Bi-Amping with Active Crossovers - Pros and Cons

Discussion in 'Audio Hardware' started by Soundgarden, Jan 9, 2022.

  1. harby

    harby Forum Resident

    Location:
    Portland, OR, USA
    Fortunately, the current-blocking impedance of an inductor increases much much more rapidly, as with the 320Hz crossover lowpass inductor I used as an example.

    Reactive components don't dissipate power via device inefficiencies when current can't flow through them...

    [​IMG]

    Inductive impedance is why you can plug an unloaded transformer into the wall and it uses 1 watt of power instead of blowing up with fire.
     
  2. Just Walking

    Just Walking Forum Resident

    Location:
    UK
    No - it is nothing to do with AC inductive impedance. Proximity effect is the tendency of the magnetic field in an inductor to force the current to flow in narrow channels within the wire at surprisingly low frequency. Found out about that when developing a supermarket security gate (in the mid 80's), class D amp and LC filter on the output. Decided to use loudspeaker inductors. Frequency was 3.3kHz. Turned it on and wondered where the smell of burning was coming from. WFT? DCR was low enough to dissipate only a Watt or so. The plastic former melted like a Salvadore Dali painting. At around 3kHz proximity effect had increased the dissipative resistance by over a factor of ten as compared to the DCR.

    There is even a Wikipedia page about it Proximity effect (electromagnetism) - Wikipedia

    There are easy ways of minimizing this by correct design of the inductor.
     
    Last edited: Jan 30, 2022
    Randoms likes this.
  3. SandAndGlass

    SandAndGlass Twilight Forum Resident

    Thank you!
     
    Randoms and Soundgarden like this.
  4. Soundgarden

    Soundgarden Forum Resident Thread Starter

    Location:
    Bend, Oregon
    Inquiring minds and all…
     
  5. Just Walking

    Just Walking Forum Resident

    Location:
    UK
    Make it out of several strands. No need for litz. All you do is keep the copper area the same, but make it out of 8 (or 10 or whatever number you like) thinner strands.

    I ended up with two screwdrivers stuck in the dirt about 10 yards apart, walking up and down with a spool of wire. Then a battery drill to lightly twist them (just to hold them in a bunch for winding). Made up the same inductance - problem solved. The security gate went into volume production - it was kind of nice to see my design in airport shops worldwide as well as the local DIY store.

    But you can buy crossover inductors that are made in that way now, so you don't have to do the hard work of making your own. They mistakenly sell them as Litz wire, which is not correct, and that it reduces skin effect, which is not the dominant mechanism. It does however massively reduce proximity effect.
     
    Randoms likes this.
  6. SandAndGlass

    SandAndGlass Twilight Forum Resident

    The HF horn is more efficient than the bass horn on the A7. This makes the HF horn louder than the bass speaker.

    Keep in mind that the A7 is a theater speaker that is a scaled down design from their larger theater speakers. They are designed to specifically project sound out into an auditorium to an audience that is fairly distant from the speaker itself.

    In this type of setting, air naturally absorbs sound energy in the higher frequencies. By the time the sound reaches you 50 to 75 feet out into the audience, the HF sound from the horns has been naturally attenuated by the air and now the level of the HF is equal to the bass and you have a balanced sound.

    If you bring that same speaker indoors and attempt to use it in a smaller space where you are sitting 10 to 15 feet away, the horns will be much brighter, because they are louder than the bass.. this obviously necessitates lowering the volume of the horn to match the level of the bass driver.

    I had my first pair of A7's built for me fifty years ago. This is what the Altec crossovers look like for the A7's.

    [​IMG]

    They were available in either 800-Hz. or 500-Hz. versions. As you notice, there is a volume control type knob to adjust the amount of HF tweeter attenuation. You just turn down from zero attenuation to the level that best matches the level of the bass output. Sounds simple?

    As I discovered back at the ripe age of sixteen, that when you decreased the level of the horn output, the horn not only decreased in volume but it decreased in resolution in addition. As you turned down the volume, the horns sound became duller and duller. Such are the problems that are introduced by using a variable resistor into the circuit.

    Schematic of 800-Hz. Altec crossover taken from the web.

    [​IMG]

    I will point out that Altec was apparently aware of this issue.

    The models that are used for playback in recording studios, accomplish this by using a less powerful magnet in the driver, thereby reducing the magnetic field on the voice coil and therefore making the HF horn not as loud. Photo is from an older Altec catalogue.

    [​IMG]

    This is why I wouldn't recommend Altec crossovers for home use. They do separate the high frequencies from the low frequencies, that in itself was not the primary issue.

    Funny, as I was looking for a photo of those Altec crossovers, I came upon this photo of my own crossovers in Google images.

    This is an upgraded crossover from ALK engineering, which is designed for a vintage Klipsch La Scala speaker and crossover at 500-Hz.

    [​IMG]

    As you can see, there is no typical "volume" control type adjustment to "shelve" the volume of the HF horns. By adjusting the jumpers on the autoformer (the transformer looking thing on the left) you will be able to attenuate the power that is routed to the 511B horn in your A7-500W's. After doing this, you have better and more balanced bass.

    By clicking on the photo in Google, it took me to the following page in our forum where I have explained this before. Check it out at the link below.

    Altec owner's upgrades, mods, short-cuts, etc.

    You might go back and reference specific posts I have made on my Altec’s and crossovers, in this thread. You might go back and review them on page 7 (A7's modified), page 13 (ALK crossovers, page 18 (HF shelving control).

    On the page I had linked to, I made mention of a site Jeff's Place, where Jeff had an Altec restoration project, including capacitors. Here is a schematic I found.

    [​IMG]

    You might check out Jeff's Place. I have provided a link on the page I linked above.
     
    Ampexed, Tim 2 and Soundgarden like this.
  7. Soundgarden

    Soundgarden Forum Resident Thread Starter

    Location:
    Bend, Oregon
    Well it is simple right? Adjustable volume knobs are the worst :) This is honestly something I've been wondering about with pre-amps, but that's another tangent altogether. Better is to have a switch with several discrete settings to choose from. Better yet is to have single value components once you know how you want things "dialed in." I'm assuming the auto-former allows that but I don't actually know how those work.
     
    Last edited: Jan 31, 2022
  8. Chris_A

    Chris_A Forum Resident

    Location:
    Arlington, Texas
    If I'm not mistaken, I believe that I'm that guy. I thought I'd dust off my account here to offer my 2¢...

    Try this: Active Bi-Amping/Tri-Amping FAQ

    You are generally correct here.

    Generally, you need to measure the acoustic output of the loudspeaker to get any reasonable hi-fi performance, i.e., ±3 dB SPL across the audible frequency band that your loudspeakers can handle. Here are some threads on that subject:

    Xilica XP active crossover integration into my system
    Using REW to Find Parametric Equalizer (PEQ) settings
    Using REW to Determine Time Delays Between Drivers

    I don't recommend higher order passive crossovers. They are subject to short term heating and aging drift in performance, as well as adding gobs of acoustic phase growth from higher frequency to lower frequency. If they are just a little bit off, they don't work very well. This is not true for DSP crossovers--using REW and a calibrated microphone (about $100) to dial everything in.

    Chris
     
    Randoms, Jim Hodgson and Soundgarden like this.
  9. Soundgarden

    Soundgarden Forum Resident Thread Starter

    Location:
    Bend, Oregon
    Thanks s much for the links and the feedback Chris. I'm obviously in research mode and likely will be for a while before pulling the trigger on anything. I hadn't seen these links before and will read through them after work this evening.
     
  10. SandAndGlass

    SandAndGlass Twilight Forum Resident

    Transformers vs. Autoformers:

    Transformers can electrically isolate one device from another to prevent hum noise due to grounding problems, and to prevent shock hazards. They can also be used to create balanced audio lines, which help to reduce noise caused by electromagnetic interference.

    Transformer basics
    A conventional transformer consists of two coils of wire wound around an iron core. The two coils are known as the “primary” (input) and “secondary” (output). Electrical current passing through the primary coil creates a magnetic field that induces an electrical current in the secondary coil. This allows two circuits to interact with each other without being physically coupled to one another. In electronics, this is referred to as galvanic isolation.

    Additionally, depending on the size and configuration of the primary and secondary coils, the voltage and current induced in the secondary coil can be different than the primary coil, allowing the voltage and current to be "transformed" to different values, stepping up or stepping down voltages.

    [​IMG]

    Autoformers
    An autoformer (also known as an autotransformer), performs some of the same functions as a conventional transformer, but there are several differences in design and performance. In particular, an autoformer has only one coil which performs the functions of both the primary and secondary windings.

    An Autoformer schematic

    [​IMG]
    Since autoformers don't have two electrically isolated coils, they don't provide electrical isolation like typical transformers do. Instead, autoformers allow for the same voltage/current transformations that typical transformers are used for.

    [The prefix "auto-" in autotransformer doesn't imply that anything "automatic" is happening. Instead, it comes from the Greek root for "auto", meaning "self", referring to the fact that an autoformer is a single coil that works with itself rather than a pair of coils working together.]

    In the photo of an ALK crossover, the transformer looking object on the left is actually an autoformer.

    [​IMG]

    If you take a closer look, you will see the tabs that accommodate push on jumpers (connectors). By connecting the jumpers to different pairs of tabs on the autoformer, you set the amount of HF attenuation. This is accomplished purely by varying the amount of inductance. There is no variable resistance in the circuit (only the low basic fixed DC resistance that is attributed to the wire itself).

    This is what ALK Engineering is saying in their description of having low DC resistance in the autoformer itself. It does not mean this is the load resistance that is being placed on the amplifier. In fact, another feature of ALK crossovers is that despite a changing impedance of the speaker drivers as the frequencies change, his crossovers provide isolation of these changes and the amplifier always sees a constant impedance at the speaker terminals.
    • [​IMG]
      Enlarge
      A specially designed autotransformer having low DC resistance (0.382 Ohms) and tight coupling provides taps to allow squawker level setting changes by the customer. The choices will be 1 to 18 dB in 1 dB steps, Attenuation is selected without soldering by push-on terminals and without cryptic setting tables. The attenuation settings are clearly written by each tap in dB units. The normal setting will be 6 dB. This adjustment will also allow the use of 8 or 16 Ohm squawker drivers without modification. NOTE: Any transformer that is larger in size is a waste of money. It will handle 50 Watts of power without batting and eyelash with virtually no loss, That is already extravagant considering the power levels through the midrange in a Klipsch loudspeaker will generally be less than ONE Watt!

    Advantages
    Because an autoformer requires fewer windings and has a smaller core, it is typically lighter and less costly than a conventional transformer with the same ratings. Autoformers have performance advantages too. They have increased power handling capability, flatter frequency response, lower insertion loss and lower distortion than conventional transformers of similar size and cost.

    Actually I like precise easy to adjust analog volume controls and keep these controls close by, at my fingertips.

    The main central point of control of my system, is provided by the Peachtree iNova. It is used as a source selector, ESS Saber DAC and SS preamp. All power amps are outboard.

    [​IMG]

    The volume control of the iNova (bottom) provides the master preamp level for the systems three power amplifiers. The Peachtree musicBox (above) allows for the relative adjustment of the rear towers. The black circular analog volume control is connected between the tube preamp and tube power amp that controls the volume level of the Altec A7’s. There are three amp / speaker pairs in the listening room.
     
    Soundgarden likes this.
  11. OhioGraySky

    OhioGraySky Well-Known Member

    Location:
    Newton Falls, Ohio
    Suggestion...
    Download Audacity and rip a cd track. Play around with the limiting tools to get an idea what in increasing levels of compression sounds like.
    I would agree that drivers are the main culprit. Core saturation of the woofer inductor can also be a problem.
     
    Soundgarden likes this.
  12. Soundgarden

    Soundgarden Forum Resident Thread Starter

    Location:
    Bend, Oregon
    So how does thermal compression affect the signal as compared with studio introduced compression?
     
  13. OhioGraySky

    OhioGraySky Well-Known Member

    Location:
    Newton Falls, Ohio
    I don't know. I would guess it would be applied more evenly vs. a driver in a certain frequency range. It sounds similar to me. Everything gets louder and shouty.
     
    Soundgarden likes this.
  14. Just Walking

    Just Walking Forum Resident

    Location:
    UK
    Thermal compression is a complicated issue.

    Normal average listening levels are quite low - one or two Watts, and usually less that 10W even in loud listening conditions. That is why SET amplifiers that supply less than 10W still sound satisfying, with reportedly excellent SQ.

    For regular transistor or large tubed amps of 100W and upwards, power and current reserves are used for transients, and supplying current to a loudspeaker with an impedance characteristic typical of a passively crossed over device.

    So under those conditions, the temperature of the voice coils are somewhat limited and thermal compression is not a major issue.

    However if you listen at an average of, say, 100W, and have massive power reserves for transients, then compression might be an issue - but that means that you average listening level, for a typical 83dB/W sensitivity loudspeaker, is 103dB - which is loud, very loud. Few domestic loudpeakers can do that without rather high levels of harmonic distortion, and thermal compression becomes more of an issue.

    Again, under loud listening conditions, thermal compression has a lag effect. When loud music starts, the voice coil is cold and there is no compression. Then over some seconds it heats up and thermal compression builds up. Then when the music goes quiet, compression is still there for some seconds as the voice coils cool down.

    For reference copper changes resistance by 0.4% per degree C. So if the voice coil temperature gets to 100C, half the maximum of (usually) 200C, the resistance increases by 40%. That sounds like a great deal; however since a loudspeaker is a current transducer, the compression can be written as 20 log (1/R) = 20 log (1/1.4) = -3dB. And that will be barely perceptible if you are rocking with high average power going into the speaker and listening at well over 100dB.
     
    Randoms, SandAndGlass and Soundgarden like this.
  15. SandAndGlass

    SandAndGlass Twilight Forum Resident

    Excellent points here...

    While commercial speakers are designed to both take a lot of power and dissipate a lot of heat energy, most home speakers are not.

    Not a real world example but one that illustrates the point, would be a speaker that is run with 1,000-Watts continuous RMS. Such a typical speaker would convert about 10% of the electrical energy into acoustic energy with the remaining 900-Watts being converted to heat energy.

    This would be similar to having ten 100-Watt incandescent light bulbs in a small enclosure. They would be giving off 900-Watts of heat. Which is quite a bit of energy to dissipate.

    A speaker would have to dissipate this energy from a concentrated place, the voice coil. This is something that is very difficult to accomplish with typical home speaker designs.

    Pared with high sensitivity speakers, that 10-Watt SET amp can more efficiently drive those speakers to fairly high volume levels, without creating high levels of excess heat.

    A much lower sensitivity speaker would require far greater power to achieve these volume levels and would therefore create more heat in the process.

    With the technology we have today, amplifiers can deliver a lot of power. Still the remaining factors of excess heat dissipation and excess mechanical vibration remain the major limiting factors for speakers.
     
    Just Walking and Soundgarden like this.
  16. Just Walking

    Just Walking Forum Resident

    Location:
    UK
    Good points. Modern domestic sealed box or vented acoustic suspension loudspeakers will typically operate at 1% efficiency. That as you say can increase to several percent with higher efficiency paper cone speakers, which might pair with SET amps.

    That leaves aside super high efficiency horns, with can get to 20-40% efficiency. But for low frequencies tend to be impressively big and need low frequency EQ PassDiy .

    Even if large horns are used, reducing the cone excursion and power levels of the drivers (so getting around thermal compression), you still aren't out of the woods. At high listening levels you run into air non-linearity in the throat region of the horn.

    There is alas no free lunch with loudspeaker transducers!
     
    Randoms and SandAndGlass like this.
  17. Just Walking

    Just Walking Forum Resident

    Location:
    UK
    It is only Nelson Pass who can come out with statements like "33 years ago I decided to build a big horn system. Constructed in the dorm library over Christmas break, the result was dubbed The Claw, a straight exponential horn 9 feet long with a 42 Hz expansion curve and a 50 sq ft mouth. We mounted a JBL LE15A woofer in the throat and used a JBL 375 compression driver and horn for about 500 Hz on up. No matter where we pointed it, the cops showed up every time."

    And from the El Pipo transmission lines "Funny things happen when your speakers are flat to 13 Hz. You have to be careful about your tone arm, your windows, your neighbors, and your bowels"
     
    Soundgarden and SandAndGlass like this.
  18. Soundgarden

    Soundgarden Forum Resident Thread Starter

    Location:
    Bend, Oregon
    First, I'll just say it's good to have a sense of what thermal/power compression actually is - both in regards to what causes it and what it sounds like.

    Second, it seems pretty clear that the driver is of greater concern than the crossover. (Assuming crossover is working properly and in-spec.)

    Third, it also seems clear that thermal compression (again with property functioning equipment) is really only a potential issue at very high volumes*.

    All that being said, it probably doesn't make much sense to spend a bunch of time and effort on reducing thermal compression from passive crossovers as a singular goal. Going to a better passive or to active may reduce thermal compression - although perhaps an inaudible improvement - but this alone is not a good reason to invest in either. But it does seem fair to include it in lists of potential benefits associated with improved passives or going active bi-amp.

    * I still wonder about bass. I don't listen to "loud music" anymore. Tinnitus. But I still love to feel bass clarity and rumble. And it takes a lot of power to drive bass effectively. So is it possible that thermal compression could still be an issue as concerns deep, strong bass, and the woofer?
     
  19. Just Walking

    Just Walking Forum Resident

    Location:
    UK
    That is why motional feedback was a big thing at one stage. You have a position sensor of woofer or subwoofer cone, so you know where the cone is. You know where the cone ought to be from the signal. The electronics then puts it there. It needs a stiff cone - but that is often the case for low frequency drivers.

    Distortion goes down by orders of magnitude, compression becomes a non-issue, and you get either a flat or tailored low frequency response to get around phase shift problems.

    This technology was introduced by Philips in 1970 (the designer of these products has just died aged 94). However the technology is alive and well still LS1s-dmf - Grimm Audio
     
    SandAndGlass and Soundgarden like this.
  20. harby

    harby Forum Resident

    Location:
    Portland, OR, USA
    One consideration for the sake of this thread's topic is not the effect of slow resistance increases in the voice coil from heating, and the resultant loss of gain, perhaps dominating in one driver. It is instead the new resistance's effect of shifting of the crossover frequencies of speaker-level passive filters.

    The lowpass filter of 4mH inductor against rising speaker resistance:
    [​IMG]
    (this does not show the gain reduction, as I don't have power dissipation analysis)

    For pondering the actual thermal levels of spouse-angering volumes:

    Thermal simulation of Loudspeakers - Chapman (pdf)

    Loudspeaker voice-coil temperature estimation - Ronny Andersson (pdf)

    A New Thermal Model for Loudspeakers. FABIO BLASIZZO (AES paywall)
     
    Soundgarden likes this.
  21. Soundgarden

    Soundgarden Forum Resident Thread Starter

    Location:
    Bend, Oregon
    Well THOSE are amazing. I'd never heard oof them before or the technology behind them. Too bad I don't have $13k to spare for the pair.
     
  22. SandAndGlass

    SandAndGlass Twilight Forum Resident

    A friend of mine back around that time had a stereo console in his families home that had a motional feedback system incorporated into the speakers. That was the only stereo I came across with a motional feedback system.

    Many if not most pro-sound systems today are processor based, where the processor maintains the speakers within spec.
     
  23. SandAndGlass

    SandAndGlass Twilight Forum Resident

    It took me a couple if weeks of listening to get the ALK crossovers dialed in correctly, through listening.

    When I got very close, I found one jumper settings were still a tiny bit too bright and the next set of jumper settings were slightly not bright enough.

    The degree of HF attenuation was achieved by changing a pair of jumpers to different taps on the autoformer. ALK provides a reference sheet that shows these different settings.

    Looking at the settings sheet, I saw that there was a pattern to these settings. I thought to myself, if I place one jumper here and the second jumper there, I might end up with something that was directly between the two published settings.

    BINGO! My guess proved to be correct and I ended up with the perfect setting.

    Your suggestion is one that had occurred to me, after taking a lot of time to manually try these jumper settings.

    Some source material may still sound a bit brite, while others a bit dull. I thought about using a rotary gang selector switch.

    [​IMG]
    These allow for multiple contact connections to be made with different positions of the rotary selector. This way, the center position could have the jumpers wired as I have them connected.

    You could then have "jumper" settings wired for a couple of attenuation settings plus and minus the correct settings, for easy HF attenuation adjustment.

    I was speaking to Al from ALK Engineering about this and he told me that none of his customers had mentioned this before.

    It was actually a project that I intended to undertake but never came to fruition...

    Zzz
     
    Ampexed likes this.
  24. Ampexed

    Ampexed Forum Resident

    Location:
    Santa Fe, NM
    FWIW I would never recommend the passive crossover route because it is wasteful, inefficient and inflexible. Active is the best way to go, especially with the Altec VOTT speakers mentioned here. There is more information about VOTT crossover/voicing on my profile page. It is important to note that these speakers need to be voiced, and the Altec crossovers, such as they are, did include some degree of HF voicing, but not nearly enough. Many people add supertweeter horns to the VOTT speakers but this was/is a band aid to get proper high frequency balance - the best way is to simply voice them correctly, then no supertweeter is necessary. It is possible to do optimal voicing with a passive crossover, but it is preferrable to do it actively.
     
  25. SandAndGlass

    SandAndGlass Twilight Forum Resident

    Not everyone in the Altec universe shares your views...
     

Share This Page

molar-endocrine