Why?
Granted I don't use them (I'm a dipole-and-omni guy at home) and I've heard a lot of bad ones, but I've also heard well implemented examples that sounded admirably "clean". They do present an example of where amplifier "matching" is important . . . they beg for ClassA or ClassD or a very carefully designed AB with near perfect bias (ie. amps with no "crossover" distortion), and with either modest excess power (to keep noise down) or an exceptionally good SNR. Compression drivers are certainly a good tool for demonstrating the "first watt" concept . . .
Without the passive xover the Tannoy woofer is 95dB/1W while the compression driver is closer to 107dB.
The eq needed is the usual 6dB/oct boost above about 5kHz and a mild notch between 2.5 and 3kHz. I use a parametric equalizer between my analogue crossovers and the amp.
In their original, passive state Tannoy claimed THD of <0.5% at 96dBSPL from 100Hz up to I don't remember what but I think it was 12kHz.
The xovers I use have a S/N ratio of about 90dB or a bit better unweighted.
I have encountered no noise problems in this set up.
My treble amp provides 175W into 8 Ohm. I think the Tannoy tweeter might be a 16 Ohm job but I never bothered measuring it, I only measured the overall FR and needed to knock 12dB off the treble amps gain.
Last edited:
The rub is that you cannot "do either to achieve it". It is easy to achieve near perfect crossovers and filters in the "active" realm that are almost impossible to accomplish in the passive because of the varying load presented by a driver. And that's without even addressing the additional options available with digital filters . . .
In these days of CAD, you can hit just about any desired transfer function with a passive. It's not 1978 anymore.
+1
The advantage of active, if there is an advantage rather than a trade off, isnt transfer function 'ease of use'. Not unless you're trying to implement 6th order or higher and/or multiple notches. As has been mentioned earlier, i think, an active design could benefit from a passive impedance conjugate. (the thread is a tad long now)
Last edited:
In theory anyway. It would be a bitch to get the active XO to track the changes in a drivers response under dynamic changes in the music or a different setting of the loud control... you can make a passive XO that does this inherently.
dave
Isn't this a bad thing? Isn't this what we're trying to avoid?
SY writes:
The problem is that in an LC or LRC passive crossover there is also another L and R in the system that is always fluctuating with the signal: the speaker drivers themselves. Not to mention the back EMF too...
So a passive XO can have the same tranfer function if it is DISCONNECTED from the driver; when it is connected to an actual drivers, things don't stay the same.
In other words, since the passive XO is directly affected by the driver's varying L and R, plus EMF, then it wil **never** be the same as using an active crossover for achieving the same XO cutoff frequency and slope.
Pardon me if i'm ignorant of something, but i thought we wanted to elliminate the problems brought by the varying load presented by the drivers.
This perhaps was true before the 1980s. Modern horns not only are free from higher order distortions and pretty controlled in 2nd and 3rd harmonic distortions, but more importantly the big efficiency increase allows the driver itself to work less and thus achieve lower distortion.
Furthermore high efficiency (>100dB) allows the use of the purest, most natural sounding amps available. Which can elimminate the nastiest non-linear distortions.
There is a reason many of the best high-loudness studio monitors in the world are using horns on their drivers.
Not all horns are created equal...
flavio, a few minutes spent with a CAD program will usefully demonstrate that poor quality drivers whose responses change (that's an inevitable consequence of parameter changes) fare poorly with either active or passive. The impedance doesn't change without frequency response, sensitivity, and distortion changing as well.
Yes, with programs like PCD you can cobble up a 10 or 15 or even 20 element "passive" crossover that will "hit [the] desired transfer function", at least close enough for audio work and without presenting too silly a load to the amplifier (and wasting too much power). But only an idiot (and a rich one at that) would build it.In these days of CAD, you can hit just about any desired transfer function with a passive. It's not 1978 anymore.
I'm not arguing that you can't build servicable two-way surrounds with passive crossovers to use with your AV receiver's built in amps. That's what I've done for mine. It takes fewer wires than an amplifier in every box would, too. But you won't find Linkwitz transforms and dipole correction and inter-driver delays quite so simple, regardless your "CAD" skills . . . and you won't find the 8.3 mH inductor that the program calls for all that easy to adjust if it turns out that 9.1 actually works better when you build it.
It's kind of quaint, though, using "modern" CAD software to design using 1978 (obsolete) technology for the build. If you're going to design it on a computer then why not implement it on a computer (DSP chip) as well?
Yes, it is a bitch trying to correct one nonlinear distortion with another (hopefully inverse) nonlinear distortion. I sure souldn't count on any simple "passive" device doing it . . .
Yes . . . at least that's generally what I want to do. And the tool for doing it is an amplifier, a device which has a constant input impedance and, if properly matched, an output that is immune to the driver's impedance curve.
Pano agrees.
DSP is cool because you can do all sorts of corrections, large or small. A poke here, a prod there and you can really whip things into shape. With passive it's not so easy, but there are some clever things you can do. I work and work at finding a crossover that does what I want with the fewest components. For example my current crossover for the VOTT A5 + tweeter has 12 parts. 3 for the low pass, 5 for the midrange, 4 for the tweeter. 5 inductors, 3 caps, 4 resistors total. Not bad for a three-way. It could be done with fewer, but would not sound right (I've tried; 1st order crossovers are not for me).
Very steep filters, which DSP is good at, fail to convince me sonically. Yes, they keep the drivers perfectly within their ideal bandwidth, but they sound segmented to me. Maybe my ear needs more of a blend between the drivers - a psychoacoustic thing. Not everyone may object to them.
high loudness studio monitors = deaf engineer = poor mix.
Please show me something better than results in High Frequency Compression Driver Evaluation thread:
These are post 1980 drivers. 4%-10% 2nd harmonic is not natural or reasonable for high fidelity, and whatever a "pure amplifier" is, it's not going to fix this.
What is mechanism for "pure" amplifiers elimination of driver distortion?
Regards,
Andrew
Far too simplified situation. I don't want to sound harsh but work in studio is divided into phase and each phase as differents needs relative to monitoring and you don't seem to understand this requirements.
In recording phase you need a playback system able to reproduce 'real' dynamic of the instrument recorded, and in case of drum in big control room highly dead (RT60 <0,3seconds with sweet spot 3/4 meters distance from loudspeakers ) high efficiency is the only way imho. Recording through small monitors (NS10 or Genelec or Adam or...you name it) give numerous artifacts and big problems during mixing with as a result = poor mix, because sound of album is made during takes, not really at mixing stage (because crap in=crap out).
For mixing situation is different as the action is relative to levels and equilibrium between instruments or group of instruments, fx, etc,etc... and small nearfield monitors well known by enginneers can give great results at moderate reproduction levels. But in this case big mains monitors are used to check work regularly ( not for high SPL but mainly for bandwith and soundstage) same rooms used as for recording with same acoustical characteristics.
For mastering you need systems as transparent as possible and without any compression artifacts on any peaks, with ability to discern very subtle corrections (eq or compressions) in great acoustics (not as dead as recording studios, more live and closer to 'livingroom' acoustics) . No nearfield allowed, mainly midfield reproduction and only high quality systems used (Dunlavy, B&W, Pmc, Atc, Klinger-Favre, Tannoy, Quested, Kinoshita/Rey Audio,...) with moderate levels of approximately 85db spl (with 20db peaks allowed: at max 105db spl peak at sweet spot, search for K-system on google).
Many infos about recording industry process and habits can be founds in Philipp Newell's 'recording studio acoustics' book on Focal Press with an impressive chapter dedicated on loudspeakers technology and choice relative to end use.
Apologize for the Off Topic.
Last edited:
I wonder how many dome tweeters stay below those THD levels when you feed them the 100-200W necessary to reach those SPL levels.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.