Here here! I couldn't agree more.
For years now I've felt that 1" dome tweeters were at an evolutionary dead end - they might have been a decent improvement at an affordable price over what was commonly available when they entered the market in the 70s and 80s, but despite ever improving measurements there is something fundamentally wrong with the way they sound. Some dome tweeters can sound good, but none sound truly outstanding.
The modern trend of pushing them lower and lower in frequency is part of the problem, and really annoys me. Even crossing a dome at 3Khz in a large system is a bit marginal IMHO, so to see them regularly being crossed over at 2Khz or even 1.5Khz baffles me, and is heading in completely the wrong direction.
I'm 100% sure someone will leap to the defence and tell me "but model X works perfectly well crossed over at 1.5Khz and is rated for it by the manufacturer!". Well, if you believe a 1" dome truly works well and sounds good crossed over at that frequency in anything other than a low SPL bookshelf speaker, good luck to you. I have a seaside resort at the south pole you might like to buy as well.
The comment Dave made about mounting right or wrong is a good one, and relates to diffraction. Unfortunately over time I've come to believe that there is no right way to mount a conventional flat faceplate dome tweeter that adequately solves diffraction effects, and that that is one of the major failings of the common dome tweeter. (Ironically its the "wide dispersion" that is frequently touted as a benefit that is to blame)
Diffraction in the treble region is very important due to the small wavelengths, as it leads to significant variations in response over a very small angular shift near on axis which can contribute inconsistency in imaging as well as harshness.
Flush mounted on a flat panel is not enough, and will still have severe diffraction artefacts from both the nearby cabinet edges and nearby drivers. A tweeter is not a point source if the outline of the cabinet and nearby drivers are also re-radiating a significant portion of the treble over the bottom 2/3 of the tweeters frequency range.
Some sort of absorptive material stuck on the baffle near the tweeter may help a bit but is kludgy and not a real fix, and doesn't work at lower frequencies.
A faceplate-less dome with a curved pod mounting (like the B&W Nautilus 802) may help somewhat but is extremely complex mechanically (for DIY) and still doesn't really solve the problem completely as there will still be reflections off the rest of the cabinet.
I think the only real solution to this problem is some full-bandwidth directivity control, with wave-guides being a particularly good approach.
Constrain the radiation pattern of the tweeter to no more than (say) 90 degrees and a lot of problems go away, chief amongst them being diffraction off the cabinet edges, which is reduced to a nearly insignificant residual amount with the remainder being able to be easily dealt with by careful cabinet design.
Eliminate the bulk of the diffraction signature and reflections, and now your tweeter behaves as a proper point source across its entire frequency range, which is beneficial for imaging and smoothness.
Controlling the directivity at the bottom end of the tweeters range means a better directivity match to the midrange/midbass driver, which usually means the crossover frequency can be higher, helping to reduce power handling needs and IM distortion.
A wave-guide will typically improve the sensitivity/power handling/distortion of the tweeter at the bottom end of its range as well, so you have a much needed double improvement in overall performance at its low end.
Depending on choice of drivers, the depth of a wave-guide often helps to better time align the drivers, meaning simpler crossover design and better coherency.
At the top end their range, all soft domes have problems with cone resonances, typically in the 6-12Khz region. At the first major resonance the centre of the dome actually moves out of phase with the edges causing a dip in the response, with long term ringing after the impulse is gone. This shows up clearly in a CSD plot.
Damping of the dome material can help a little but not solve the issue. The basic problem with a dome at high frequencies is that there is nowhere for the bending waves to be terminated as there is with a cone driver.
With a dome the impulse starts at the outside and propagates towards the centre, but then keeps going past the centre to reflect off the other side. It's not possible to add sufficient damping at only the centre to terminate the bending waves, and if you make the entire dome lossy to allow the wave to dissipate before reaching the far side, the high frequency response is significantly sacrificed.
In a cone driver the source is at the middle, bending waves travel outwards, and have the opportunity to be terminated and absorbed at the edge after the sound impulse has already been radiated from the cone. Most drivers don't completely absorb the bending waves at the edge termination due to impedance mismatches with practical materials, but at least the topology allows for it.
The dome topology, driven at the outside, is simple and easy to build, but its far from ideal at high frequencies. Another approach is a rigid metal dome which gets around the above issue, but now you have a giant supersonic resonance that (according to some findings) can inter-modulate with in-band signals.
Of the best domes I've heard, I think the metal domes sound better, but I still find them a bit fatiguing to listen to.
For the top end of the treble I really like the sound of good ribbons - so far I haven't heard any dome that simultaneously has the crispness and smooth fatigue free characteristic of a ribbon, I find dome tweeters either sound a bit dull, or if they're crisp then they also sound fatiguing and sometimes a bit harsh.
The problem with ribbons is they need to be crossed over even higher than a dome, for example I cross my 50mm ribbons over at 4Khz, and wouldn't cross them over any lower for intermodulation and power handling reasons. Many ribbons are also wave-guides, so you can get a lot of the advantages of the wave-guide at the low end and the ribbon at the high end in the same driver.
If you're willing to cross over at 4Khz or above to a very good midrange driver a ribbon is hard to beat IMHO, otherwise if a lower crossover frequency is needed then a larger format non-ribbon wave-guide driver is probably a better choice.
Both are superior to any 1" dome tweeter IMHO, and I think if a few more people woud venture off the cone + 1" dome design rut they would discover that there is a lot more out there and that they're missing out on some good design approaches.
Last edited:
I'm not sure where you're drawing the line between horn loaded and wave-guide loaded, but the Aurum Cantus G2 is 96dB/w/m with a small built in wave-guide, the larger G1 is claimed to be 102dB/w/m, both are hardly "low efficiency". I haven't measured the latter, but I have the G2 and they do indeed measure 96dB compared to other known drivers.
http://www.aurumcantus.com/aurumcantus-tweeter/aluminum ribbon tweeter G2.htm
http://www.aurumcantus.com/aurumcantus-tweeter/aluminum ribbon tweeter G1.htm
Last edited:
Not having looked at it in much detail, the ribbons seem to be rated in volts and use a transformer. This may not be the same as "efficiency" which must be done in "watts" not volts. I am not sure how this would all factor out, nor how this data is measured. Is the 2.83 volts input to the transformer?
After all this I have to say that the best commercially available speakers I have heard (ATC SCM50Pro and above) use surprisingly cheap 1" dome tweeters.
They've changed tweeters now but the ones I heard used Vifa D27TG-46-06.
On the other hand I personally use Tannoy DCs with a 2" compression tweeter and a substantially lower crossover point at 1.2kHz rather than ATCs 3.5kHz and I am extremely happy with them.
They've changed tweeters now but the ones I heard used Vifa D27TG-46-06.
On the other hand I personally use Tannoy DCs with a 2" compression tweeter and a substantially lower crossover point at 1.2kHz rather than ATCs 3.5kHz and I am extremely happy with them.
I can't speak for the Fountek drivers, but the sensitivity of the Aurum Cantus ribbons is rated in terms of input watts not 2.83v, and all of them have built in step down transformers to make the input impedance very close to 8 ohms as the foil resistance itself is under half an ohm.
Rated input impedance of the G2 is 8 ohms, and I measure a minimum input impedance of 7.8 ohms on mine at around 3Khz, rising to a maximum of about 8.5 ohms at 20Khz. (In fact it stays between 7.8 and 8.5 ohms from 800Hz to 20Khz, rising to 15 ohms at 200Hz and falling again below that)
So they are indeed 96dB/w/m as measured at the input terminals, no magic tricks or hocus pokus
Last edited:
Quote from good old wikipedia...
Driver ratings based on the SPL for a given input are called sensitivity ratings and are notionally similar to efficiency. Sensitivity is usually defined as so many decibels at 1 W electrical input, measured at 1 meter (except for headphones), often at a single frequency. The voltage used is often 2.83 VRMS, which is 1 watt into an 8 Ω (nominal) speaker impedance (approximately true for many speaker systems). Measurements taken with this reference are quoted as dB with 2.83 V @ 1 m.
Translation 2.83V @ 1 m = 1 watt
Driver ratings based on the SPL for a given input are called sensitivity ratings and are notionally similar to efficiency. Sensitivity is usually defined as so many decibels at 1 W electrical input, measured at 1 meter (except for headphones), often at a single frequency. The voltage used is often 2.83 VRMS, which is 1 watt into an 8 Ω (nominal) speaker impedance (approximately true for many speaker systems). Measurements taken with this reference are quoted as dB with 2.83 V @ 1 m.
Translation 2.83V @ 1 m = 1 watt
So they are indeed 96dB/w/m as measured at the input terminals, no magic tricks or hocus pokus
That's also what I've seen in "HOLM made" measurements. Myth confirmed!
Slight correction to myself, not all are 8 ohms - I forgot the G2si and G3si are 6 ohms, however the G1, G2, and G3 are all 8 ohms.
I do watch myth busters but I was reading you too literally, so didn't catch the reference
Yes exactly, the G2 (and G1) are almost exactly 8 ohms so the 1W rating = the 2.83v rating.
Edgar Villchur who passed away recently invented the dome tweeter and dome midrange. AR3, the first to use either was the world standard of accuracy in its day and can be seen in the Smithsonian Institution as an example of excellence in American engineering and manufacture. AR3 was featured in a number of successful LvR demonstrations Roy Allison conducted and I was lucky enough to attend 2 of them. AR3a superceded it and it became a world standard too. AR LST increased lateral dispersion and power handling capacity of the AR3a by incorporating 4 dome midranges and 4 dome tweeters including some at panels angled 45 degrees in either direction to the front. Dome tweeters of this type are not in current manufacture and those who have these speakers and need replacement drivers are usually out of luck and must settle for substitutes with substantially inferior dispersion.
I think that we are all away of the history, but I for one would like to think that we have progressed somewhat in the last 50+ years.
When there are more successful demonstrations that show that a recording of music can be made to sound very much like the real thing by direct side by side comparison to the satisfaction of experienced critical listeners of live music I will agree but not before. Until than I conclude that the endless variants of the same ideas are all flawed. IMO while materials have improved and methods of testing and analysis have also improved, the ability of recordings to convince serious listeners that they are hearing live music remains unachieved. Even the Allison demonstrations were highly contrived. listening to commercially made recordings AR3 was also unconvincing.
Wouldn't even "good" tweeters cost less than an equivalent-quality woofer or midrange simply because there is less material used in its construction?
BTW this may nauseate some audiophiles but Cerwin-Vega is the only brand I know of that uses a soft-dome tweeter paired with a...well, I think it's a horn because it appears much too large to be a wave guide: for example this 3-way model with 12" woofer. Crossover point for the tweeter is listed as 2.4kHz, and 380Hz for the midrange driver.
BTW this may nauseate some audiophiles
but Cerwin-Vega is the only brand I know of that uses a soft-dome tweeter paired with a...well, I think it's a horn because it appears much too large to be a wave guide: for example this 3-way model with 12" woofer. Crossover point for the tweeter is listed as 2.4kHz, and 380Hz for the midrange driver.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.