Page 2 of 2 FirstFirst 12
Results 16 to 19 of 19

Thread: BSS FDS-388 and 366 Omnidrive

  1. #16
    Senior Member
    Join Date
    Apr 2003
    Location
    Richmond Hill, Ont.
    Posts
    4,682

    Talking Oh The Irony !

    Hi Jim

    My 2 cents ( okay there's more than that ) on this topic is;

    I'd be more concerned with maintaining correct phase relationships than "time-aligning" those two components.. I realize that's an obscure / meaningless statement so I'll give you an example of a present dilemma I'm trying to work out a solution to . It all goes back to my playing around with DC blocking capacitors ( hence the title of my injoke - Ian should appreciate this ) & listening to the sonic signature of different dielectrics .

    Some background; I use a 24 db/octave, LR slope type , active crossover ( 1st generation Behringer ). My only HF EQ is a simple RC network that depresses the mids and effectively adds a 6db per octave "boost" that fights against the 9 or 10 db/octave roll-off of my big 2441 drivers. This isn't a "flat" curve but the net result sure does sound musically correct to me. This setup on a small round-mouthed horn gives stellar results, like pin-point imaging and great depth of sound-stage. No synching up of the woofer & compression driver is involved or is a contributing factor in my musical enjoyment. The voice coils have about a 4" offset .

    BUT - add in those damn DC blocking caps and my great soundstage collapses or compresses. For a couple of weeks I thought I was listening to a particular quality of the caps themselves until I twigged that this problem happens with all my types of DC blocking caps. The lurching-trail of logic has become that I am now reacting in a very negative fashion to an LR circuit that has been tragically mucked with giving the Hi-pass secton a 450° phase shift. This is easy to do with active components . For two years I've listened ( more or less ) to music with no DC blocking caps in place. So, the 90° phase-shift added onto that of the 360° hipass section has created something quite distracting to my ears. I can time-align the 2 sections but my ears tell me its far from the same. And I think the analog "bucket-brigade" delay section of the Behringer is a linear-phase type that begins correct in the crossover area and then just keeps adding increasing delay ( and phase offsets ) into the low pass section. In a couple of weeks , when work patterns permit, I'm going to turn my DC blocking caps into a 12 LR style network ( gloomed ontop of the choosen active frequency ) - then flip the drivers polarity and see if this does the trick for my ears by restoring the lost LR phase relationships . Going all passive may be a better option .

    Obviously the domino effect of one change can be quite staggering .

    Can I ask why you are concerned with time-aligning? It's mostly relevant ( to vertically alligned transducers ) within a single horizontal plane ( while these pesky phase anamolies will exist over a much broader listening area ).

    If you go digital; I'd suggest a CD player with a digital out allowing you to keep it digital until you get to the best sounding DAs in your setup . This could also mean buying Apogee DACS and using them after the crossover if it has a couple of AES digital outs - which I can't remember if it does .

    regards <. Earl K

  2. #17
    Administrator Robh3606's Avatar
    Join Date
    Apr 2003
    Location
    Rocinante
    Posts
    7,984
    Hello Ian still

    Something seems weird there. If you look at the polars for the 4430 the 15 degree tilt looks to be way too much. You see the best at 0 and 10 up. At 20 you start to see some problems. Also if you apply that logic to the 4345 you have one axis from the midrange to the compression driver pointed at the ceiling. Much more tilt than the 4430 because of the much longer length of the horn. The compression driver to slot is pointed at a side wall. I must be missing something You have to include the crossover slopes in there too don't you?? 12db network Isn't the compression driver 90 degrees ahead and the woofer 90 degress behind so you invert phase. So you effectivly have just " physically moved" the driver 1/2 wavelength.


    Rob
    Last edited by Robh3606; 10-09-2003 at 03:32 PM.

  3. #18
    Senior Member Ian Mackenzie's Avatar
    Join Date
    Apr 2003
    Location
    Australia
    Posts
    7,129
    Robh3606,

    (Going a bit off topic here...)

    Yeah, sorting this out its like discerning French & English mustard, a mouthful of each and you can't work out which is which. (also both make you sneeze and blast hot wind)

    My reference to the tilt was from the 4430 White paper in the AES.

    Just looked at those curves myself, the 10+ degree up curve is quite smooth +-2 db and next one is +20 grees with 5 db variations. If you go down there are deeper troughs for sure, but only over a narrow band, I think this is JBL point and they would be eq'd out at the desk listening possy anyways.

    I appreciate your comments regards the 4345, I'm not sure what the deal is as there are no known off axis curves that I am aware .

    But looking at the 3145 network intuitively, the 4345 uses sharp high pass filter slopes in the upper mid and UHF regions of 18 db and this would mitigate the lobing effects in the crossover regions to a large degree. The slopes are also asymmetric 12db low pass/18db high pass and wired in phase.

    The lobes are only an issue in the region of driver overlap, this is why the alignment is a big deal with 1st order systems like Duntech, Dynaudio etc.

    It is normal practise to amend the LC values to minimise the lobing effects on axis or +- 30 degrees off axis. Having said that the 4345 is normally run with the slots on the outer left and right, could be the lobe is pushes inward in the crossover region

    One can also reasonably assume the network in the case of the 4345 was smoothest on axis by design, and was not intended as a monitor with flat power response.

    When I am in the test phase of the 4345 diy project I will certainly run some on/off axis curves and post them.

    It would be interesting if Giskard can slooth from memories from GT on this topic.


    Ian

  4. #19
    Senior Seņor boputnam's Avatar
    Join Date
    Apr 2003
    Location
    northern california
    Posts
    7,007
    Originally posted by Ian Mackenzie
    Yeah, sorting this out its like discerning French & English mustard, a mouthful of each and you can't work out which is which. (also both make you sneeze and blast hot wind)
    Then, of course, there's always VEGEMITE, the attributes of which completely evade me...

    DamnYankee
    Attached Images Attached Images  
    bo

    "Indeed, not!!"

Thread Information

Users Browsing this Thread

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

Posting Permissions

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