SuperTweeter Questions

[Alonzo]

So I'm thinking of adding supertweeters to my systems.  What's everyone's opinion on there use?  Do you use a seperate amp or bi-amp your (...) (speakers of choice)?  I have Frugal horns and Sexy speakers, both could use a little help in the high end.  I've got some Tangbands laying around, I'm going to stick in a cardboard tube with a 3 to 5 uf cap to see how they sound.
Alonzo

 

[4krow]

Well, there is a lot to consider when adding to a speaker system. You would not necessarily want to bi-amp here(but I'm sure there are those who would argue). And it would help knowing things like the frequency response of your present system, and the frequency response of the added driver as well. Then there is efficiency to work with between the two drivers. Not least here is the crossover frequency(if the tweeter doesn't have the ability, i.e. a 'regular' design, not piezo).

 

[Clark B.]

I used to use the dual outputs of my fpIII to drive paramour II's into fostex FE166ES-R's and also a sex amp powering T90AEX's rolled off w/.68 fostex caps. The line level VC of the sex amp seemed a tad more transparent than the L-pads I tried.  Also, stereo volume control for the tweets rather than matching l-pads by eye.

Not sure of the long term correctness of this arrangement as it was amongst many things I was experimenting with at the time

Cheers!

Clark

 

[Paul Joppa]

Of course, a reasonable match of sensitivities is a good starting point.

Ideally you would have both tweeter level and crossover frequency adjustable, since there is usually a lot of experimenting involved.

The most widely successful high-pass crossovers for supertweeters seem to be either second order at 8-10kHz, or first order at 15kHz of higher. Below that crossover, the tweeter interference pattern with the main system becomes too obvious.

In fact, if the tweeter is much more sensitive than the woofer, sometimes it works to do first order very high, like 50kHz or even more, using that to adjust the sensitivity. I suspect this works because the tweeter actually does roll off, at least in power, even if the on-axis response is extended.

In some cases, people have had better success aiming the supertweeter away from the listening position, so it fills the reverberant field without drawing attention to its location.

Many systems (but not all!) have enough treble rolloff that you don't need a lowpass filter on the main system. Note that if the main system has nasty treble (as opposed to a lack of treble) the supertweeter won't fix that, and you'll probably have to cross much lower to solve that problem. That's a whole 'nuther ballgame.

 

[earwaxxer]

I'm not crazy about the "super tweeter" concept. I heard the new Magnepan 20.7's and I think the super tweeter just adds confusion to the sound. Certain sounds and instruments sound great and super real. But, then when the frequencies go mid range, the mind senses something is wrong. I would use horns for the midrange and ribbons for the highs if you have to use horns. Crossover the ribbon about 5K and you got your super tweeter and then some.

 

[2wo]

I mostly listen to a pair of Hornshoppe Horns with Fostex 126en, I think.

I have a pair of the T-90 tweeters. Yea, they go higher, I can get a bit more air if I turn them to the rear.

but for me, all in all, I am happy without them...John   

 

[Grainger49]

I have always used a cap high pass into the supertweeter.  It doesn't present a bad load to the amp and is a gentle enough roll in.

 

[Alonzo]

Well, thanks for all the info.  I tried them out last night, with both speaker combinations and with 3 different amps and 3 capicator values (.66, 1.5, 4.7).  None were bad, the best of the bunch was with the tweeters augmenting the Sexy Spkrs on an EL84 amp.  Since my horns are what I mostly listen to, I"ve got to do my homework to make this match up properly. 
The Tangbands are 700 - 30 KHz, the FE126En's have a pretty broad range.  I like just the cap high pass but I like Clarks use of the VC for a L-Pad, I see lots of experimenting over the summer.

Clark/John, are the T90's worth the expense to upgrade over Tangband 1744S?

 

[2wo]

Alonzo, I sent you a PM...John

 

[Clark B.]

Not sure about the cost/benefit ratio, have never tried the Tang Band.

The stock T90A can sound a bit hard on the very top until crossover cap and diaphragm have aged quite a while.  I'm still really sensitive to that stuff though because of the still young age of my ears (perhaps? or its just a personal preference?) 

I was using the much higher end T90A-EX in my own system until recently.  Finally sold them last year along with my FE208ES-R drivers because my Feastrex's don't need a supertweeter.

Went back to visit an old customer who had some early speakers of mine that used T90A's and FF165k's  and they were working REALLY well together after six years in the field.  Sometimes the best tweak is to just leave your system alone, I guess!

The speed/dynamics of the T90 is very nice.

Cheers,

Clark

 

[Gerry E.]

Of course, a reasonable match of sensitivities is a good starting point.

The most widely successful high-pass crossovers for supertweeters seem to be either second order at 8-10kHz, or first order at 15kHz of higher. Below that crossover, the tweeter interference pattern with the main system becomes too obvious.

In fact, if the tweeter is much more sensitive than the woofer, sometimes it works to do first order very high, like 50kHz or even more, using that to adjust the sensitivity. I suspect this works because the tweeter actually does roll off, at least in power, even if the on-axis response is extended.

I do this to some degree.  My Jensen RP-302s tweeters are about 3 - 5dB more sensitive than the ALTEC 756Bs mid-range/woofers.  Therefore, I let the 756Bs run full-range and use a .47uf cap on the 16-ohm RP-302s.  That starts the roll-off at approx. 21,140hz.

I also tried .68uf (14,600hz), .56uf (17,740hz) and .33uf (30,100hz) but the .47uf was "just right".  Here's a question though - is raising the roll-off frequency a "legit" way to compensate for driver sensitivity differences?  In other words, would speaker design purists approve of this method?  Not that I need anyone's approval but I'm curious from an academic standpoint.  Thanks.

Gerry   

 

[Paul Joppa]

@Gerry, acadeically it's trash. But if you allow for both drivers rolling off in the treble (as well as your ears ...) then you can make an argument for it. No driver is really flat above 20kHz, IMHO - even if you can come up with a measurement that seems to be so, it won't work that way in the real world.