Results 1 to 11 of 11

Thread: Ducatista47's crossover upgrade project

  1. #1
    Senior Member
    Join Date
    Apr 2003
    Location
    Australia
    Posts
    7,942

    Ducatista47's crossover upgrade project

    Upgrading the stock 3145 network in the JBL 4345

    A photo essay.

    Owner Ducatista47, Project support Ian Mackenzie
    Date: Sept 2006 - Nov 2006

    Publishing this project has been held over for a variety of reasons but it is now appropriate to post what happened. That is before we get too old to remember and our hard drives and pc's become obsolete!

    The decision to undertake this project was initially discussed in August/September 2006.

    In case you are wondering why upgrade a stock 3145 network?

    I wanted to measure, study and audition the stock network and hence what other JBL 4345 owners experience.

    At the time Clark was very interested in upgrading his newly acquired 4345 monitors to match the purity of the rest of his contemporary audio system. It was a matter of timing and this is how the opportunity for the project came about.

    The plan is to meet up with Clark in September this year to see how it all worked out.

    Enjoy

    Images 1-5 below illustrate the stock network in the JBL 4345 system and removal prior to shipping to Australia
    Attached Images Attached Images      

  2. #2
    Senior Member
    Join Date
    Apr 2003
    Location
    Australia
    Posts
    7,942
    The initial phase become a real exploratory journey as I was able to look far deeper into the way these networks were engineered and this will be photo illustrated with some discussions later in the following posts .

    Within this phase I also conducted some subjective evaluations of the stock network and compared them to the equivalent crossover designs.

    By gaining an appreciation for the way the networks were configured and the technical / engineering constraints of the stock design I was able to envisage an upgrade path on several levels. This including termination’s, switching and signal routing and component selection and voicing.
    Attached Images Attached Images      

  3. #3
    Senior Member
    Join Date
    Apr 2003
    Location
    Australia
    Posts
    7,942
    Following the exploratory phase of the project we determined it was viable up upgrade the networks and make significant improvements to the subjective performance of this particular system.

    Accomplishing this step was the most challenging aspect of the project.

    A key consideration was to find a practical means for implementation of the improvements with an approach that would be least invasive to the original network. The sheer complexity of the wiring and switching and the requirement for fitting in new parts in a pre existing board layout was a real challenge.

    All the exisiting wiring loom has been left intact. The terminations of the unused looms have been careful taped up. It was discovered that significant resistive losses are experienced in the operation of biamp switch. The action of the switch effects both the active signal and ground so it had to be completely bypassed (heavy white wires)

    Just seemingly little things like finding a suitable binding post and carefully fitting it was no easy task.

    Removal of the old capacitors was actually quite easy. The old hot melt glue came up quite easily with some gentle easing with a putty knife. Application of the new hot melt glue was also easy and everying fitted fine with a small bridge made from perspex for one of the capacitors on the HF pcb.

    Images below: Voltage drives of stock network, shematic, parts installation.
    Attached Images Attached Images      

  4. #4
    Senior Member
    Join Date
    Apr 2003
    Location
    Australia
    Posts
    7,942
    As part of our Capacitor selection decision process there was extensive consultation with Jeff at Sonic Craft to create a reference baseline network. Jeff is regarded as one of the foremost in the field of loudspeaker crossover capacitor upgrades.

    The reference network used in the evaluations was fitted a Mundorf Supreme Silver and Oil capacitors. Jeff regarded this the only worthwhile upgrade from the Clarity SA Capacitors (non bypassed) and previously used iterations of Mylar, Mylar bypassed (+biased), Hovlands, Auricaps and Solen Fast caps (+biased) for the 4345 system.

    During the development phase various capacitors (as mentioned above) were swapped out in situ within the stock 3145 network using a jig constructed high quality flying leads to verify the authenticity of the results.

    We also compared the stock network and an original equivalent network using identical capactors. I was not able to verify any subjective difference between them other than the stock network was judged inferior in performance without the capacitor upgrade. There are some variations on the shape of the voltage drives depending on the position of the L pads.


    Images: Final capacitor selection, flying lead jig for reference tests, voltage drives of networks under test, JBL DX1 active crossover, Passlabs X250.5 reference amplifier
    Attached Images Attached Images      

  5. #5
    Senior Member
    Join Date
    Apr 2003
    Location
    Australia
    Posts
    7,942
    It should be noted that the outright results came from more than mere capacitors upgrades but from more than a dozen other modifications to the signal path of the 3145 stock network.

    These modifications required an extensive study of electrical characteristics of the network, of the layout and signal routing which is quite complex as mentioned in the previous posts.

    As a measure of the complexity an error was found in the connection of the inbound signal polarity to one of the HF/UHF Pcb lug / crimp terminations. It was like this when Clark opened up his 4345's for the first time. He did not disclose this until I examined the networks so I could verify the error. The result was on one channel the horn and slot radiator were wired out of phase. This would serverly impact on the tonality and imaging of the system. I am unaware if the previous owner was aware of this fault.

    In summary the upgrade path required a review of the termination’s, signal routing and switching and parts quality.

    This lead to improvements with new bindings posts that had to be carefully fitted to the rear terminal plate. The stock terminal posts used a steel bolt to hold the on the rear plate while on the inside held in place a large crimp lug.

    An improved signal routing scheme that split the mid and high/uhf signals direct from the binding posts direct to critical placement on the main and high/uhf pcb was required. This took a lot of thought and review of the existing wiring.

    The result is a direct line signal path that completely bypasses all the lossy effects of the switching and delivers the signal for both the active and earth direct to specific locations adjacent to the filter elements on the pcbs thus eliminating any cross talk.

    We also discovered the stock networks loop the output of the filters back to the pcbs via the L pads and then back to the drivers. Crimp/lug connectors are a neat and simple means of connection. But in the interests of the purity of the audio signal properly prepared direct soldering is a better approach while the number of termination should be minimised where possible.

    Of course the best results regardless of the network can only be obtained with very careful setting up of this type of system.


    It was suggested that a new wiring loom be prepared that was soldered direct to the pcb tags and routed directly to the drivers via the L pads or fixed value resisters. The series resisters in the midange fixed pad were upgraded to Mills quality audio power resisters.

    Images: Pcb wiring terminations, old binding post, new binding post and direct signal routing.
    Attached Images Attached Images      

  6. #6
    Senior Member
    Join Date
    Apr 2003
    Location
    Australia
    Posts
    7,942
    Conclusion

    If asked what is the difference.

    The stock 3145 sounds safe but in comparison it sounds fluffy and plastic with an artificial quality compared to the modified/upgraded network which gave the system significantly more clarity and a natural tonal contrast without sounding bright and unnatural like many other capacitors we evaluated.

    Please note unless you have the resources, appropriate skills and experience I would recommend a re build of an equivalent network by a qualified technician rather than attempt to modify a stock 3145 network.

    I must say my involvement with Clark during the project has been a lot of fun and has resulted in a mutual friendship.

    For those interested Clarity SA Cap and Auricap bypass modification is becoming a well regarded combination among those who are interested in the ultimate diy loudspeaker capacitor upgrade. The price / value ratio is indeed impressive.

    We felt in the end it was the most appropriate conversion given the original stock network was also configured with bypass capacitors (polypropolyene bypassing mylar).

    Finally, and this applies to any tinkering or tweaking of your crossovers. Its all relative to the rest of your equipment and you should spend time time investigating what will work best in your system.

    Ian

    Response from Clark:

    http://www.audioheritage.org/vbullet...&postcount=130


    Ian has indeed been kind enough to upgrade my 3145 networks. He simplified the circuit by eliminating or bypassing the pieces rendered unnecessary by biamping. That and upgrading some parts of the conducting path reduced insertion losses significantly. The capacitors were then upgraded to Clarity cap SA's.

    Since I had to be without them for a while, I had been listening through Stax headphones in the meantime. I was expecting the 4345's to be quite a step down from that, but the differences in quality were quite subtle. The improvement is really that great!

    The detail, despite not yet being bypassed, was much finer and sharper than before. Imaging was better and the sound was just plain more natural. I was sceptical if bypassing would improve things much, but since the network was designed to operate that way, I pressed on.

    I installed Auricap .01mf caps in all the bypass positions. The quality again leapt upward. Detail went through the roof, getting much more out of my SET amp and the wonderful transducers the 4345 holds. More, frankly, than I thought possible from these speakers I keep reading are out of date and obsolete. The imaging was much better still, and Patricia Barber's CD's sounded every bit as good as they did through the Stax. Better in some ways, actually. The combination of the superb balance of the system with the level of detail, and the naturalism of the voicing and the tone, is unreal.

    I am especially impressed because the improvements make them much better 4345's, but still 4345's. If they had ended up sounding great but like something else, I would not be so happy with the results. All the strengths and great character of the monitors remain, but improved in every way.

    Please don't dismiss these upgrades as a poor man's alternative to new networks, bypassed or not. I don't know if more money would yield better results, but the improvements gained in this way are spectacular. If your sources and electronics are up to the task, you will be amazed what the old 4345 can do for you.

    The only downside is if your sources or amps are crap, you will hear the weakness right now. I switched out pieces to experiment and quickly found out what gear was up to it and what stuff was not. This system is now very, very revealing.

    At some future date Ian and I may get around to documenting this step by step. Make no mistake, Ian is very good with this stuff, and frankly he did pretty much everything. And I am so grateful.

    Clark
    __________________
    Attached Images Attached Images  

  7. #7

  8. #8
    Senior Member
    Join Date
    May 2007
    Location
    SoCal
    Posts
    128
    Ian,

    Probably a difficult question, but how do you think this stock upgrade compares to the charge-coupled 3145 crossover.??

    BTY... Althoough not in the same league, I have just completed an upgrade of a N2400 x-over using ClarityCap SA and Alphacore Inductors. Honestly they sounded like sh*! at first ( i really thought i f*cked up) but now after a few weeks of breakin-in there is soooo much more detail, texture and realism. The stock vintage x-overs sounded good but vailed ...and as you said ...'safe'.

    Ignorance is Bliss!....with 100dB speakers
    Attached Images Attached Images  

  9. #9
    Senior Member
    Join Date
    Apr 2003
    Location
    Australia
    Posts
    7,942
    CC works well but it depends on everything else in your system.

    It comes down to individual preferences. We figured it out and came up with this after a lot of listening and a lot of thought.

    If there was a one answer fits all there would be very few loudspeakers and equipment offerings in the market.

    Finally please dont turn this thread into another audio blog for opinion sayers.

    The is thread is purely about Clark's project.

    Ian

  10. #10
    Junior Member
    Join Date
    Dec 2019
    Location
    Czech Republic
    Posts
    2

    Jbl 4345

    Huge greetings to all fans and connoisseurs of beautiful JBL 4345! I recently sold my favorite 4350 and decided to build copies of the 4345, but I want to use the 2-inch 375 drivers instead of the 2421.If someone can give me any advice, recommend changes to the crossover scheme, I will be very grateful. With huge greetings to all participants! Gennady.

  11. #11
    Senior Member Odd's Avatar
    Join Date
    May 2009
    Location
    Oslo, Norway ------ Now Portugal
    Posts
    809
    You will find a lot of good information in member 4313b's thread, here
    43XX (2235-2123-2450-2405-CC 3155)5235-4412-4406-4401-L250-18Ti-L40-S109 Aquarius lV-C38 (030) 305P MkII

Thread Information

Users Browsing this Thread

There are currently 1 users browsing this thread. (0 members and 1 guests)

Similar Threads

  1. L100T crossover Capacitor Upgrade
    By L100t Owner in forum Lansing Product DIY Forum
    Replies: 35
    Last Post: 10-29-2010, 06:35 PM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •