|
Flamingo-V The Flamingo-V is based on the Tang Band W6-789E. I first tested this woofer early last year. I didn’t follow up after that because it is one of those woofers that gives me nightmares. My Thiele & Small is not anywhere close to Tang Band’s specs. After procrastinating for almost 2 years, I went ahead and installed the W6-789E onto a 13 liters Bass reflex for one more try.
The plot in Fig 1 is the RAW response of the W6-789E. On the right, the cone breakup is visible, starting at 3kHz. Below 500Hz is a nearfield measurement. This is an approximation of what the woofer response is without my room reflections. Notice the peak at 150Hz. That’s caused by their runaway Thiele & Small. Tang Band’s Qts is 0.46. My Qts is 0.61. That’s a huge gap. By right, with this high Qts, I should be using a sealed box but since my test box is a bass reflex, I’ll start with it. If the peak is unbearable, I can always seal the port.
My initial approach was to use a 2nd order filter to suppress the cone breakup. But after thinking it over, I decided to bite the bullet and go for a 1st order all the way. If it works, it’ll be fantastic. The Blue plot in Fig 2 is the W6-789E with a 1st order network.
The Red plot in Fig 3 is of the Peerless H26TG45-06 horn loaded tweeter with a 1st order high pass filter.
As I had hoped, the two drivers crossed properly. There are no cancellations in the passband.
Fig 5 is the final frequency response of the Flamingo-V. The response is practically flat from 500Hz to 15kHz except for the sharp notch at 5.5kHz. That’s not from the tweeter. It’s caused by the cone breakup peak of the W6-789E.
The Step Response (Fig 6) threw up a surprise. The W6-789E has a very fast transient. At 60%, there’s a break, followed by another one at 95%. The apex is at 200 microsec. The Step also reveals my tweeter is wired in the opposite phase.
The Waterfall plot shows a clean treble. There’s a bit of artifacts at 5kHz. This is not from the Peerless tweeter. It’s actually from the cone breakup of the W6-789E.
The Toneburst plot in Fig 8 is another representation of the Waterfall. This time, the z-axis is in cycles. The light blue slices at 5kHz are those seen in the Waterfall plot.
The Spectrogram (Fig 9) shows no smearing at 5kHz. The treble in the Flamingo-V is actually very clean. Below 3kHz, there are two spots at 2kHz but they are already -50dB below the fundamental. There is a longish streak at 1.3kHz but it’s even weaker, about -65dB. This is quite an impressive Spectrogram.
The distortion in the treble is lower than normal. From 2kHz upwards, it’s -60dB below the fundamental. From 2kHz to 1kHz, the 3rd harmonics (Violet plot) dominates. During auditioning, I couldn’t hear these odd harmonics as they are still at -50dB below the fundamental.
The Flamingo-V is a very easy load for power amplifiers. The lowest impedance is 6Ω at 200Hz. This would be rated at 8Ω Nominal Impedance in most speakers. The electrical phase is well behaved from 200Hz to 20kHz. Sound of Flamingo-VThe Flamingo-V is quite a special speaker, well suited for Jazz and Vocals. The clarity in the midrange is impressive. No smearing, no veil. Vocals are crystal clear and focused. Treble contains no harshness or anything that irritates the ear. No aggressive sibilance. Diana Krall’s fans will love the Flamingo-V. I am listening to the Girl in the Other Room as I’m writing and it sounds like she’s right in front of me. And I’m not playing loud. Turning back the clock, I put on Cliff Richard. His cover of The Twelfth of Never is scary. You can make out clearly that he’s close miking. And that vibrato and vocal control. Absolute mastery. They don’t sing like this nowadays. I am really glad I finally worked on this Tang Band W6-789E. In all honesty, I never expected that I’ll like this woofer. But when she’s paired with the right tweeter, the speaker sounds wonderful. Lastly, the Flamingo-V’s sensitivity is quite high, about 91dB. This will appeal to chip amp and tube users. Unless otherwise stated, all measurements were made with the mic at 36 ins, tweeter axis. Impulse Window=5ms. No smoothing applied. |
November 21, 2020HIFI DRIVERS, Projects
Fig 1 – W6-789E RAW Response • Baffle Width=8-1/2″
Fig 5 – Flamingo-V Frequency Response
Fig 6 – Flamingo-V Step Response
Fig 7 – Flamingo-V Waterfall
Fig 8 – Flamingo-V Toneburst Energy Storage
Fig 9 – Flamingo-V Spectrogram
Fig 10 – Flamingo-V Harmonic Distortion • RED plot=2nd Harmonics • VIOLET plot=3rd Harmonics
Fig 11 – Flamingo-V Impedance