I don't see how you can arrive at anything less than 2nd order, due to the inherent high-pass roll off of the full range. AFAIK, the only thing you can really do is put that roll-off to use and aim for something close to a 2nd order (preferably -6dB Linkwitz-Riley IMO), and even that is often difficult/impossible due to baffle step, the driver used, and the enclosure type. When it comes to the low-pass on the woofer, I suppose one could easily just use a series 1st order (leaving excessive output above the crossover point), and this might actually be beneficial in some cases (I plan to try this out with my FF85K project). However, regardless, with the natural roll off of the full ranger (not to mention that of the woofer), I honestly don't see how anything remotely similar to actual 1st order slopes can be achieved, except for maybe the lowpass on the woofer, if its range is wide enough. Other than that, none of the benefits of 1st order crossovers can really be obtained. If I'm wrong and missing something here, someone please clue me in.
..Since I've recently acquired a pair of FF85K's, let's take something like this for example: Fast
Thinking along the lines of slopes and Qtc, let's consider that a series 1st order slope adds a Qtc of 0.1 to the inherent high-pass roll-off characteristics of the FF85K and its enclosure. A 2nd order Linkwitz-Riley slope equals 0.5 Qtc, and a 2nd order Butterworth is 0.707 Qtc.
The FF85K by itself (according to the spec sheet) is already right around 0.5 Qts. So, mounted to a flat wall in an anechoic chamber, one should already expect a frequency response resembling something at least somewhat similar to that of an LR2 high-pass with a -6dB point at around 130Hz:
http://img444.imageshack.us/img444/6505/ff85k05qtc.png
Add to this the fact that the FF85K won't be in mounted in a wall, and that a small enclosure of some sort must be used in the system, which will inevitably increase the natural high-pass slope. For this example, we'll give it a sealed enclosure of 2 liters. We now end up with a Qtc of around 0.6:
http://img607.imageshack.us/img607/9047/ff85k2l.png
Now, let's take the baffle dimensions of the Klang+Ton FAST Monitor (37 x 22.6 cm), enter them into the Edge baffle diffraction simulator, along with the 80mm (3in) FF85K, place the driver near the top center of the baffle, and we end up with this:
http://img803.imageshack.us/img803/6186/ff85kfastedge.png
..A straight slope of -2.5 dB from 400 to 200 Hz, plus another -2 dB from 200 to 100 Hz, plus another -1 dB below 100 Hz. About -5 dB between just 100 to 500 Hz, due to baffle step. Add all of that to the pre-existing 0.6 Qtc high-pass slope of the FF85K in its 2L sealed enclosure, and you're probably more likely to end up with something in between 2nd and 3rd order at best, and this is BEFORE adding any electrical high-pass network. Add a series 1st order into the mix, and you'll probably end up with something resembling at least 3rd order high-pass, likely even higher.
Again, to achieve the benefits of a 1st order series network, you need a FINAL slope of 0.1 Qtc throughout a considerable amount of bandwidth, do you not? And that goal is impossible. With the example above, if the woofer had usable response out to 20 KHz (then it wouldn't be a woofer!), you could use a series 1st order to cross at around 2KHz, and that might yield some of the advantages of first order slopes IF the drivers are angled just right and IF your head is in a specific position.
Again, if I'm wrong here, let me know.
Edit:
BTW, I'm not attempting to prove a point or anything. I come across comments like this from people I generally regard as knowledgeable on the subject, and begin questioning my thinking. I've been looking for woofers to mate with these full-rangers, and I figure Dave probably has quite a bit of experience with them. Also, I hooked up the FF85K's lastnight to a spare Adcom I had sitting around, just to get a good idea of them. The drivers were just open, but with the backs stuffed into a pillow. No tweeter needed with these for sure! If anything, they might be a little hot on the top end. And while delicate, they actually took power a little better than I'd expected.
Thinking along the lines of slopes and Qtc, let's consider that a series 1st order slope adds a Qtc of 0.1 to the inherent high-pass roll-off characteristics of the FF85K and its enclosure. A 2nd order Linkwitz-Riley slope equals 0.5 Qtc, and a 2nd order Butterworth is 0.707 Qtc.
The FF85K by itself (according to the spec sheet) is already right around 0.5 Qts. So, mounted to a flat wall in an anechoic chamber, one should already expect a frequency response resembling something at least somewhat similar to that of an LR2 high-pass with a -6dB point at around 130Hz:
http://img444.imageshack.us/img444/6505/ff85k05qtc.png
Add to this the fact that the FF85K won't be in mounted in a wall, and that a small enclosure of some sort must be used in the system, which will inevitably increase the natural high-pass slope. For this example, we'll give it a sealed enclosure of 2 liters. We now end up with a Qtc of around 0.6:
http://img607.imageshack.us/img607/9047/ff85k2l.png
Now, let's take the baffle dimensions of the Klang+Ton FAST Monitor (37 x 22.6 cm), enter them into the Edge baffle diffraction simulator, along with the 80mm (3in) FF85K, place the driver near the top center of the baffle, and we end up with this:
http://img803.imageshack.us/img803/6186/ff85kfastedge.png
..A straight slope of -2.5 dB from 400 to 200 Hz, plus another -2 dB from 200 to 100 Hz, plus another -1 dB below 100 Hz. About -5 dB between just 100 to 500 Hz, due to baffle step. Add all of that to the pre-existing 0.6 Qtc high-pass slope of the FF85K in its 2L sealed enclosure, and you're probably more likely to end up with something in between 2nd and 3rd order at best, and this is BEFORE adding any electrical high-pass network. Add a series 1st order into the mix, and you'll probably end up with something resembling at least 3rd order high-pass, likely even higher.
Again, to achieve the benefits of a 1st order series network, you need a FINAL slope of 0.1 Qtc throughout a considerable amount of bandwidth, do you not? And that goal is impossible. With the example above, if the woofer had usable response out to 20 KHz (then it wouldn't be a woofer!), you could use a series 1st order to cross at around 2KHz, and that might yield some of the advantages of first order slopes IF the drivers are angled just right and IF your head is in a specific position.
Again, if I'm wrong here, let me know.
Edit:
BTW, I'm not attempting to prove a point or anything. I come across comments like this from people I generally regard as knowledgeable on the subject, and begin questioning my thinking. I've been looking for woofers to mate with these full-rangers, and I figure Dave probably has quite a bit of experience with them. Also, I hooked up the FF85K's lastnight to a spare Adcom I had sitting around, just to get a good idea of them. The drivers were just open, but with the backs stuffed into a pillow. No tweeter needed with these for sure! If anything, they might be a little hot on the top end. And while delicate, they actually took power a little better than I'd expected.
Last edited:
In the last effort FR rolloff was about 100 Hz (FF85KeN), XO was > an octave & a half higher than that. Bafflestep does not, i believe affect the phase response, just the FR and power response.
dave
Exciting speaker,if thinner and add one subwoofer , very good 2.1 sysytem for
home audio.
we have made some type of 2.1 digital processing amp, we are looking for good speaker tie-in sale.we mean good curve and good sound.
Thinner speaker box will better for flat panel TV,of course good subwoofer necessary.
we believe it is bigger market in China, now almost all family have flat panel TV even in countryside in China.
I'm struggling with the transition from full range to woofer. The woofers I am considering play pretty flat and pretty high, but the full rangers (all 4" and 5") are a problem on the low end. In a sealed box, none of them have an f3 of lower than about 100Hz. If I raise my crossover point from 300Hz to 400Hz, this would give me a two-octave crossover transition, but then it would raise two other questions:
1. Is it likely I would notice the intrusion into the critical 300Hz-3500Hz band?
2. The FR models show a rolloff past f3 of not quite 12dB per octave. Would that change the first-order characteristics of the full ranger low end to a noticeable degree?
1. Is it likely I would notice the intrusion into the critical 300Hz-3500Hz band?
2. The FR models show a rolloff past f3 of not quite 12dB per octave. Would that change the first-order characteristics of the full ranger low end to a noticeable degree?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.