|
Flamingo2 – W6-789E with BC25SC08-04
Active Bi-amped
The Flamingo2 is meant for those who demand something better than a full range W6-789E. In this 2-way, I opted to use the Peerless BC25SC08-04 mini horn tweeter mainly because I will be crossing it at 3.5kHz. It would be a huge waste to use a $50 tweeter because I’m not utilizing the lower frequencies.
However, this sub $20 tweeter is not without limitations. While I can’t fault it’s performance when crossed at 3.5kHz, this is not a tweeter for passive crossovers. I shall explain further.
Peerless BC25SC08-04 Frequency Response
Fig 1 – Black plot = RAW. Blue plot = 2uF
The Black plot in Fig 1 is the RAW response of the Peerless BC25SC08. So far so good. The Blue plot is with a first order high pass, a single 2uF. Note the tweeter doesn’t roll off nicely. Worse than that, there’s a sharp peak at about 1.7kHz.
This peak is caused by the high impedance at the tweeter’s resonant frequency. To eliminate it, I had to use a conjugate (LCR) network. Add in the component cost of a third order filter, you’ll end up the passive crossover costing more than the tweeter itself.
If I’m going to spend so much money for a tweeter crossover, I should at the very least use a tweeter that cost just as much, if not more. This is one of the problems facing budget tweeters and woofers. Try to save on crossover parts and you end up compromising on the performance.
Is there a solution? Yes, run it active. No crossover cost. The money saved is better used for an active crossover and an additional power amplifier. More so when the amplifier is for the tweeter. A LM3875 or a TDA7293 chip amp is very cheap.
 Fig 2 – BC25SC08 Black plot = RAW. Blue plot = 2uF. Red plot = 3K5 (18dB/oct)
The Red plot in Fig 2 is of the BC25SC08 crossed actively at 3.5kHz (18dB/oct) with my Synergy Bi-amped kit. Any active crossover can be used. It’s out of convenience that I use the Synergy because I designed it specifically for speaker designs.
See the difference between the Red plot and the Blue one. It will cost a lot to get the Red plot with a passive crossover. Just look at the crossover with the SuperMicro.
Tang Band W6-789E at 3.5kHz LP
 Fig 3 – Black plot = RAW. Blue plot = 3.5kHz (18dB/oct)
Now that the Peerless BC25SC08 is out of the way, let’s turn our attention to the Tang Band W6-789E. The Black plot in Fig 3 is the RAW response whereas the Blue plot is with my Synergy active crossover at 3.5kHz (18dB/oct) low pass. The entire treble hump has been rolled off. On top of that, the crossover slope is a nice straight line.
Tang Band W6-789E and Peerless BC25SC08-04 Crossover
 Fig 4 – Blue plot = W6-789E at 3.5kHz (18dB/oct) LP. Red plot = BC25SC08-04 at 3.5kHz (18dB/oct) HP.
Fig 4 shows the W6-789E and BC25SC08 crossing over acoustically at 3kHz. Note the roll-off slopes of both the woofer and the tweeter. They are straight and symmetrical.
Flamingo2 Frequency Response
 Fig 5 – Summed Response of W6-789E and BC25SC08-04
Fig 5 is the Frequency Response of the Flamingo2. Measurements below 500Hz are in Nearfield. We can see from this that the Peerless BC25SC08 tweeter is now doing the treble. How much difference does this make.
REW Cumulative Spectral Decay of Flamingo2
Fig 6 – Waterfall plot of Flamingo2
The Waterfall plot in Fig 6 shows frequencies above 4kHz have less artifacts. A 1ms time line is a highly magnified view. In less than 0.77ms, the actifacts are already down by -30dB.
ToneBurst Energy Storage of Flamingo2
Fig 7 – Light Blue slices are Excess Energy
The Blue slices in Fig 7 are the Excess Energy of the Flamingo2. The bulk of them are from 3kHz~6kHz. That happens to be where the crossover resides. The question is, are they coming from the tweeter?
ToneBurst Energy Storage of Peerless BC25SC08-04
 Fig 8 – Peerless BC25SC08-04 at 3.5kHz (18dB/oct) HP.
Fig 8 is the Excess Energy of the Peerless tweeter crossed at 3.5kHz at 18dB/oct. There are some Blue slices but not much. In other words, the additional Blue slices in Fig 7 are from the W6-789E. They are probably from the treble hump.
REW Spectrogram of Flamingo2
Fig 9 – Wavelet of Flamingo2.
The Flamingo2 Spectrogram in Fig 9 shows a clean tweeter response. Frequencies above 4.5kHz have virtually no artifacts. Below that, from 4kHz~1kHz, they die off after 6ms. For all intents and purposes, they are not a problem. You’re not going to hear anything at 6ms.
Harmonic Distortion of Flamingo2
Fig 10 – Red plot=2nd harmonic. Violet plot=3rd harmonic.
The Harmonic Distortion (Fig 10) of the Flamingo2 is very similar to the W6-789E in Full Range. There’s still a slight 3rd harmonic hump from 1kHz~2kHz. Distortion above 3kHz is largely very low at about -55dB below the fundamental.
the Flamingo2 in use
Is there a difference adding a Peerless BC25SC08-04 tweeter. Yes, definitely. No doubt about it.
The Flamingo2 is for listeners that value sound quality. The vocal clarity is still there but the treble now has a certain sweetness that was absent before.
This doesn’t mean the Flamingo is inferior. Not at all. I will have no hesitation installing the Flamingo in retail outlets like boutiques and restaurants. It will provide high quality music at minimal cost. With it’s high sensitivity, more cost savings can be achieved by using chip amps.
For a passive 2-way version, I will have to change the Peerless tweeter to something superior. Otherwise, it’s hard to justify the crossover. Another project perhaps?
Note: All measurements were made with the mic at 1m, tweeter axis. Impulse Window=5ms. No smoothing applied. |
