Super Tweeters

[izzy wizzy]

What are they all about? I got to hear some Taket-Live gizmos in my own system and another Tannoy system that's very different.

These things https://www.dagogo.com/taket-live-super ... e_vignette

In my friends system which is digital; Roon streamer on a NUC, Topping D90 and Pre, Neurochrome amps, it sounded a bit more lush to me. My friend reckons they did a whole lot more much like the review above. In my system, I got an almost instant pain in my head/sinuses. Unbearable.

In my friends system there is no information above 20kHz so they can't be dealing with "signal" as they kick in at 20kHz or so the spec says. There can't be a brick wall filter beforehand so maybe it's stuff a bit lower than 20kHz. I realise extra HF can make the sound stage seem bigger; more air. Are they making stuff up and spaying it around?

In speaker design AFAIK, the tweeter has to be close to the driver below otherwise there's all sorts of comb/lobe additions and cancellations. These things are normally positioned a fair way from the other drivers as they sit on the top of the box. Seems to ruin the idea of a coaxial driver.

My question; is this some kind of effects generator or is something else going on? Kinda glad they didn't do anything positive in mine

[chris661]

There are a few potential effects of adding supertweeters:

- Wider HF dispersion
- Increased HF output
- Depending on the HPF being used, they might well be adding some harmonic distortion to the lower treble, resulting in more upper-treble.


I can imagine that some combination of the above would result in more "air"/"shimmer"/whatever, but only one of the effects above could be duplicated via a simple HF boost to the existing speakers.

IMO, a well-designed speaker shouldn't need a supertweeter. If you're stuck using a 1.5" dome tweeter, or a compression driver & horn combo that goes narrow >12kHz, then the wider HF dispersion may be an audible benefit, if that's the presentation you prefer.


Chris

[Nick]

Just remember that one octave down from 20kHz is 10kHz, and if the super tweeter has a first order roll off below 20kHz, then there will be -6dB added signal at 10kHz. Even with a 20kHz brickwall there will be content at 10k. Adding extra signal at some distance from the original tweeter is going to create all sort of filtering effects. Just go and listen to what adding a flanger to a signal sounds like. Also that looks like its going to have a very different dispersion pattern from the tweeter in the loudspeaker face.

[izzy wizzy]

Good points. Thanks guys. Seems like they are effects generators and some may like, some not.

I suppose if there's nothing else to fiddle with, then these things and cable swapping can be part of the hobby. Different but better?

[rowuk]

My experience is that I do not "hear" the super tweeter when doing a frequency sweep. I just hear what it does to instruments like trumpet, percussion, violins, piano, oboe. Those instruments have an appreciable portion of "articulation" to get the individual notes started and the general "close miking" picks up an unrealistic proportion of that.

[Daniel Quinn]

Nicks is the first time I have read an explanation of supertweeters that doesn't involve the magical realm of belief

[Nick]

Nicks is the first time I have read an explanation of supertweeters that doesn't involve the magical realm of belief

That may be because I am neither trying to sell some or to justify the purchase

[Scottmoose]

'twas ever thus.

There are lots of reasons why a supertweeter can (as in can) have an audible influence. But almost invariably not for the reasons touted by the sellers, magazines, hi-fi fans etc. A few, in no particular order:

• It depends what you actually call a supertweeter. They aren't all the same thing. Back in the day, with something like a Spendor BC-1, then its main Audax dome tweeter was about done by 12KHz - 14KHz, & the additional unit supplied the missing ~1/2 octave or so.

• The reason Nick describes. And there are a lot of elements to that, as he says. Attached is John Atkinson's FR measurements of the Harbeth HL5XD, on the vertical axis with the supertweeter, averaged over a 70 degree horizontal angle. Not a great set of measurements, but they're okay for purpose here. It's a useful example because despite appearances, that Harbeth doesn't actually work quite the same way as the BC-1. The Harbeth main tweeter can do 20KHz all by itself -it doesn't need any help to do that at all. But the supertweeter is also operating above 10KHz, even though there's no real need for it to do so. Unlike the old BC-1 from the '70s, it isn't just supplying output when the main tweeter has given up the ghost. You can see & hear the result: because you've got two tweeters operating over at least the 10KHz - 20KHz frequency range, you get some fairly severe variations in the response above 10KHz due to the phase-related cancellations. The tweeters receive & reproduce the same signal, but they are located at physically different distances from the listener (or the microphone in this case), so the minor delay with an otherwise identical signal results in cancellations that vary in severity with frequency and the listening position. Assuming your hearing is good enough for this region -you'll hear it. The same thing applies to any supertweeter that's added to a system and operates over the same range as the main tweeter. The details will vary depending on where you put it, but the basic principle holds. You might like what it does (fair enough), or you might not (ditto). One favourite is a rear-firing supertweeter -Terry Cain did that with his SuperAbby for e.g.

• A more esoteric effect may (may) come from a dome resonance in the supertweeter. For e.g., say you have a metal-dome supertweeter that has a ruddy great bell mode at 26KHz like the Harbeth has. You get distortion amplification from that dome resonance at 1/3 of that frequency if it (either, actually) is triggered, sharpening the sound & adding some 'false air', 'false detail' / 'glare' / whatever term you prefer. It'll amplify the 5th order HD at 1/5 that as well. This is more an issue for main than supertweeters, since they're usually being rolled off by that point, & I suspect it's a fairly minimal effect compared to the other potential issues, but it's a physical mechanism that exists, and may have an effect depending on how it's implemented, & how good the supertweeter's motor & distortion performance is.

• The electrical load the amplifier 'sees' will change by adding a supertweeter. Since the electrical characteristics of the amplifier-wire-speakers form a summed circuit, by adding another element, the impedance load will change -how much depending on the supertweeter. You might for e.g. get a new impedance peak at, say, 10KHz (to pull a number out of the air). If the amplifier has a high[ish] output impedance / low DF, the actual system response is likely to follow the impedance curve, so you may start seeing shifts in behaviour that way.

None of which is to knock supertweeters -just to point out a few functional reasons why they can have an audible effect of some kind, which in the case of tacked-on additions, is likely to vary quite a bit in different systems & with different positioning.

[Nick]

The electrical load the amplifier 'sees' will change by adding a supertweeter.

I have to admit I was thinking like a loudspeaker designer there and assuming the amplifier was perfect and could provide full power into either a capacitive or inductive nail. (at the sane time )

[Scottmoose]

Well, you covered it anyway when you mentioned the extra signal effects.

Agreed -a lot (far too many) of speaker designers tend to assume a voltage source / near infinite damping, which is probably a bit short-sighted as they've lost sight of the system context. Not all though, in fairness. Until a couple of years ago, Fostex for e.g. assumed most of their E / En / ESigma series units would be used on the end of an amplifier with an output impedance of about 2.5 - 6ohms, so when you designed an enclosure for them, ideally you factored that into the alignment. Lowther were / are often quite similar in that too. Not that it matters, that's also why I tend to aim as far as possible for relatively easy / unreactive loads with multiway systems, so the damn thing stays as consistent as possible between different types of amplifier, and it also isn't demanding rapid current swings, which some might struggle to cope with.