hi,
Often its said that 1st order crossovers do not provide sufficient protection especially for the tweeter. I have read that signals down to 100hz get through at almost full power. That sounds pretty tough for a tweeter to cope with. How do you make a design like that work? I have seen that B&W used 1st order slopes on a number of their models, how do their metal dome nautilus tweeters cope with all that energy?
Often its said that 1st order crossovers do not provide sufficient protection especially for the tweeter. I have read that signals down to 100hz get through at almost full power. That sounds pretty tough for a tweeter to cope with. How do you make a design like that work? I have seen that B&W used 1st order slopes on a number of their models, how do their metal dome nautilus tweeters cope with all that energy?
well, usually the tweeter impedance shoots really high at its resonance, the resistance is so high there that instead of the tweeter rolling off 6db at say 3khz, it goes fairly flat down to say 1.5khz. Then you get carried away with the volume, then oops, why is the tweeter goings schhh, schhhhh on voices ?
And yes, A lot of power gets through a 6db filter. A driver moves 4 times farther as you drop an octave, even a tweeter. And you tend to get more and more power the lower you go in music on the freq spectrum. I think a 6db filter barely counteracts this.
You can build a rugged tweeter with a low resonance. I think rule of thumb is to cross 2-3 octaves above its resonance (for a 6db).
I've been wanting to play with 6db tweets (then you can get time aligned) but havn't yet.
Norman
And yes, A lot of power gets through a 6db filter. A driver moves 4 times farther as you drop an octave, even a tweeter. And you tend to get more and more power the lower you go in music on the freq spectrum. I think a 6db filter barely counteracts this.
You can build a rugged tweeter with a low resonance. I think rule of thumb is to cross 2-3 octaves above its resonance (for a 6db).
I've been wanting to play with 6db tweets (then you can get time aligned) but havn't yet.
Norman
The correct xover choice is determined by the tweeter you wish to use , exactly where you wish to implement the freq. of the xover and power distribution considerations. Ferrofluid and improved construction methods have eased the constraints imposed by first order networks and it useful to consider the quasi second order as a preferred type.
If you look at the current B&W diamond 800 series it seems they use first order crossovers. Are these speakers time aligned? I dont usually think of B&W as a proponent of the phase coherent speaker system and they dont advertise it. Everybody says first order speakers are difficult to do and cause all sorts of problems yet it seems they have pulled it off. How have they managed to avoid all the problems people talk about like cone breakup, tweeter damage, as well as the fact that you need drive units which are fairly flat well beyond the crossover points?
What actually happens if the drivers are not wide bandwidth but roll off quickly? You would get an overall steeper slope however does it mean the phase is no longer phase coherent?
What actually happens if the drivers are not wide bandwidth but roll off quickly? You would get an overall steeper slope however does it mean the phase is no longer phase coherent?
It all depends on the tweeter. Look at their specs. Take their advice. Personally, I would not use a "typical" 1" dome at first order below 4 or 5K. Distortion. Yes, you can take advantages of the natural roll-offs of the drivers. Smart designs often do, but the are NO help with protecting the driver from power. A tweeter with a second order LF roll off at 1K with a first order crossover may give the third order acoustic slope, but for power, it is still a first order crossover.
Don't second guess what big companies that design and build their own drivers may do in some specific speaker and think that is a general rule. You will fail. Been there, done that. Paid for my errors.
Don't second guess what big companies that design and build their own drivers may do in some specific speaker and think that is a general rule. You will fail. Been there, done that. Paid for my errors.
The point about managing the resonance of the tweeter is critical. If you implement a trap to flatten out the impedance of the tweeter at resonance and also across over at a sensible frequency - greater than 3khz for most tweeters, then a first order crossover is no great feat.
Getting the trap right is critical and is something most novices balk at.
The difference in energy a typical tweeter sees with and without a trap on a first order crossover is significant. Totally audible and I would expect it to impact the ruggedness of the driver under abuse. To be blunt the crossover just does not work properly with the variable impedance of an untreated tweeter.
If you don't want to implement the trap (and to be honest I think it is best used for any passive crossover!) then you would be best using a higher order crossover to minimise excitation of the tweeter at resonance.
Getting the trap right is critical and is something most novices balk at.
The difference in energy a typical tweeter sees with and without a trap on a first order crossover is significant. Totally audible and I would expect it to impact the ruggedness of the driver under abuse. To be blunt the crossover just does not work properly with the variable impedance of an untreated tweeter.
If you don't want to implement the trap (and to be honest I think it is best used for any passive crossover!) then you would be best using a higher order crossover to minimise excitation of the tweeter at resonance.
I thought the acoustic slope was what counted? Third order acoustic is third order crossover regardless of the components used to get there, no?
B&W uses pseudo-LR2 slopes. Tweeter damage is avoided since it's an LR2 at 4khz, which means it's already ~30+db down at resonance and I would imagine even more down than that in actuality. (it may be an LR2 in the crossover region, but probably shifts to 4th order out of band)
As for avoiding cone breakup, they don't. Their woofer plays right through it with a crossover around 4.2khz and a cone breakup at 3.8khz.
As for flat... the end speakers aren't flat IN band, so why would they care about out-of-band behavior?
For excursion, acoustic slope matters. But limiting thermal buildup, electrical slope matters too. Also, even if the tweeter is inefficient at x frequency, there is still "attempted motion" being produced from the voice coil at those frequencies - it's just being stifled by the "enclosure" of the box.
Last edited:
No. Not for power, distortion, or staying away from resonance and breakup modes. You have to meet all of these requirements.
I was just testing my most recent set. It uses second order electric in combination with the natural roll offs for effective 4th order at only 1700. (modification to the Zaph SR-71). At my listening levels, it works. I would not push it really hard though.
A lot depends on how loud your 'normal' playing will be. I used 1st order elect on the design at the link below:
Introducing the "INTIMATES" (high WAF & quality sound)
When companies like Dynaudio and Thiel use first order slopes, they use a network to flatten out the impedance curve. If they didn't do this, the *acoustic* slope would not be first order, due to the problem you describe. Due to all these components, the crossovers are actually quite complex and expensive, despite only having one cap in series with the tweeter. (IE the impedance compensation network is more costly and expensive than everything else.)
I know that Danley uses phase coherent crossovers in the Synergy horn, and he's discussed how they work, but I've never figured it out. As I understand it, the Danley xover is phase coherent but it's NOT first order. These days I'm leaning towards MiniDSP as it's so cheap and so simple to use.
I may be mistaken, but I think the deal with Danely's horns is simply like this:
1. Take a few drivers, and make a system with filters that result in the usual 180° or 360° degree phase difference between drivers (assuming they are time-aligned to start with).
2. Offset them in space to bring the phase difference back to 0°.
3. Now everything's fixed, but wait, now nothing works right off-axis because your offsets don't stay constant when you move! Enter the synergy horn.
1. Take a few drivers, and make a system with filters that result in the usual 180° or 360° degree phase difference between drivers (assuming they are time-aligned to start with).
2. Offset them in space to bring the phase difference back to 0°.
3. Now everything's fixed, but wait, now nothing works right off-axis because your offsets don't stay constant when you move! Enter the synergy horn.
Last edited:
If I'm not mistaken, Bill Waslo and Jason Winslow have a grasp on how the crossovers work in the Synergy horn. They've posted some things that detail how the filters work, but it's over my head. I understand horns quite well, but phase is something that I'm slooooowly learning.
It's definitely an important part of the sound; I've always noticed that Dynaudio and Thiel speakers have a type of imaging and articulation that sounds different than typical speakers with high order crossovers.
I always use 1st order slopes on my tweeters. You need decent quality tweeters, preferably underhung with good Xmax but generally it's no drama if you deal with the impedance resonance.
Typical average power is <1W for most listening.
Power is not much of an issue. Say you have 1st order at 4kHz and feed it 2.83V RMS at various frequencies:
4 kHz : -3dB => 0.5W
2kHz : -6 dB => 0.25W
1 kHz : -12dB => 0.06W
500Hz : -18dB => .015W
Nowhere close to frying a tweeter especially since the majority of music energy is below 500Hz.
As for excursion, a decent 25mm with first order at 4kHz will stay within it's linear range below 100W. The other drivers in the system will be distorting before the tweeter does.
Here's my 4way system's displacement limited power handling. The tweeter is the Seas Magnum (T29MF)
Typical average power is <1W for most listening.
Power is not much of an issue. Say you have 1st order at 4kHz and feed it 2.83V RMS at various frequencies:
4 kHz : -3dB => 0.5W
2kHz : -6 dB => 0.25W
1 kHz : -12dB => 0.06W
500Hz : -18dB => .015W
Nowhere close to frying a tweeter especially since the majority of music energy is below 500Hz.
As for excursion, a decent 25mm with first order at 4kHz will stay within it's linear range below 100W. The other drivers in the system will be distorting before the tweeter does.
Here's my 4way system's displacement limited power handling. The tweeter is the Seas Magnum (T29MF)
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
when I started building speakers I use to use the capacitor at the crossover frequency 4.7uf and I lost a lot of tweeters.
Then when I was taking consoles apart I noticed that 90% of every tweeter had caps in the 2uf range.
I started using 2uf's and I quit burning out tweeters and as a plus I normally did not need a resistor to drop the volume.
I am also a fan of adding light bulbs in series to protect the tweeters. Some claim they can hear that but I doubt it.
Then when I was taking consoles apart I noticed that 90% of every tweeter had caps in the 2uf range.
I started using 2uf's and I quit burning out tweeters and as a plus I normally did not need a resistor to drop the volume.
I am also a fan of adding light bulbs in series to protect the tweeters. Some claim they can hear that but I doubt it.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.