cut from domestic breadboards. The terminations are all 1 mm 100% pure Aust copper.

Ian

[Ian Mackenzie]

I measured all the filters using the JBL test Jig. The results where within 5 % or the tolerance of the components. The horn filter was perhaps a tad sharper with the -17 db point at 8500 (9000 spec).

I then measured the individual response of each driver at a nearfield distance of 1 foot to completely remove room effects and focus on the bandpassfilter actions.

The L pads were adjusted for 0 db and the analyser scale set for 0db. By adjusting the scale I was able to determine the average of the horn was within primarily a 4 db window (+-2 db). Not bad for a 30 year old horn design.

The 2405 was also measured for the L Pad at 0 db on axis as was the 2122H midrange driver. .

The overall full response was then measured in the listening room at about 6 feet on axis with the 2307/2308 horn lense assembly.

Given the room is completely un treated and I had equipment scattered everywhere it is remarkably flat and free from significant peaks and dips (with the exception of the comb effects around 8,000-10,000 hertz which as we know are unavoidable and in my view not audible).

If anyone is interested in the technical aspects of making these crossovers or would like some construction assistance please send me a pm.

Ian

[Ian Mackenzie]

These next two graphs are off axis at approximately 30 degrees taken at 0.5 metre and 1.5 metres off axis with the microphone on the vertical axis of the 2307/2308 horn lense assembly.

There is some tappering off of the horn but the response is still remarkably smooth in both cases. With some careful use of equalisation it would be hard to argue a case for a CD horn and even without any equalisation the results are excellent given many modern systems can't hold a candle to these results.The combined left and right channel off axis would be interesting but I do not have this facility at the present time.

Ian

[Ian Mackenzie]

For those interested the system is pictured below, the JBL Professional Series 4345. This is in fact a clone I build about 18 months ago.

Ian

[Ian Mackenzie]

Hi Earl,

Thats a fair question. My understanding from the data I have and from audible results is that they are. By virtue of their size and the construction I can't imagine why they would not be film foil. Metalised foil types are by design significantly smaller..The Auricaps are case in point but thye are also lower voltage. I am not proposing to destroy one to satisfy curiosity. I would point out however the generic description has nothing to do with is it good or bad..there are good and bad example of both. The Auricaps are regarded as some of the best capacitors available and I think they are ahead of the Claritycaps in some areas but they are more expensive, the Mundorfs Gold line is possibly better again but the are 4x more expensive again.

The Claritycaps have a clean, detailed and perhaps slightly warm sound.

You really have to take the plunge and try some out and this have been beaten to death countless times before on the forums.

The subject here is about this particular crossover and its application in the JBL 4345 and 4344 monitors.

Ian

[yggdrasil]

Very nice Ian.


The coils seem to have the formers in place even after mounting. Is this so? And if so - are you using the formulas for air coils?

[johnaec]

Hi Earl,

Thats a fair question.
(I'm assuming a message was deleted here, since I see nothing from Earl before this reply).

Please don't delete messages after they're posted - it just totally breaks the continuity. http://audioheritage.org/vbulletin/images/smilies/banghead.gif

If you must, *edit* them - at least a marker will be in place...

John

[4313B]

Thanks Ian.

Feel free to repost the schematic.

Let me know what Mills resistors you want and I can include them with the diaphragms.

[Chas]

Great and interesting work, Ian. I hope one day to do something along these lines, sans the 2245H.

Earl, this thread reminds me of your comment eons ago about Solens with JBL's not being to your taste. My original Giskard equivalent to the 3133 network used Solens initially. About a week or so ago I updated them with Hovlands. A very big improvement :yes: .

[4313B]

Biased or unbiased Solen's Chas?

[Ken Pachkowsky]

Ian

This invites the question, how would you compare the sound of this passive design to your active crossover?

I ask this simply because I was blown away by the performane of the BBSM15's I reluctantly sold in November. They were all passive with one Boulder 500AE driving them. Perhaps that they were raised 6 feet off the floor and slightly tilted forward made a difference ( I found myself thinking maybe getting mine further off the floor would do the same after all, they were designed to be soffit mounted). The BBSM15's were truly stunning.

Ken

[Chas]

Biased or unbiased Solen's Chas?

Unbiased, I was lazy.....:p

[4313B]

Just curious. I'm not a fan of unbiased Solens either. Just wanted to keep a frame of reference going. JBL used biased Solens in several statement systems and viewed them favorably.

[Ian Mackenzie]

Ken,

The system is biamped at 290 hertz and the 2122H, the 2307/2308 and the 2405 have a passive crossover. The biamping in the way that I am doing it is superior to full passive operation.

I think Earl deleted his remarks because he is skeptical or had doubts about the Claritycaps and it was just a shrt question answer thing.:hmm: . I don't think he has any interest otherwise in this remarkable crossover or the system.

One day I will make a charge coupled Solen Network.:)

Yggdrasil,

I kept the cores in there because I found removing then altered the inductance. It is also difficult to remove a solid core.

For next project I am am looking to build a universal coil winder just for grins. Making thing takes a little time but its fun and relatively inexpensive. The really fun part was chopping up the Bread Boards..nice colours.

The response curves perhaps woud be even smoother if I moved some items in the room.

Ian

[Ian Mackenzie]

Here the schmatic of the New Equivalent Giskard posted elsewhere on the forums .The original Equivalent Equivalent is shown for comparison.

Ian

"DCR values are extremely important and this model is based on Jantzen 18 and 20 AWG coils.

Use gangs of resistors to achieve proper power handling.

Orange is "old equivalent", green is "new equivalent", white is actual "tapped autotransformer" model.

Orange was done from 1981 Engineering Design Specification voltage drive using datapoints - all L-Pads open - all 8 ohm dummy loads.

There is enough slop in the L-Pads that both are viable - the new model should be cheaper to build with less parts."

[Ian Mackenzie]

Please post any questions you may have.

Non of this happened in 5 minutes and there was a lot of thought put into it particularly by Giskard. We should also respect the original work by the designer/ engineer, JBL's Greg Timbers.

So if you need a few tips or information not posted please ask.

Ian

Long Live the JBL 4345, Three cheers for Greg Timbers, Three cheers for Giskard.
:cheers: :cheers: :cheers:

[boputnam]

Great Thread, Ian. Kudos to the design team - all of 'em. I wish I could hear them.

The black baffle embarrasses me. :o: Let me see what I can do.

[Ian Mackenzie]

Thanks Bo,

Its an old picture!

Ian

[Earl K]

Hi All


I think Earl deleted his remarks because he is skeptical or had doubts about the Claritycaps and it was just a shrt question answer thing. . I don't think he has any interest otherwise in this remarkable crossover or the system.

(A) ( John ) Post Deletions ? Yes ;


- I deleted two posts because ( as noted above ) they were of a conversational manner, which I knew was off-topic ( to the central N3145 chat ).

- My intent to delete ( after asking & receiving a reply ), was stated clearly in the first line of each post.

- Since Ian decided to let his answer stand on record / I guess I was hasty in my deletions . Sorry if that had people looking under stones .

(B) (Ian ) N3145 ;

- I actually do have some interest in circuits similar to this ( incorporating 2-pole to 3-pole transitions ) / just not much interest in this one specifically ( as mapped out with these LC values ).

- Equivalent circuits ( to a tapped autotransformer ) are very interesting. I have to admit, I don't know how to create the "equivalent" pair of inductors ( using a calculator and the posted formulas ). Now, a bandpass circuit consisting of a; series coil followed by a shunt coil followed by a series cap followed by a shunt cap, is an extremely interesting adaptation of a circuit in s far as it "breaks the standard topology rules of passives' arrangement". I wonder how many at LHS members' have listened to this arrangement and it's audible deviations from more "normal" characteristics in circuit dampening .
.
- I prefer larger diaphragmed compression drivers without lenses in front of the horn. So that preference precludes directly implementing the N3145 for any of my projects .

- Last Nov. I spent quite a lot time persuing ( with calculator & empirical RTA testing ) the "best-fit" LC values that would enable the raising of the crossover point up into the 10K range ( for the horn-driver/horn-driver combos ) . I left this effort unfinished. It'll need to remain unfinished for the near future .

- I fully intend to include a tweeter within my final MTM efforts ( switchable into & out of the hipass circuit since there are a few occasions where I enjoy its' contribution ) .

- The tweeter driver will be a small format ( small diaphragmed ) compression driver, instead of the usual ring-radiator type. I've had quite good success , ( level matching ), with the inclusion of ( from Italy ) two different polyester diaphragmed drivers. I'm still looking for "THE" small / "efficient" ( @ 1.25" diaphragmed ) driver .

- At some point I'll want to also tackle minimizing the acoustic nulls in the crossover area caused by the acoustic overlaps.

- I suspect that a High "Q", series LCR trap ( paralleled across the load ) & set to the correct band-width will help achieve the necessary steep slopes out in the tweeters stop-band. This frequency dependant "trap" can be "hidden" from the amp by putting it behind ( driver side ) any reasonably valued pad. This position will protect the source amp from (seeing & ) having to deal with a frequency dependant short-circuit .


- The ( horn ? ) will be a simple diffraction slot ( a la the WestLake approach ) but smaller in depth due the octave higher Xover point . The "horizontal" orientation of its' lips can fit into the tight MTM component pack that I have to contend with . My horn "tests" were done with just the metal throat portion from the original 2344 horn . That throat alone provided enough loading for my purposes .

(C) ClarityCaps, FWIW; Tony G. is auditioning these caps now & will eventually describe his impressions in a 2006 update at Humble HomeMade HiFi . (http://www.humblehomemadehifi.com/)

- I intend to try them out since I don't own any large value Film & Foils and these seem affordable. Also, I'm terminably curious when it comes to the sound of caps.
- I was going to order some ( yesterday ) from Ians' Australian contact until I discovered they don't accept online, Visa or PayPal.
- Shortly, I'll pursue the money-order angle ( Western Union I guess ) .

:)

[Ken Pachkowsky]

Ken,

The system is biamped at 290 hertz and the 2122H, the 2307/2308 and the 2405 have a passive crossover. The biamping in the way that I am doing it is superior to full passive operation.

Ian

I can appreciate the active 290 hertz x-over from active to passive, but how would you compare the 290 hertz+ passive performance to that of the active x-over?

Just curious

Ken

[Ian Mackenzie]

Ken,

I don't know yet as I have no means of a full active sytem at this point.

I would say it would be difficult to mimic what has been done with the passive crossovers, that is why so often people get a different result when they attempt to plug and play with a active system..its is not that simple

Ian

[Ken Pachkowsky]

Ken,

they attempt to plug and play with a active system..its is not that simple

Ian

I would sure second that. As you are aware I have experimented with many different active systems and so far always end up back to the original. Perhaps that will change in a few months:) :blah: .

Ken

[4313B]

I would say it would be difficult to mimic what has been done with the passive crossovers, that is why so often people get a different result when they attempt to plug and play with a active system..its is not that simpleOften true. Once one gets the voltage drive of the active or line level passive solution identical to the speaker level passive solution one should be good to go. One can't really do that with off the shelf crossover cards in off the shelf production equipment.

[andresohc]

I am looking into getting the caps for this circuit and will be checking with E Speaker.com soon. Has anyone tried ordering Clarity Caps from them yet?

http://gallery.bcentral.com/GID4956567DD509579-CLARITYCAP.aspx

[Ian Mackenzie]

News:

http://www.audioheritage.org/vbulletin/showpost.php?p=129259&postcount=194