WDXRS Supertweeters

[Greg.]

Read my last but one post. I have now decided they are not worth bothering with in my system, however, I know several people who absolutely enjoy their inclusion with their speakers. Each to their own.

[VantheMan]

After testing the Batpures with my WD25Tex this weekend my conclusions are :
1. Better connected to the tweeter outputs of my external crossovers than the full range input on said crossovers-
2. I hear no difference if placed on rear baffle instead of in full view on tweeter face plate
3. I´d rather have them connected than not because sound seems less grainy

All in my system , my room, my well worn cartridge and my lugs yesterday afternoon. Tomorrow, who knows ?

My back aches from reaching round the back of my shelves so I won´t be repeating this testing lark.
Up the Rams

[VantheMan]

I have decided to fit the batpures semi.permanently on the rear baffle firing upwards and maybe slightly tilted back toward the rear wall. I shall drill through the back of the speaker’s ”neck” and install a female mini jack socket flush with the rear baffle and connect direct to the Millennium terminals inside the speaker so no nasty dangly bits of wire. The batpures themselves will be stuck onto a beechwood doorstop in the shape of a quarter of an orange with 2 adjacent flat sides, one for the batpures and the other for the male minijack so they will resemble the backpack active crossovers of yesteryear a la ARC or Isobariks. Then I can experiment to my heart’s content firing left right or vertical or of course remove them if necessary. A dab of blutak will prevent inadvertent rotation.
Sounds good anyway. Photos will be posted when finished but don’t hold your breath waiting.

[VantheMan]

Probably not much use to many here, but I recently lent my Taket Batpures to a mate who runs the Wd20s with the concentric cones and I am astounded by the improvement I hear, much, much greater than the effect with my Wd25TMillenniums so maybe concentric tweeters do benefit more than a separate tweeter which is something I think I read years ago somewhere and would explain wgy Tannoy sell them. In fact it may well have been Peter who mentioned the fact that the advantages of a concentric driver (point source stereo etc.) could be, to a certain extent, offset by the disadvantage of having the tweeter recessed inside a bigger bass cone. Anyway my mate´s system´s sound has improved no end in clarity etc. in fact even deaf old me found a notable increase in the volume level across the board which I have never noticed when using them with my Wd25Ts- Weird.

[petercom]

My take on the super-tweeter issue has always been the same:

1. We can't hear above the upper limit of our senses
2. There's not much (musically) happening above 15kHz other than a few instrumental overtones
3. Enhancing rise time or transient attack is meaningless for frequencies above the upper limit of hearing
4. Most so-called super-tweeters work by adding to the high frequency output in the room.
5. Super-tweeters are still working from mid-treble frequencies upwards

Go back to basic physics. To start with a lot of super-tweeters are sold with first order crossovers. Think about what this means:

For a 'crossover' frequency of 20kHz the treble output will be -6dB at 10kHz, -12dB at 5kHz and still audible at -18dB at 2.5kHz.

So you can see and hear that the super-tweeter is just adding to overall treble output in the room. This is why VantheMan's friend with WD20s heard a big 'improvement' because the treble unit buried in the bass cone of the concentric unit has limited dispersion. So he's hearing an increased level of treble power and detail in the room with a corresponding subjective loudness level too.

It's also why Tannoys, with Dual-Concentric drivers, are always 'improved' by super-tweeters.

Despite Greg's outspoken comment on expense of the WDXRS, this design was based on a Fountek ribbon which is not a cheap film driver like the Takets. If you think a highly specified ribbon driver mounted in a nice piece of oak is expensive, take a look at the Tannoy ST300MG!

If I was doing the same design exercise today I'd use an AMT (my friend at TrueAnalog makes these superb units: http://www.trueanalog.com/product_detail.php?id=16) because of its very wide dispersion. I'd aim for a second order crossover at 15kHz or above (adjust to taste).

Bear in mind that dome treble units, ribbons, planar films, AMTs are all different, not just to each other but from different manufacturers, designers and everything in between. So to say one device is 'better' than another is ignoring the obvious - a good driver is a good driver and a bad one is awful, no matter the technology employed.

Probably the best place to put a super-tweeter is on the top of the speaker cabinet, flush with the front baffle, to give the best chance of integration with the main drivers. Or you could face it upwards just to increase the treble power level in the room without interfering with the direct sound from the main drivers.

There's nothing 'weird' about it!

[Richard]

Great to hear from you Peter!

What you say makes complete sense and gives a good explanation of Van's findings. The mention of "super tweeter" suggests high frequencies being added to a speaker system yet I too hadn't considered the driver would be adding output lower down as it a rolled off, as you clearly describe.

In the past, diy designs would also suggest an easy option of adding such as a Coles 4001 so I suppose similar effects would apply. I don't know the Taket or how Van's friend added them (crossover components and connection) but think if they were fed from the tweeter it would be vastly different (cumulative roll offs) to if they were added as a separate connection, again depending on actual component values.

[Black Stuart]

Hello Peter,
interesting and useful post. Your suggestion of super tweeter placement reminds me of someone I knew in the early 70s' who used to design and build speakers. I listened to his personal speakers which he designed for classical music (the only kind he listened to).

I had never seen speakers like these. They were 3-way with the 18" bass units and mids each in their own cabinet as were the tweeters. He used various materials to isolate the tweeter from the mid unit, it was an excellent sound helped by being in a very large room with high ceilings.

He offered to let me use his workshop and equipment and create a design suitable for my musical tastes - like an idiot I refused as I was travelling a lot and 'possessions' were not part of my lifestyle then.

He used separate cabinets for each transducer to avoid inter-reaction, it worked but is probably not a commercial possibility.

[petercom]

Originally Posted by Black Stuart

He used separate cabinets for each transducer to avoid inter-reaction, it worked but is probably not a commercial possibility.

Commercially this has been done many times but is nearly always a sales disaster unless put together in a cohesive format like Focal Scapa Utopia.

The best I've heard of this format were the original Dahlquist DQ10 (before the higher sensitivity 'upgrade') which had separate baffles for each driver.

In practice isolating each driver completely from the other is horrendously difficult if you want the speaker to be saleable and not look like a collection of bits slung together (as most do).

The benefits are not as great as you might think. For most designs the percieved audible gains are quite small and it's better to focus on optimal crossover design rather than concentrating your efforts on controlling vibration.

In high end speakers with advanced drive units it is, however, a valid selling point. I found a superior way in Mission Pilastro by making the cabinet so inert that the drive unit vibration was effectively negated (see Stereophile review).

[petercom]

Originally Posted by Richard

In the past, diy designs would also suggest an easy option of adding such as a Coles 4001 so I suppose similar effects would apply. I don't know the Taket or how Van's friend added them (crossover components and connection) but think if they were fed from the tweeter it would be vastly different (cumulative roll offs) to if they were added as a separate connection, again depending on actual component values.

Generally the Coles 4001 was added to extend the HF response of a unit like the Celestion HF1300 which has limited bandwidth to 13kHz. Calling it a 'super tweeter' really meant, as far as the Spendor BC1 was concerned, that it took over from the main treble unit and covered the upper harmonics of percussion instruments like cymbals, triangle, glockenspiel and so from 13 - 30kHz.

And you can see how the 4001 added to the HF power of the HF1300 in this analysis: http://www.troelsgravesen.dk/vintageBC1.htm

The much later, so-called 'super tweeter', extension above the audible limit was coined simply because Sony recommended that SACD replay required associated components that had response extensions to 100kHz. I think they really meant that components like amplifiers should not have bandwidth limiting (as Naim did, for example), because the immediate benefit of SACD was to remove the brick wall filter of conventional CD at 22kHz with all its associated filter ringing effects.

Many numpties assumed that speakers should also not be bandwidth limited in this way! As a result many Japanese distributors demanded that speakers MUST have responses to 100kHz or they couldn't be sold in the asian market.

Tannoy quickly responded (as their biggest high-end market was Japan) with a 'super tweeter' cobbled together using an Audax dome in a fancy enclosure.

As I was being asked by the Mission Japan distributor at that time to design new speakers that were 'SACD compatible' I asked Audax what this magic dome was that could reach 100kHz? It turned out that it was a standard dome with a tin-can resonance at 28kHz. This meant that the second order harmonic distortion carried the output to 56kHz and the third order harmonic occurred at 84kHz.

Of course no-one could easily measure the response of this super-tweeter (B&K mics normally used in anechoic chambers are generally limited to 30kHz) so nobody questioned this.

What amused me was the various reviews (some by critics who should have known better) who were amazed by the subjective response when adding this super tweeter and came up with all sorts of stupid ideas about how we could hear the extended response.

At the time there was a lot of rubbish talked about 'feeling' the ultrasonics (through your skin or eyes or whatever) or the faster rise time of transients having some sort of subjective effect (though, if you look at how the human ear works, it's pretty obvious you can't actually hear this).

What nobody investigated was, as I pointed out, that the output of these super tweeters added to the power response in the room. And so the myths continue!

[Greg.]

Hello Peter, Good to see you here once again and sharing your considerable knowledge.

Originally Posted by petercom

Despite Greg's outspoken comment on expense of the WDXRS, this design was based on a Fountek ribbon which is not a cheap film driver like the Takets. If you think a highly specified ribbon driver mounted in a nice piece of oak is expensive, take a look at the Tannoy ST300MG!

I don’t know if you recall our original conversation on this point, but my comments to you at the time were specifically about the cost of the Oak block you were marketing as a mount for the super tweeter and not the cost of the Fountek ribbon tweeter itself. At the time I was very aware of the expense of quality super tweeters. My ‘outspoken comment’ is meant to be mainly a reference to the enclosure you were at the time selling which, realising your market was the DIY community, in my view was seriously over priced. Most DIYers would make their own mount anyway. That is what I told you at the time and remains the case even today.

I have listened to several systems with super tweeters fitted. In some cases they do bring a sound improvement (as VantheMan has found) and in others they just change the sound slightly but don’t bring an improvement. It’s yet again suck and see and I completely acknowledge that their effect may well be related to the actual listening room structure. In my room, the TakeT BatPure’s altered the sound but I wouldn’t say they offered an improvement so they have now been sold on. Phil Y and I trialed his Townsend super tweeters in my system and they made no difference to the sound.

This was with my WAD KLS3 Mk3 (Millennium) which I continue to love. Thank you so much for your publication of the upgrade you did to Guy Pettigrew’s KLS3’s. Acting on it has brought massive benefit to my sound and I love it every time I switch on