Not sure if anyone is still paying attention here but I was just reading a review of the PMC OB1 loudspeaker in which the reviewer states that the crossover is wonderfully designed and produced a plot of phase coherence that is the best he has ever seen. According to the review, the crossover maintains the drivers in perfect phase with each other except for a small abberation at one point in the frequency spectrum. Given all that has been stated above by the first order crowd, how can this be given the use of 24db fourth order slopes? I'm sure I'm missing something here as this discussion lost me a long time ago. I'm just wondering how a steep slope design like the one in the OB1 can produce such perfect phase relationships as measured in the review.
First Order Crossovers: Pros and Cons
I wonder if some folks might share their expertise on the question of crossover design. I'm coming around to the view that this is perhaps the most significant element of speaker design yet I really know very little about it and don't really understand the basic principles. Several of the speakers I have heard in my quest for full range floorstanders are "first order" designs. I have really enjoyed their sound but do not know if this is attributable primarily to the crossover design or to a combination of other factors as well. In addition, I have heard that, for example, because of the use of this crossover configuration on the Vandersteen 5 one has to sit at least 10 feet away from the speakers in order for the drivers to properly mesh. Is this really true and if so why? Another brand also in contention is the Fried Studio 7 which also uses a first order design. Same issue? Could someone share in laymans terms the basic principles of crossover design and indicate the advantages and disadvantages of each. Also, what designers are making intelligent choices in trying to work around the problems associated with crossover design? Thanks for your input.
116 responses Add your response
Dodgealum...There is no magic! No matter who designs the crossover each section, which is worth 6 dB, causes a 90 degree phase difference between Hi and Lo. Both Hi and Lo phase vary with frequency, but the difference remains 90 degrees. Of course, as you go away from the filter break frequency (crossover frequency) the amplitude of the Hi and Lo signals diminish so that phase is not a big issue. So, a 12dB crossover has a phase difference of 180 degrees, and if you hook the tweeter up with reversed phase the audio output will be back in phase. With the 24dB crossover that you mention the phase difference is 360 degrees, which is the same as zero, so the drivers, both connected with normal polarity, will be back in phase. So why do people like 6dB crossovers? The audio output of the drivers may be phase shifted from the electrical signal, and, at a selected frequency, can also be adjusted by positioning of the tweeter with respect to the woofer. So, with tweeks, a 6dB crossover can avoid the 90 degree phase shift, at least at the crossover frequency which is where it matters most. FWIW, 6dB crossovers also require the minimum of crossover components. A passive 24dB crossover is a tough nut to crack, and therefore rare, but is easily implemented in electronic crossovers, where it is very common. |
A 2nd-order x-over with the tweeter's leads reversed gives you an in-phase output if you're considering _absolute_ phase (polarity). The 4th-order x-over is similar. However, you're still seeing a phase shift. If you look at the electrical input and output with something other than a simple sine wave, it becomes obvious that the output is different from the input. The classic statement I keep repeating is "a 1st-order x-over sums to a piece of wire." This is not true of other standard x-overs. It's possible to add a delay in an active x-over to put things closer to normal, and this is even easier when the x-over is digital. Technologies like DEQX hold a lot of promise. It's like having your cake and eating it too. :) |
Hi Eldartford. Thank you for your comments. Just a quick note, however. A second-order filter is not two first-order crossovers in series, nor "cascaded". The parts in a second-order filter, the capacitors and inductors, are of different values than those used in a first-order circuit. Best regards, Roy Johnson Founder and Designer Green Mountain Audio |
greenmountainaudio@adelphia.net...Yes the component values, for a particular frequency, differ from a first order filter because the cascaded filters are working from and into different impedances. But they are still two series filters. In the case of an electronic crossover, fourth order L/R for example, each of the four filter sections includes a buffer amp, so that each stage is similarly driven and loaded, and in this case the component values are the same. |
Ok, not to be over the top, but if speakers are time coherent and each cone, if I understand correctly, is 180 degrees out of phase from the other, wouldn't a perfect set-up only work with bi-wiring, where you can flip one of the sets of wires for the cone you want to put back into the same phase? Or am I wrong? |
In honor of the impending release of a diy version of the Nelson Pass designed First Watt active crossover, I thought I’d bump this amazingly informative thread on crossover design... looking forward to building the FW crossover and using it to facilitate bi-amping of my Coincident PREs. https://www.diyaudio.com/forums/pass-labs/165756-pass-labs-b4-crossover-questions-58.html |
@murphythecat That is exactly what my Coincidents due with the midrange and tweeter. Midrange is connected directly to amp and the tweeter just has a single 4.7uF cap in front of it. But then they've got an additional two 12" woofers after an inductor on each channel... I'll be using Nelson's crossover to skip the inductor and play a bit with the crossover frequency (they don't fall away until 2000hz and I'm getting some resonance up there with my current sub amps). |
Transducers are not flat though there operating range they roll off the 1st order utilizes the reduced levels and since it overlaps do to its slight -6db per octave roll-off you can fill in these dips in response by allowing other driver to gently roll off overlapping and filling in. A 1st order is not for every loudspeaker type and I would say 1st orders require more costly drivers without peaks and severe breakup modes as well as smooth gentle roll-off. I find horns can take advantage of 1st orders since the horn itself has a fairly severe drop off at its operating range you can use a 1st order without worrying about to much overlap. A 1st order can reduce power handling and peak SPL but its advantages in-phase and lower parts count can have sonic benefits. |
Absolutely wonderful read! Glad this thread was bumped recently back from 2005 or I would never have found it. I am an electrical engineer myself with significant signal processing background, and I truly appreciate the clarity of Roy’s and Karls explanations here. They are absolutely correct. At the same time, it is easy to see how without formal study of electrical signal mathematics it might be difficult to follow or even counterintuitive. Nowhere is this as obvious as the discussions relating to Phase coherency vs Time coherency (i.e. simultaneity). Time coherency is clearly the stricter condition as Roy points out. I also really liked Karls explanation of the sum of the high and low frequency drivers output after first order crossover as the sum of two phasors separated by 90 degrees at all frequencies. I think I would have further reinforced the fact that it is *perfectly fine* to add these "out of phase" signals because the result is always a constant phase at the average of the two individual phases, which is actually ideal. Without some background in signal processing it might be easy to think these two drivers were fighting each other and causing distortion. Karls post really helped clear this up. Finally, it was good to see it pointed out correctly that the spatial lobing problem is a multiple emitter issue, and not a crossover issue. Steep crossovers merely narrow the region over which multiple emitters "exist" for that frequency. Thanks for the great read. I hope there is more discussion on this. I have a pair of speakers with a first order crossover now myself (Dynaudio Special 40), and they really do have a very special sound. Yes, I hear the lobing, but yes, I also hear the outstanding clarity which I assume is the phase and/or time coherency of the drivers in the sweet spot. Still trying to decide what I think of them overall, but so far very impressed. Btw, there are a lot of other LC components on the crossover boards (presumably used for impedance correction of the drivers?). Would like to learn more. |
I have heard that, for example, because of the use of this crossover configuration on the Vandersteen 5 one has to sit at least 10 feet away from the speakers in order for the drivers to properly mesh. Any multi-way, non-coaxial speaker needs a minimum distance to be heard as intended. The differences in arrival times as well as the angle between the drivers themselves is reduced with distance. Imagine for instance if you are 1' from a 2-way speaker. You can probably hear the tweeter and woofer separately. I am not sure why Vandersteen recommends what they do, but I could see this with a variety of filter designs. The other thing for people to keep in mind with drivers and crossovers is that the slope and angle is a function of the driver behavior AND the crossover components. This is true whether you use an active or passive crossover. Best, E |
| Post removed |