The big technical downside to active crossovers

Discussion in 'Hi-Fi Stereo Speakers' started by lindsayt, Aug 17, 2018.

Tags:
Thread Status:
Not open for further replies.
  1. lindsayt

    lindsayt
    Active Member

    Joined:
    Feb 9, 2011
    Messages:
    299
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    28
    Ratings:
    +48
    Let’s have a look at what’s inside a “simple 3 way analogue active crossover”.

    [​IMG]



    Source: Simple 3-Way Active Crossover Circuit Diagram | Circuits Diagram Lab

    If we trace the signal path from the input to the midrange output we can see that it goes through IC1, IC4A, IC4B, IC5A, IC5B, IC7A.

    Looking at the parts list for this circuit we can see that that’s 1 x TL071 and 5 x TL072’s.

    If we then look at the Texas Instruments Functional block diagram on page 26 of their TL07xx series Data Sheet:
    http://www.ti.com/lit/ds/symlink/tl071.pdf

    we can see that the signal passed through something of the order of 8 transistors in each TL071 or TL072 op amp.

    This means that this active crossover adds 48 active amplification devices (transistors) to the signal path!

    For reference purposes, passive crossovers add 0 active amplification devices to the signal path.

    For further reference purposes, my Phillips CD753 CD player has a DAC chip and then an analogue section that consists of half an op amp per channel. So let’s call that 4 active amplification devices. In my main system I have a stepped attenuator (passive) pre-amp with 0 active amplification devices and then a SET amplifier with 4 active amplification devices (2 x valves and 2 x transformers) in the signal path. If I were to use a solid state integrated amplifier I’d have something of the order of 8 active amplification devices in the signal path, in addition to the 4 from the CD player.

    My conclusion from this is that when comparing active to passive crossovers, one should be aware of the SHOCKLINGLY HUGE number of amplification devices that active crossovers may add to the signal path.

    OK that was one example of a 3 way active crossover. In a 2 way 4th order active crossover we can expect to have 4 op amps (c32 transistors) in the signal path. Whilst a 2 way 8th order active crossover will have 8 op amps (c64 trasnsistors) in the signal path!

    Please feel free to examine the circuit diagrams of any active crossovers that you may own to determine how many active amplification devices your active crossover is adding to the signal path.
     
  2. Ugg10

    Ugg10
    Well-known Member

    Joined:
    Apr 17, 2009
    Messages:
    2,933
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    136
    Ratings:
    +1,400
    I guess the questions is - what effect does this have on the sound when comparing active to passive - I assume you are asserting that a passive system should sound better as it has less active components that all have a level of error/distortion that is additive and affect the sound quality.

    From my own experience I have moved from a traditional active preamp>poweramp(s)>passive speakers to a passive preamp>active speaker setup. The two system had a similar retail price when new (bought second hand) and were in the mid to lower end of the market (c. £1500-2000 ten years ago). My A/B listening with the same source (CCA + dual Wolfson WM8741 DAC) and in the same room on the same day concluded that I prefer pretty much all aspects of the active system and so have moved that way. Noting this was in my room with my ears and my brain - the critical point here !

    But, I am sure if someone else did the test they may come to the opposite conclusion.

    Your thoughts ?
     
  3. jamieu

    jamieu
    Active Member

    Joined:
    Jul 3, 2010
    Messages:
    132
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    31
    Location:
    London
    Ratings:
    +58
    And how many IC's and transistors did that album you're listing to on your setup pass though in the studio before it was even mastered onto a delivery format?

    Obviously the more complexity you put in the signal path the greater the room for error, but complexity of electronics shouldn't necessarily equate to a loss in sound quality if designed well.
     
    Last edited: Aug 17, 2018
  4. Deleted member 781788

    Deleted member 781788
    Guest

    Products Owned:
    0
    Products Wanted:
    0
    Ratings:
    +0
    Your circuit is complex, true, but modern op-amp chips (even some older ones ne5532) are very very good indeed.
    Passive crossovers are not that simple either.
    A cheap ferrite cored inductor, sounds inferior to a ribbon air cored one for various reasons, similarly a non-polar electrolytic capacitor sounds inferior to a quality polypropylene equivalent - that means expensive ! ! by a few 100% s.
    Also in chip domain, phase accuracy, accurate filtering is much easier achieved than passive.
    An active system has poweramps connected directly to voice coils, meaning the amp can control the cone much better than in a passive system, and the poweramps do not have to deal with reactive-capacitive loads, so they can function better and made cheaper.
    I agree that a properly designed loudspeaker cabinet/driver that requires only a simple crossover circuit to function, would sound better in passive compared to an active version . (provided quality components are used in the passive crossover).
    But a poorly designed (often small) loudspeaker cabinet/driver would sound better if it is active, because the design shortcomings can be ironed over using complex active circuitry.
    Usually large driver (8 inch or more) quality speakers tend to be passive, since the back-emf of a large driver is less, driver matching is easier ( bass, midrange and HF), so they require simple first order crossovers.
    Almost all active speakers (sub 5K) tend to be smaller boxes! (with exceptions)
    Magico - B&W are probably the best reference manufacturers I can think of, only do passives.
    I am personally a purist in my circuit approach, that's why I have moved away from Actives to tube/valve amplifiers and passive speakers.
    The only chips in my system are in my DAC, but unavoidable.
     
    • Like Like x 1
    • Agree Agree x 1
    • List
  5. Deleted member 781788

    Deleted member 781788
    Guest

    Products Owned:
    0
    Products Wanted:
    0
    Ratings:
    +0
    True but the answer to your question is tried and tested one:
    - "Anything that happens to the music at production level, is part of the creativity and can not classed as anything else."
    The roaring of an electric guitar is nothing but a deliberate distortion and not an actual distortion.
    The job of a playback system (HiFi) is to be true to this reference and not beyond.
     
  6. jamieu

    jamieu
    Active Member

    Joined:
    Jul 3, 2010
    Messages:
    132
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    31
    Location:
    London
    Ratings:
    +58
    True, but a convenient cop out as well ;-)

    My point is that what matters is the effect on the signal, not how many electronic components it has passed though.

    Get a scope out and see what effect a particular crossover has on a signal. My guess is that both do a similarly good job when well designed and paired. But as you point out, a well-made active crossover will have the edge in dealing with compromises in cabinet size or drivers, which overall will likely result in a more accurate sound reproduction.

    As for larger 3-way cabinets I think the jury is probably still out, particularly as there is an inherent, but not unreasonable human trait to gravitate towards the minimal (great equations tend to reduce to a simplistic form) and hence the search for that zen combination of passive crossover and drivers. But there’s also a convincing case for dividing up the frequency range electronically, where crossover slopes and phase alignment can be perfectly adjusted for, then letting a single amp drive each driver, particularly as power requirements rise.

    When well designed, both probably lead to the same place, I guess it comes down to which road you want to take to get there. The analog lane, traditional and picturesque, or the digital highway, all concrete and blinking lights ;-)
     
  7. Trollslayer

    Trollslayer
    Distinguished Member

    Joined:
    Feb 11, 2007
    Messages:
    24,251
    Products Owned:
    0
    Products Wanted:
    1
    Trophy Points:
    166
    Location:
    Poole
    Ratings:
    +9,556
    It takes a lot more than a good oscilloscope to understand how well audio equipment is performing.
    As to the number of transistors, it is far more complex. Saying 'transistor' is like saying 'car' - Bugatti or Trabansk?
     
  8. Deleted member 781788

    Deleted member 781788
    Guest

    Products Owned:
    0
    Products Wanted:
    0
    Ratings:
    +0
    Every component (active or passive) degrades the sound, Active ones do more.
    There are whole threads about active vs passive preamps.
    Again, it is irrelevant how many components the music goes through at production, once it is finalized, it must be degraded as little as possible from there on.
    You do have a point for album re-releases, but that's another argument.
    There is a separate argument for benefits of larger cabinet/drive diaphragm combo's that needs an entire thread, suffice to say, they need simpler passive crossovers (means fewer components) and therefore shorter signal paths.
    As you agree, active can correct ( I call it ironing over) compromises in speaker design.
    I put it to you that it is a case of "horses for courses".
    quality speakers have less compromises in design, so they sound better in passive, and they tend to be large.
    Smaller speakers, do have compromises in design, so Active works for them.
    How easy is it to upgrade the amps in an Active system ??
    it's a doddle in a passive system!
     
  9. Trollslayer

    Trollslayer
    Distinguished Member

    Joined:
    Feb 11, 2007
    Messages:
    24,251
    Products Owned:
    0
    Products Wanted:
    1
    Trophy Points:
    166
    Location:
    Poole
    Ratings:
    +9,556
    Couldn't have put it better. :smashin:
     
  10. lindsayt

    lindsayt
    Active Member

    Joined:
    Feb 9, 2011
    Messages:
    299
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    28
    Ratings:
    +48
    Yeah, but in the case of most modern active speakers, they use analogue op amp based active crossovers.

    Op amps and general purpose circuits. The Texas Instrument op amps given in the example in the opening post can be used in a variety of applications. They are not specifically tailored for the specific purpose of being in the crossover of any particular active speaker. They are a cheap solution.

    Using an op amp based active crossover is like adding 6 cheap solid state amplifiers in series -one after the other - to your system / signal path.

    For an example of the harm done by adding all these devices, look at figure 19 on page 24 of the Texas Instruments TL07X data sheet. Look how the signal overshoots by 20% and then undershoots by 5%. Now multiply that by 6 times - one op-amp after the other. What you then have is a musical effects box and not musical fidelity!
     
  11. lindsayt

    lindsayt
    Active Member

    Joined:
    Feb 9, 2011
    Messages:
    299
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    28
    Ratings:
    +48
    I agree entirely about it being horses for courses too.

    However, it's very clear that active crossovers are not technically better than passives.

    Not when op amp based active crossovers - as used in the majority of active speakers sold today - add such a huge amount of active amplification devices to the signal path.
     
  12. Abacus

    Abacus
    Active Member

    Joined:
    Jan 19, 2015
    Messages:
    642
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    67
    Location:
    Torbay
    Ratings:
    +423
    A passive system can only ever take things away, whereas an Active system can add and subtract things, now as very few things can fixed by just taking things away, the active system has a better chance of getting everything right.

    Think of it like a car, the old ones had carburettors (Passive) which did the job; newer ones have fuel injection (Active) which improves the performance in all departments despite being more complicated.

    Bill
     
  13. lindsayt

    lindsayt
    Active Member

    Joined:
    Feb 9, 2011
    Messages:
    299
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    28
    Ratings:
    +48
    "An active system has poweramps connected directly to voice coils, meaning the amp can control the cone much better than in a passive system,..."

    We need to be very careful with the language used - to avoid making misleading statements.
    The power amps in typical active speaker DO NOT control the cones much better than in a passive system.

    What they do is provide more DAMPING in active speakers than passive speakers.
    We can provide the same or more DAMPING in passive speakers by altering the volume of air inside the speaker cabinet!
    Check out online speaker enclosure calculators, such as this one:
    Speaker Box Enclosure Designer / Calculator

    From a technical point of view, what's the better way to provide more damping to a bass driver:
    (a) put it in a smaller cabinet and use an active crossover that will add 4 times more (not very good quality) active amplification devices to the signal path than what we had before.

    or
    (b) keep our signal path as pure and simple as it was before and make our cabinet internal volume bigger.


    From a marketing point of view (a) may be better. From a technical point of view, (b) is better.
     
  14. lindsayt

    lindsayt
    Active Member

    Joined:
    Feb 9, 2011
    Messages:
    299
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    28
    Ratings:
    +48
    Abacus, would you like to make a comment that is actually on topic?

    That actually responds to the points made in the opening post?

    Please re-read the opening post. Please examine the circuit diagram. Please go check out the Texas Instrument data sheet. And then feel free to make a comment on all of that.
     
  15. Trollslayer

    Trollslayer
    Distinguished Member

    Joined:
    Feb 11, 2007
    Messages:
    24,251
    Products Owned:
    0
    Products Wanted:
    1
    Trophy Points:
    166
    Location:
    Poole
    Ratings:
    +9,556
    Amplification isn't a problem. Noise, amplitude and phase distortion etc. would be.
    Without amplification your speakers wouldn't make any sound.
     
  16. Trollslayer

    Trollslayer
    Distinguished Member

    Joined:
    Feb 11, 2007
    Messages:
    24,251
    Products Owned:
    0
    Products Wanted:
    1
    Trophy Points:
    166
    Location:
    Poole
    Ratings:
    +9,556
    Proof please. Bearing in mind the speakers will be driven by an internal or external amplifier.
     
  17. andy1249

    andy1249
    Distinguished Member

    Joined:
    May 7, 2006
    Messages:
    8,523
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    166
    Ratings:
    +2,579
    Simply counting up the number of amplifiers in a chain and calling the higher number “ bad” is fundamentally incorrect.
    Design is all important.

    All electronic components, active or passive, have some fundamental noise which must be accounted for in the design of any amp or circuit.
    Note “noise" in this context means any distortion or unwanted alteration of the required signal

    Here is a paper from 1994 which covers the basic noise formulas and techniques for eliminating noise.
    https://pdfs.semanticscholar.org/d770/d93b92abf067aab6e469ba816da0d205777e.pdf

    In particular, note section XII “ Noise reduction with parallel input devices” which applies to all “ low noise” op amps.
    Quote from section XII
    “ noise can be reduced to any desired level if N is made large enough”.

    In short , Whether Passive or Active is better in terms of crossover circuits depends on circuit design and the quality of the components used.
    The Quality of any circuit is determined by its input vs output characteristics based on actual vs required performance and not on how many, or the type of, components that are in the chain.

    Active can be better or worse than passive and vice versa , it all depends on the design.
     
  18. Abacus

    Abacus
    Active Member

    Joined:
    Jan 19, 2015
    Messages:
    642
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    67
    Location:
    Torbay
    Ratings:
    +423
    It’s fully on topic, as an active crossover can be designed to balance everything out within the speaker, whereas a passive system can only ever take a stab.

    You are just trying to prove that sticking more things in the path can make things worse, which as any competent designer will tell you is a complete load of poppycock, unless of cause you are such a poor designer that you do not take the characteristics of the components into account in the design, hence ending up with a poor performance. (The post reminds me of H-Fi cables manufactures who pump out one point and ignore the rest so that they can charge a fortune for a bit of cheap cable with a fancy name on it)

    The circuit and components suggested in the original post are just one of a myriad of designs and components out there, and to try and convince people that active is bad by using just one example is misleading at best and total misinformation at worst.

    Bill
     
  19. Ugg10

    Ugg10
    Well-known Member

    Joined:
    Apr 17, 2009
    Messages:
    2,933
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    136
    Ratings:
    +1,400
    Innocent question here.

    Does it make any difference if the crossover or filtering (in its widest sense) goes after the power amp stage or before I.e. is acting at line level volts or at speaker level volts?

    Here is a schematic of my speakers, being active the power stage acts directly on the speaker whereas in a passive setup the crossover is between the power amp and the speaker.

    https://mackie.com/sites/default/files/PRODUCT RESOURCES/SPECS/Spec_Sheets/HR824MK2_SS.pdf
     
  20. Trollslayer

    Trollslayer
    Distinguished Member

    Joined:
    Feb 11, 2007
    Messages:
    24,251
    Products Owned:
    0
    Products Wanted:
    1
    Trophy Points:
    166
    Location:
    Poole
    Ratings:
    +9,556
    Not really.
     
  21. stevelup

    stevelup
    Well-known Member

    Joined:
    Aug 13, 2002
    Messages:
    6,710
    Products Owned:
    1
    Products Wanted:
    1
    Trophy Points:
    136
    Location:
    Swindon
    Ratings:
    +1,221
    Many active speaker designs, especially high end ones, are fully in the digital domain these days so this is becoming less and less relevant.

    Going back to analogue...

    It is just as possible to design an excellent active crossover as it is to design a crappy passive one.

    Likewise, a really good and well designed passive crossover will give better results than a rubbish active one.

    This is a great theoretical discussion, but there's no point in getting religious over it...
     
  22. Deleted member 781788

    Deleted member 781788
    Guest

    Products Owned:
    0
    Products Wanted:
    0
    Ratings:
    +0
    We agree to disagree on that point.
    There is a firm belief & engineering facts, that in Analogue music reproduction, every component added degrades the sound to various degrees, Active components have the added practical disadvantages of PSU noise, overload noise/distortion, IM distortions and so forth.
    Passive components, though not perfect, are not prone to some, unless badly designed and manufactured.
    BUT, YES, design is very important, but it was understood that we were not talking about bad designs. A good designer, always counts the number of components in his signal path.

    At any rate, the circuit offered here is not necessarily indicative of the crop, often the filtering is built into the individual poweramp(s) by limiting their frequency response, for the driver they are connected to.
    Hence Active crossovers can be made very simply, using few components within the design of the amp.
     
    Last edited by a moderator: Aug 18, 2018
  23. Trollslayer

    Trollslayer
    Distinguished Member

    Joined:
    Feb 11, 2007
    Messages:
    24,251
    Products Owned:
    0
    Products Wanted:
    1
    Trophy Points:
    166
    Location:
    Poole
    Ratings:
    +9,556
    I mainly deal with digital electronics but a quick look at the schematic at the beginning suggests that filtering has been added to match the response of the speakers, something an external amplifier cannot do.
     
    • Like Like x 1
    • Agree Agree x 1
    • List
  24. lindsayt

    lindsayt
    Active Member

    Joined:
    Feb 9, 2011
    Messages:
    299
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    28
    Ratings:
    +48
    You have quoted me highly selectively.

    The big point that I made - abundantly clearly - was that the power amps in active speakers do not CONTROL bass drivers in speakers better, they DAMP them better than in a passive speaker.

    It's all about semantics and using the word that most accurately describes the situation.


    The big problem with using the word CONTROL in this context is that it strongly implies some sort of intelligence on the part of the power amplifier. Some sort of measuremnet or observation that the power amplifier is doing in an active speaker that it can't do in a passive speaker.

    The proof of what I'm saying is there in the circuits of the power amplifiers used in active speakers. They are just dumb circuits (in the vast majority of cases). There is no control circuitry.

    All that you have is an electrical effect due to the power amplifier being connected directly to the bass driver that results in more damping of the driver.


    The same or more damping - or CONTROL if you want to be really lazy and imprecise on your definitions - can be provided by changing the design of the speaker cabinet (by changing the volume of air in the cabinet).

    This therefore makes the supposed advantage of active speakers having better CONTROL AKA DAMPING completely null and void - when you can get just as much or more CONTROL or DAMPING by designing the speaker cabinet properly. And in so doing - avoid putting dozens of transistors in the signal path!
     
  25. stevelup

    stevelup
    Well-known Member

    Joined:
    Aug 13, 2002
    Messages:
    6,710
    Products Owned:
    1
    Products Wanted:
    1
    Trophy Points:
    136
    Location:
    Swindon
    Ratings:
    +1,221
    I am wondering if you realise just how patronising you sound?
     
    • Like Like x 1
    • Agree Agree x 1
    • List
  26. lindsayt

    lindsayt
    Active Member

    Joined:
    Feb 9, 2011
    Messages:
    299
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    28
    Ratings:
    +48
    In a passive speaker we can get the passive crossover to match the response of the speaker drivers by padding down the most efficient driver - usually the tweeter - with resistors.

    Resistors have a much more benign effect on system sound quality than active amplification devices such as transistors. And we're looking at adding 2 resistors instead of 32 to 64 transistors!

    And then we get can get into a discussion of what's more important:
    (a) flattening the frequency response curve by the odd db here and there around the crossover region.
    or
    (b) preserving the clarity and dynamics of the signal by not putting 32 to 64 additional transistors in the signal path!
     
  27. Trollslayer

    Trollslayer
    Distinguished Member

    Joined:
    Feb 11, 2007
    Messages:
    24,251
    Products Owned:
    0
    Products Wanted:
    1
    Trophy Points:
    166
    Location:
    Poole
    Ratings:
    +9,556
    There aren't 32 to 64 transistors, there are ten monolithic op amps. this is different when it comes to noise etc.
    How about looking at how the filters in that circuit interact and try to compare that with the selectivity of a passive crossover?
     
  28. Trollslayer

    Trollslayer
    Distinguished Member

    Joined:
    Feb 11, 2007
    Messages:
    24,251
    Products Owned:
    0
    Products Wanted:
    1
    Trophy Points:
    166
    Location:
    Poole
    Ratings:
    +9,556
    Oh, and add phase control plus variations in behaviour at different power levels.
    Hint - passive crossovers are more susceptible to this.
     
  29. Ugg10

    Ugg10
    Well-known Member

    Joined:
    Apr 17, 2009
    Messages:
    2,933
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    136
    Ratings:
    +1,400
    Reading through this discussion is very interesting.

    If I were to start again with a clean sheet on my “system” (because all of this need to be tempered by it being in a system and the usual adage - crap in crap out) and streamed music being my only source I would personally use, I would keep the signal in the digital domain as long as possibly (carrying out crossover, pad, notch filters etc) until I get the frequency profile and driver delays (although this can be done passively by offsetting the drivers) to where I want it to be and only then d-a the diagonal and amplify it.

    So, digital stream > crossovers/filters/delays > d-a converter > passive attenuators > power amps per driver > speaker drivers.

    Therefore, given the uprise in streaming and the emergence of cd quality/lossless streams the I suspect we may see more “digital active” speakers in the future.

    This is very possible with the use of kit like minidsp/Dirac etc. already but in multiple boxes.

    I have had (possibly mad) thoughts about a diy speaker based on the PMC OB1 (3 way transmission line) but with three amps per speaker being fed by a minidsp 2x4hd each, but that is for another thread.

    This obviously does not work with analogue inputs as there is an a-d conversion prior to the crossover etc. which I guess this just removes one problem and adds another.
     
  30. Trollslayer

    Trollslayer
    Distinguished Member

    Joined:
    Feb 11, 2007
    Messages:
    24,251
    Products Owned:
    0
    Products Wanted:
    1
    Trophy Points:
    166
    Location:
    Poole
    Ratings:
    +9,556
    It might seem picky but "increase" is better.
    As to the whole audio chain, it is worth remembering that AV receivers do a lot of work provided you carry out a room calibration and I don't know if the phase variations in an active speaker system would affect the room calibration.
     
Thread Status:
Not open for further replies.

Share This Page

Loading...
  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.
    Dismiss Notice