This Harbeth model gets a very nice write-up from HR in the new S'phile. It looks in many ways to be a solid implementation of classic BBC design principles, but I'm puzzled by its use of a 22mm supertweeter crossed in above 10kHz. I think this feature is based on another decades-old BBC model, but this is the ONLY current Harbeth model that has such a thing.
My question is, why bother? Because the center-to-center distance from the main 26mm tweeter is necessarily greater than one wavelength at this frequency, and because of additional phase error in the crossover, the measured response above 10kHz is a nasty-looking mess, both on and off axis. It hurts my ears just to look at it!
It's not like it's necessary to resolve the top octave, either. SEAS makes (and Harbeth uses) a number of 26mm tweeters that extend flat well beyond 20kHz. With the high, 3.5kHz crossover from the woofer to the main tweeter, an off-axis dip followed by a flare is already inevitable, but may be an acceptable tradeoff for keeping all vocal fundamental frequencies coming from the same driver. This may result in greater coherency.
But I can make no such case for the use of the supertweeter. Any hypotheses on why Harbeth chose this design approach, and what advantages it might offer?
This post is made possible by the generous support of people like you and our sponsors:
Topic - Why supertweet? (Harbeth SHL5+XD) - Brian H P 12:17:05 08/14/23 (30)
- RE: Why supertweet? (Harbeth SHL5+XD) - Krav Maga 23:11:29 08/17/23 (2)
- Update: - Krav Maga 00:41:27 08/26/23 (1)
- My understanding, from multiple sources, is that the smaller tweeter is of titanium. - MWE 09:52:02 08/26/23 (0)
- RE: Why supertweet? (Harbeth SHL5+XD) - BCR 05:36:50 08/17/23 (0)
- RE: Why supertweet? (Harbeth SHL5+XD) - MWE 18:53:46 08/14/23 (0)
- Supertweeters add a sense of "shimmer" or "air" to the sound. - peppy m. 17:40:53 08/14/23 (1)
- Or "edge" or "shrillness" or "grain," if that's what's on the recording - Brian H P 18:56:28 08/14/23 (0)
- RE: Why supertweet? (Harbeth SHL5+XD) - BCR 16:10:37 08/14/23 (0)
- Tweeter, Not Supertweeter - Inmate51 14:14:43 08/14/23 (3)
- RE: Tweeter, Not Supertweeter - Tre' 17:16:37 08/14/23 (2)
- Got chuckle on "ferrofluid cooling tweeter" isnt that White Van marketing - Edp 22:32:00 08/14/23 (0)
- RE: Tweeter, Not Supertweeter - Inmate51 19:03:25 08/14/23 (0)
- "The BBC Heritage"??? - John Marks 12:59:43 08/14/23 (6)
- RE: "The BBC Heritage"??? - hahax@verizon.net 21:06:57 08/14/23 (5)
- RE: "The BBC Heritage"??? - Pat D 16:53:44 08/15/23 (3)
- RE: "The BBC Heritage"??? - hahax@verizon.net 21:31:10 08/15/23 (2)
- RE: "The BBC Heritage"??? - Pat D 17:29:55 08/17/23 (1)
- RE: "The BBC Heritage"??? - hahax@verizon.net 08:26:12 08/19/23 (0)
- That would seem to cover the subject quite well. nt. - MWE 05:34:43 08/15/23 (0)
- The graphs are painfull,but most listeners wont hear it - Edp 12:41:26 08/14/23 (10)
- I wonder if that upper-octave graph would look cleaner on a different measuring axis. nt - MWE 15:44:58 08/19/23 (1)
- See JA's horizontal/vertical off-axis plots. It doesn't. (nt) - Brian H P 15:16:29 08/22/23 (0)
- RE: The graphs are painfull,but most listeners wont hear it - hahax@verizon.net 21:11:08 08/14/23 (6)
- Waveforms - Inmate51 11:46:02 08/15/23 (5)
- RE: Waveforms - hahax@verizon.net 21:35:31 08/15/23 (4)
- RE: Waveforms - Inmate51 07:38:36 08/16/23 (3)
- RE: Waveforms - hahax@verizon.net 21:30:16 08/16/23 (2)
- RE: Waveforms - Inmate51 05:12:25 08/17/23 (1)
- RE: Waveforms - hahax@verizon.net 08:27:27 08/19/23 (0)
- Which again raises the question, why do it? - Brian H P 13:30:00 08/14/23 (0)