Any tweeters out there that are will handle a 1st order acoustic crossover@2500Hz (or so) for reasonable levels (100 db)?
Not counting super expensive Bliesma T34 😱
Maybe Wavecor TW030WA11/12?
Or SB29RDNC?
Not counting super expensive Bliesma T34 😱
Maybe Wavecor TW030WA11/12?
Or SB29RDNC?
I don't think that those tweeters have ferrofluid in the gap. Check Seas
And... Acoustic x-over?! Who knows...
And... Acoustic x-over?! Who knows...
I use the wavecor and SB with good results but with 2nd orders. The tw030wa are a good bet and independent measurements look very good. What mid or woofer are you thinking about.
Yes, acoustic. A single cap usually gives a 2nd order acoustic roll-off with most tweeters even tho it's 1st order electrical.
No woofer yet, but finding a suitable woofer for a 1st order acoustic XO is less of a challenge 😀
Take a look at this one. Very efficient so it can be padded down and it has a low Fs. I've been very happy with it although I use it with a 2nd order Xover.
Morel Cat 378 https://www.morelhifi.com/product/cat-378/
Morel Cat 378 https://www.morelhifi.com/product/cat-378/
Last edited:
Electrodynamic loudspeakers by nature display a 2nd order acoustic highpass because of acoustics and the dynamics of a mass and spring system. There is something called radiation impedance, which tends to lead to a built-in highpass function. So I cannot imagine what you try to reach, unless you would compensate for it by brute force. Which would invariably lead to deceased tweeters.
I'm not so sure about that. I used a pair of Morel CAT 378 tweeters with 1st order xovers in PA speakers I made that were used very loud with a live rock band. Total band level was typically around 105~110 dB. So how loud would that make the PA? Not sure but apprroaching 100, if not at 100, dB. But I do agree. 100 dB is excessive in the home and not likely to be needed or reached. Anyway, I blew a woofer on those PA speakers so disassembled the pair rather than rebuild them. But the tweeters still sounded great and I now use that same pair in two way monitors I built for my living room. In any case, experience tells me 100 dB is doable with a hi fi tweeter, if not desirable or really needed.
I don't think anyone is expecting the first order response to hold up that far away from the crossover frequency.
Indeed the true difficulty of a first order acoustic design is how many octaves below fc can you maintain the 1st order acoustic response. For the reasons you mentioned this is a lot harder than you might assume.
With a 2500Hz fc some tweeters are flat till ~500Hz which would indicate roughly two octaves of actual 1st order overlap going down to the midrange. You'd need a midrange flat to 10kHz to accomplish the same thing in reverse. Obviously the crossover is going to be more complicated than a single cap.
I'd be more concerned about tweeters that can actually give you a decent transfer function rather than their absolute ability to play loud but this isn't my speaker.
Contenders, imo, would be.
D3004/602010
D2905/970000
SB26ADC-C000-4
TW030WA tweeter series
All of these have low distortion and very extended lower responses. Screw the impedance peak, due to lack of ferrofluid, any of these are going to need an LCR to destroy the impedance peak anyway.
Personally I like the idea of the small scan as it can get the C2C distance as low as possible but it's the least happy at volume. The TW030 will most likely play the loudest but that faceplate is huge. The ADC has pretty impressive xmax for its size but needs it with the small SD, the 97 is excellent but probably not worth the extra vs the ADC.
You're going to need to use something like the scanspeak 10F as a midrange though.
Never found a dynamic tweeter I would cross below 2K even with a second order crossover. For a 1 inch dome, 2200 to 2500 is usually more suitable. Excursion becomes the limit. The SB's seem to do a tiny bit better than most. There are a few 1 1/8 domes out there that may work a bit lower and shallower. I was trying to find a ribbon or AMT that looked well behaved that could cross as low as 1700 to make integration with the ER18's easier, but not in my price domain.
Why a first order? That would imply similar on the woofer and finding even a 5 incher that behaves above 2500 or so near impossible. As the overlap is so great with first order, it makes a huge mess of the phase alignment. First order you have to deal with impedance spikes of the tweeter and resonance spikes in the mid-woofer. 3rd or 4th can often avoid them.
Why a first order? That would imply similar on the woofer and finding even a 5 incher that behaves above 2500 or so near impossible. As the overlap is so great with first order, it makes a huge mess of the phase alignment. First order you have to deal with impedance spikes of the tweeter and resonance spikes in the mid-woofer. 3rd or 4th can often avoid them.
Another thing to bear in mind with low order crossovers is distortion:
To quote Lynn Olson
" All direct-radiators (tweeters, mids, woofers, etc.) have diaphragm excursion that increases with decreasing frequency at a rate of 12dB/octave. (Drivers are constant-acceleration devices.) To prevent the excursion from actually increasing below crossover, you must use at least a 2nd-order (12dB/Octave) crossover ... even this rate simply keeps the excursion constant below the crossover point."
To quote Lynn Olson
" All direct-radiators (tweeters, mids, woofers, etc.) have diaphragm excursion that increases with decreasing frequency at a rate of 12dB/octave. (Drivers are constant-acceleration devices.) To prevent the excursion from actually increasing below crossover, you must use at least a 2nd-order (12dB/Octave) crossover ... even this rate simply keeps the excursion constant below the crossover point."
Hmm.... I have great respect for Lynn Olson but..... if that were true enough to imply a necessary hard rule on xover slopes, then how can the very many commercial 1st order designs be viable? How can my own past experience be viable? A perfect example of a well functioning 1st order design is the venerable Dynaco A-25of which they sold 100's of thousands. So I'm not sure how one can say a 12db/ocatve must be used. Something is not adding up with that concept.
1. order is a hard ask no matter what.
Do also consider the 2 Peerless Corundum Domes:
DA25TX00-08 (fs 633 Hz, 0,7 mm Xmax)
DA32TX00-08 (fs 573 Hz, 1.0 mm Xmax)
Waveguides will give a boost in lower freq which will also help.
I'm trying out the DA25TX00-08 in a Visaton WG148r ... but will cross over by DSP ..... soon ready for first listening test 😉
Do also consider the 2 Peerless Corundum Domes:
DA25TX00-08 (fs 633 Hz, 0,7 mm Xmax)
DA32TX00-08 (fs 573 Hz, 1.0 mm Xmax)
Waveguides will give a boost in lower freq which will also help.
I'm trying out the DA25TX00-08 in a Visaton WG148r ... but will cross over by DSP ..... soon ready for first listening test 😉
If you measure and map the distortion, you find it actually skyrockets much above 1/2 X-Max. Now realize X-max is a very vague spec, so you have to measure. I use 96 dB as an arbitrary level and design my crossover to be under the elbow.
Waveguides actually don't boost, they provide more efficient coupling into the air. So, the result is a slightly extended low end limit. The driver is doing the same thing. Just to be clear on the physics.
I was very interested n the larger Peerless, but it's breakup is brutal and that is an issue I believe is causing the "glare" problem I have been chasing.
Why comercial first order?
Two very good reasons:
They are cheap. ( valid)
Slick page editorials try to convince you they are better so they can fill a full page review, put the kickback in their pocket and still sell the full page ad in the back.
Why so many DIY? That too is simple. Folks who do not understand the physics or have not done the tests.
The other , IMHO really stupid thing I see all the time is trying to use the drivers roll off on both ends as part of the crossover. Flat does not work as well as doing it correctly. Often they wind up having to add a notch here and there using more parts than if they had done it correctly in the first place.
Waveguides actually don't boost, they provide more efficient coupling into the air. So, the result is a slightly extended low end limit. The driver is doing the same thing. Just to be clear on the physics.
I was very interested n the larger Peerless, but it's breakup is brutal and that is an issue I believe is causing the "glare" problem I have been chasing.
Why comercial first order?
Two very good reasons:
They are cheap. ( valid)
Slick page editorials try to convince you they are better so they can fill a full page review, put the kickback in their pocket and still sell the full page ad in the back.
Why so many DIY? That too is simple. Folks who do not understand the physics or have not done the tests.
The other , IMHO really stupid thing I see all the time is trying to use the drivers roll off on both ends as part of the crossover. Flat does not work as well as doing it correctly. Often they wind up having to add a notch here and there using more parts than if they had done it correctly in the first place.
How about Heil ESS? can go much lower...
https://essspeakers.store/collectio...al-great-ess-heil-amt-air-motion-transformer™
IMO the only hope for the OP's idea is to use a driver that is not exactly a tweeter, but can be used as such: a small full range driver. Using a true tweeter with a first order acoustic rolloff as low as 2.5kHz is just... a bad idea all around.
With a cap it ends up as a 2nd order acoustic roll-off.
Probably not. Maybe 96 db is a more reasonable aim.
The SB26ADC looks like a decent contender. Seems the small scan is a little short on excursion.
I was hoping on finding a beefy tweeter with a small faceplace to match with a small 4-5 inch woofer for close spacing.
Then I have a little more slack for crossing higher if necessary.
Tons of 5 inch woofers that behave perfectly fine with close to no breakup.
Impedance spike of tweeter has to be dealt with regardless (typically).
As for phase mess, are you talking about lobing? If the target function is met then phase should be pretty flat for a 1st order no?
They are interesting tweeters for sure! Just wish they had a small faceplate neo version. That ferrite faceplate is just huge.
Their frequency response is quite hard to work with, so I've never seriously considered them.
Some of these new 1.3 inch tweeters with good excursion capability is fine with 2nd order at 1500 Hz.
Then 1st order at 2500-3000 Hz should be fine for reasonable spl levels.
I admire your confidence, but I fear there may be a problem with power input to the driver. Here is why:
Even a large format dome tweeter might have Fs=500Hz and Q will be low, like 0.5. For example, Wavecor TW030 series has Fs=450, Q=0.49. This means the tweeter's unfiltered response is already dropping off and is down 6dB at 500Hz (for my hypothetical unit). Your first order acoustic slope is actually less steep than the tweeter's own roll off and at some point power attenuation will drop to zero. A 3k first order response is down about 18dB by 375Hz but the tweeter's own response might be down by 10-12dB by that point, meaning input power is only being attenuated by 6-8dB. At some point attentuation will fall to zero as the tweeter's 2nd order response overtakes the desired first order acoustic response.
A tweeter's voice coil is very delicate and it doesn't take much input power to heat it up. The only reason that you can drive a loudspeaker with hundreds of Watts and the tweeter survives is because the crossover filter limits the power delivered to it. With a first order acoustic response, the tweeter will see a lot of out of band (lower frequency) content. Sure, if you only play with limited amounts of power it will survive, etc. but why do it?
Even a large format dome tweeter might have Fs=500Hz and Q will be low, like 0.5. For example, Wavecor TW030 series has Fs=450, Q=0.49. This means the tweeter's unfiltered response is already dropping off and is down 6dB at 500Hz (for my hypothetical unit). Your first order acoustic slope is actually less steep than the tweeter's own roll off and at some point power attenuation will drop to zero. A 3k first order response is down about 18dB by 375Hz but the tweeter's own response might be down by 10-12dB by that point, meaning input power is only being attenuated by 6-8dB. At some point attentuation will fall to zero as the tweeter's 2nd order response overtakes the desired first order acoustic response.
A tweeter's voice coil is very delicate and it doesn't take much input power to heat it up. The only reason that you can drive a loudspeaker with hundreds of Watts and the tweeter survives is because the crossover filter limits the power delivered to it. With a first order acoustic response, the tweeter will see a lot of out of band (lower frequency) content. Sure, if you only play with limited amounts of power it will survive, etc. but why do it?
Just because a tweeter rolls off at 12db at a certain frequency with a 6db xo doesn’t mean it isn’t trying to make those frequencies. The acoustic output drops but the dome still tries to move too far. The Xmax will be exceeded. The AMT idea is the best in my estimation as the large one can be used down to 1200 easily so 2500 is no problem with a 6db xo. They sound great too.
2nd the morel cat378
With its 5.6uf (don't need z-peak filter even), it is -6db @ 2khz and -12db @ 1khz........................
I ran it with my onkyo crossed @ 200hz (to eminence b102 and the cat378), as loud as it would go, didn't blow.
With its 5.6uf (don't need z-peak filter even), it is -6db @ 2khz and -12db @ 1khz........................
I ran it with my onkyo crossed @ 200hz (to eminence b102 and the cat378), as loud as it would go, didn't blow.