Page 4 of 6 FirstFirst ... 23456 LastLast
Results 46 to 60 of 76

Thread: 4-Way Active Crossover Options

  1. #46
    Senior Member
    Join Date
    Apr 2003
    Location
    Australia
    Posts
    7,942
    One of the bug bears with using an active crossover is the issue of ground loops that can intrude as a hum or buzz. In most but not all cases this is mitigated with balanced equipment inputs and outputs.

    However, the balanced system relies on the absolute cancellation of the common mode noise signal on the differential signal. This is very much determined by the absolute impedance balancing of your balanced output of your equipment and balanced inputs. A mismatch of 1% can have serious detrimental effects on the effectiveness of balanced connections.

    Secondly, balanced equipment inputs are in face much noisier than unbalanced inputs. A balanced input typically has a signal to noise ratio of -104 db. An unbalanced input is typically -117 db. This has to do with a trade off between common mode noise rejection which requires high impedances to work and the inherent noise of a high impedance circuits. The noise performance can be significantly improved but it’s more expensive to design and manufacture so most equipment manufacturers not bother to tell you this.

  2. #47
    Senior Member
    Join Date
    Apr 2003
    Location
    Australia
    Posts
    7,942
    As a rule the noise level should be at least 70 db below the lowest signal level in normal operation. Not the manufacturers rated noise performance at the rated output. The reason is that in practice your signal from your preamp could be -52 db below the rated output of your 200 watt Macintosh power amplifier.

    To prove this where the power amp has a balanced voltage gain of 23 db and a rated output voltage of 40 volts for 200 watts rms into 8 ohms, the input voltage required the rated output is 2.83 volts from your active crossover and your preamp.

    But for a mere 0.1 volt output into your loudspeakers which is quite audible the input signal to your power amp from your active crossover is -52 db below the 2.83 volt input level for the rated output of 200 watts. If your noise floor is at or around this level then your dynamic range is 52 db which is not really acceptable. As l mentioned above your noise floor should be way below the threshold of audibility. That threshold might be another -20 below that 0.1 volts of noise at your loudspeaker terminals. That’s -72 db below the 2.83 volts input for the rated output of the power amplifier. The power amps noise level might be -110db.

  3. #48
    Senior Member
    Join Date
    Apr 2016
    Location
    Winnipeg Manitoba
    Posts
    279
    Quote Originally Posted by Ian Mackenzie View Post
    Not many posts here in recent times.

    Just wondering if anyone wants to share their active crossover experiences?
    After starting this post I did end up buying an Xilica XP-4080. I ended up parting out these speakers and building a pair of 4345 clones. I used the Xilica with the 4345s for a bit, but running four amplifiers was a real pain. I then built a passive crossover for the 4345s and used a 2-way Bryston 10BLR.

    My current set-up is a pair of M2 clones with a Crown DCI 4x2400N. I got rid of my pre-amp and DAC and just use a BSS BLU-USB connected to my amp with a Cat 6 cable, nice and simple. My Mac keyboard is the volume control and I have a Furman RS-2 remote control panel and MP-30 to turn on and off the amp. The amp is in my basement right below the speakers. It had really loud fans so I had to move it to a different space.

    To sum it up I would never do a four way active system again.

  4. #49
    Senior Member
    Join Date
    Apr 2003
    Location
    Australia
    Posts
    7,942
    The story does not end there.

    If your crossover has a rated noise level of -110 db then it’s connected to your power amp with a rated noise of -110db the noise level jumps to -104db. How do? The residual noise level of the crossover and the power amp sum together or double. This is a +6db level increase. If your crossover has a noise level of -104db the crossover will be the limiting factor. In practice this noise level is only -32 db below the example above. When your turn your preamp right down to just audible it’s around -35 on the volume scale.

    With these relatively sensitive JBL systems that sensitivity is your friend for those dynamic peaks but your foe with that residual noise floor.

    So how do l reduce my noise floor?

    As equipment designers we look to what is referred to as low impedance design. In audio circuits there are a number of noise sources. Johnson noise is the noise generated by resistance values in the circuits. To reduce the noise by 10 db we need to bring these circuit resistance value down to 1000 ohms or lower. However the monolith Opamps used in this kind of equipment begin to loose linearity and the distortion level rises rapidly.

  5. #50
    Senior Member
    Join Date
    Apr 2003
    Location
    Australia
    Posts
    7,942
    Another problem arises. The monolithic op amp only has a small fraction of output current in class A. It therefore has to drive this load in class B. But class B leads to higher distortion and noise of the circuits power supply.

    The brute force way around this problem is to use parallel monolithic opamps which is a practice in some mixing desks. The more elegant and superior solution is to employ a discrete operational amplifier. A discrete op amp can drive a 75 ohm load is some cases all day every day. The trade off is cost and power consumption. Your power supply will need to be re designed and upgraded. So the overall manufacturing cost goes up and you have to pay a more.

    What this comes down to is don’t skimp on your active crossover. It kind of defeats the purpose of the exercise overall. Particularly if you have a high or premium grade of other audio equipment.

    Digital is one path but take Riley Casey’s comments into consideration.

  6. #51
    Senior Member
    Join Date
    Apr 2003
    Location
    Australia
    Posts
    7,942
    My analogue crossover l have mentioned will use discrete operational amplifiers and low impedance design. The estimated cost ex works will be around US$2400 mark. In adjusted US$ the JBL DX1 would be US$2110 today. This thing will cover the pre determined loudspeaker category and diy / project users with a suitable measurement kit. I will be looking to obtain a BSS OmiDrive or evaluation. I think it’s a case of being aware of developments in this area as it is the gate way to your end result.

    Getting something really good and knowing how to use it properly can in fact mean a much shorter journey in terms swapping and changing out horns and drivers. This is because your finding out first time around what’s really what. You definitely don’t want to be stuffing around too long because it will become wearing and ultimately frustrating.

  7. #52
    Senior Member
    Join Date
    Apr 2003
    Location
    Australia
    Posts
    7,942
    I have seen some people upgrade the legacy JBL 5234/5235 active crossover with new monolithic opamps in the past.

    This is a good idea where the Opamp used was a TL074. This Opamp and is dual package TL072 have a much inferior common mode distortion performance when used in Sallen Key active filters (1) (2).

    Unfortunately there are relatively few good audio opamps in a quad version without looking at a SMT package. Texas Instruments now manufacturer superior monolithic opamps if their marketing can be believed. However, not many of their Soundplus range of opamps are only specified for their rated performance in 2k ohm loads. All of these op amps rely on complex distortion cancellation circuits to produce low distortion in class B operation. We can’t see these cancellation systems because Texas Instruments do not disclose the Opamp schematics.

    Class B operation is used because of its efficiency. Because everything in a monolithic Opamp is contained on a tiny silicon substrate problems associated with heat and temperature must be avoided.

    (1) Douglas Self . The Design of Active Crossovers.
    (2) John Caldwell . Distortion and source impedance in Jfet opamps

    https://www.ti.com/lit/an/slyt595/sl...-df-pf-null-eu

  8. #53
    Senior Member
    Join Date
    Apr 2003
    Location
    Australia
    Posts
    7,942
    In the previous post l discussed some of the performance limitations of monolithic opamps in audio applications.

    But this is not really new news. Dean Jensen, an astute and very talented American audio engineer realised the limitations of monolithic opamps in the sound recording and motion picture industry back in the 1970’s. What he did off his own bat was thoroughly research the subject of operational amplifiers and then designed a discrete operational amplifier and patented the design. Jensen then made the patent public. The Jensen 990 discrete op amp became the industry standard. All discrete operational amplifiers used in professional recording microphones are based on the original work of Dean Jensen.

    Dean Jensen worked in the installation of large format studio mixing desks in the 1970’s. Jensen went on to establish Jensen transformers because he was dissatisfied with the performance of audio transformers of that era. Bill Whitlock, a close friend of Jensen refers to their technical efforts and that of many others as the sound of the 1970’s. Whitlock reasoned it was not one particular piece of equipment but the collective sum of the recording process that gave analogue recording a particularly natural sound that has been adopted as the vinyl sound today.

    Today recording studios with analogue equipment are highly revered by well known artists.

  9. #54
    Senior Member
    Join Date
    Apr 2003
    Location
    Australia
    Posts
    7,942
    So why are discrete operational amplifiers not more widely used in professional and consumer electronics. The answer they are used in premium equipment but that comes at a cost. The question then is is it justified? Only the user can answer this question. If you own a quality record collection, a quality turntable and associated audio equipment you will probably justify the cost over the mass market lower cost monolithic Opamp based equipment.

    If your largely a digital based user who streams largely from a provider and has dsp based equipment you might not justify the cost.

    The cost of a discrete Opamp as component starts at US$50.00 and goes up from there. In comparison a NE5532 monolithic opamp costs US$0.30. That’s a whole magnitude difference in cost and some equipment would have ten or more opamps. If you cringe at the price then think about the cost of manufacture and inventory. This is why such manufacturers are small boutique businesses. Their products that they manufacture are highly sort after by those who place value on this level of equipment.

    One way of looking at is to draw an analogy with maintaining your jbl driver collection. These days you can buy a new replacement woofer for the cost of the repairs. But once it’s gone then it’s gone forever and that sound you cherished.

  10. #55
    Senior Member
    Join Date
    Apr 2003
    Location
    Australia
    Posts
    7,942
    Riley Casey commented in another thread that some in built Eq in an active crossover can be quite useful.

    Without doing a deep dive into what EQ can and can't fix as a rule EQ can't fix phase cancellations caused by room modes which cause deep narrow dips. This is also true of phase cancellation in the crossover region due to differences in path lengths in sound waves reaching the listening position.

    EQ can however be useful to moderate some room induced effects in the low frequency response. These effects depend on the location of the loudspeaker and the listening position in a room.
    Typically bass cut is required to reduce a peak causing boom or a broader shelf cut action where the loudspeaker is located near a floor and wall boundary. These effects can be in the order of 3-10 db. Other room effects can occur between 60-150 hertz as a series of peaks or dips where sound waves wrap around the enclosure and either cancel or add. The best approach is to use modest cut to subdue these effects. So this EQ might be presented by an LP shelf EQ and a parametric EQ. Measurements are best used to assess where the EQ is required.

    Other room effects are more subtle but can influence the perceived clarity of the system. For example it is not uncommon for a room to have a resonance in the 250-340 hertz region which can mask midrange clarity. This can be dealt with using a fixed frequency low Q EQ and a switchable level of attenuation.

    EQ can also be useful where the actual monitor has a broad frequency response variations both on and off axis. This can be assessed using high resolution measurements on and off axis and averaged to provide broad low Q EQ in subtle amounts of 0.5db to 2.00 db correction in defined areas. The correction can be preset on a plug in EQ card in the crossover or a user access switchable corrections in incremental amounts to their own taste. This is the power of a high quality active crossover. The key is having very accurate data from which to make these corrections. In most cases its simply a case of measuring a system in a defined environment accurately and then determining the EQ correction. This can be done with a test system. Unfortunately an accurate set of measurements is unlikely to be accomplished effectively by the user at home unless they are enthusiastic.


    This is a much more tidy approach than offering the users a knob farm to fiddle with. Your time is far better spent listening to your music than messing around with EQ adjustments when your not sure what your really doing.


    The legacy systems have somewhat sloppy and imprecise L PADS.

    There are several ways of dealing with this to obtain a level overall tonal balance. The L pad can bypassed with precision calibrated fixed pad for the mid, horn and slot. Or the calibrated pad could have three or five switched positions. I can organise that to replace the variable L PADS. a descent Greyhill or Lorlin rotary selector switch and a small pcb with some MILLS resisters is a better way. The mid LPAD can be a fixed pad when bi amping.

    Alternatively the user can conduct their own measurements using REW and trim the L PAD settings. I am inclined to view this as messy.

    By adopting a structured approach all these small variables can be defined and measured. The result is a significant improvement in L+R tonal balance matching and stereo image. In my experience this kind of approach makes a surprising subjective improvements. I experienced this when Greg Timbers set up and bi amped a 4331 system at his home and we were very pleased with the results. Its s a very good loudspeaker if properly set up. So you imagine the improvements with a more sophisticated system.




    .

  11. #56
    Administrator Robh3606's Avatar
    Join Date
    Apr 2003
    Location
    Rocinante
    Posts
    8,170

    Post

    EQ can also be useful where the actual monitor has a broad frequency response variations both on and off axis. This can be assessed using high resolution measurements on and off axis and averaged to provide broad low Q EQ in subtle amounts of 0.5db to 2.00 db correction in defined areas.
    Hello Ian

    Just adding a little clarity. For Constant Directivity systems, think 4430 where the on and off axis response are very similar you would get the most benefit. With system using beaming horns or systems with abrupt changes in directivity think 4333, Urei 811 this will not work as well. So it's important that the user understands the ramification's of trying to "fix" the overall power response and the influence that directivity plays on the effectiveness of EQ.



    This is a good read

    http://www.audioheritage.org/vbullet...me-1-Number-11

    Rob
    "I could be arguing in my spare time"

  12. #57
    Senior Member
    Join Date
    Apr 2003
    Location
    Australia
    Posts
    7,942
    Hi Rob,

    That is exactly right.

    That’s why l think offering some preset EQ options that have been thought through carefully is possibility. The measurements l or Greg would do on a test system such as in the case of the 4333 would need averaging and then related to how the user positions the system and their room acoustics. Two or possibly three different preset of EQ would be available depending of the users set up.

    People tend to ask too much of EQ according to Nelson Pass. If you try to knock out a kink in the response completely it can cause other problems. What Pass found with the late Dr Linkwitz in the LX Linkwitz crossover was that the more gentle analogue EQ compared to the abrupt DSP EQ was indistinguishable in carefully listening by Dr Linkwitz. The consensus was the analogue was subjectively better compared to the mini dsp. That’s really interesting. Dr Linkwitz was a very fussy and a very smart guy so it just goes to show what can be done when it’s done carefully and correctly.

    It’s a space that needs a thorough examination or trials by users for feedback

    I have measured some polar responses of the 43XX system and they track not bad over defined arch. But of course toe in has an impact. JBLs suggestion of a 5-10 degree toe is in fact because the 4344-45 measurements are smoother with small toe in and you get less differential reflections off the walls. They know their designs well. LF problems in rooms is where EQ can make obvious clean up.

    Used sparingly to solve specific issues EQ is a useful tool in your tool box.

  13. #58
    Senior Member
    Join Date
    Apr 2003
    Location
    Australia
    Posts
    7,942
    An extension of classic analogue EQ in the studio recording process is the selective addition of tonal colour. This is done using various techniques which exhibit saturation. The effect is what Neve call silk in their high end equipment. API and other legendary studio equipment manufacturers have this genre of analogue equipment used by major studios. The intention is to make an otherwise lifeless digital stream more natural. In the past some consumer CD players used broad bush tube buffers. The effect is more selective and precise in analogue EQ.The audiophile does this by trading in and out various brands of equipment until they feel comfortable with the subjective sound. There is therefore application of analogue EQ for the digital streamed program source.

  14. #59
    Senior Member
    Join Date
    Apr 2003
    Location
    Australia
    Posts
    7,942
    When the weather here clears l will post some driver and system measurements with and with out EQ for illuminative purposes.

    Another application of structured design is to provide precise EQ for different compression driver diaphragms, alternative drivers, horns and wave guides.

    This give the adventurous Thinking Cap Customer the opportunity to explore with confidence the subjective qualities of alternatives. People who enjoy high fidelity sound reproduction are generally curious individuals. The above pre determined EQ gives this interest group the opportunity to efficiently sculpt their own sound. Offering the ultimate frontier. But done right the first time “in a systematic and controlled manner”. That’s a phrase straight out of a JBL brochure.

    The implementation of these approaches can be arranged by plugging in a card specific to the users drivers and system in use. The thinking’s been done. A small amount of user engagement might be offered for fine adjustments within a defined range. A 0.75db change in sound level is generally accepted as the perceived difference in sound level. (1). Wow l did that and l like it better! But the subjective effect of such small measured differences can have a noticeable effect on tonal balance with low Q equalisation.

    (1) Douglas Self. On psycho acoustics and subjectivism.

  15. #60
    Member
    Join Date
    May 2022
    Location
    Ontario Canada
    Posts
    86
    I'm using a JBL DSC280 in my main setup.
    Works very well for my needs

Thread Information

Users Browsing this Thread

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

Similar Threads

  1. jBL-UREI 811C crossover options?
    By dubkarma in forum Lansing Product DIY Forum
    Replies: 2
    Last Post: 12-15-2017, 10:49 AM
  2. Crossover options
    By migman in forum Electronic Crossovers
    Replies: 3
    Last Post: 03-28-2013, 09:19 AM
  3. 4722-HF (2432H) to 2226H crossover point/slope (active crossover)
    By Jonas_h in forum Lansing Product DIY Forum
    Replies: 6
    Last Post: 04-25-2012, 12:17 AM
  4. Passive Crossover Options
    By upstreme1 in forum Lansing Product Technical Help
    Replies: 1
    Last Post: 05-01-2009, 03:14 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
  •