The SuperMicro-DX is my first attempt with the Peerless DX20BF00-04 tweeter. It is one of the few 3/4″ dome tweeters that extends down to 1kHz. Other features that attracts me is the low Fs of 900Hz and a damped impedance peak at resonance. And finally, it is loaded in a very short horn. This not only increases the sensitivity but more importantly, avoids the issues associated with baffle reflection.
Peerless DX20BF00-04 Frequency Response
Fig 1 – Black plot = RAW. Red plot = 2.6kHz (24dB/oct)
The Black plot in Fig 1 is the RAW response of the DX20BF00-04 surface mounted in the SuperMicro box with a baffle width of 5″. Microphone is at 24 ins, on tweeter axis. Impulse window at 5 msec, no smoothing applied.
The Red plot is the DX20BF00-04 with a 24dB/oct electronic crossover set at 2.6kHz. The resulting roll-of is close to a text book electrical filter. It’s very smooth, no bumps anywhere.
Tang Band W3-1364SA with Peerless DX20BF00-04
Fig 2 – Black plot = Summed Response. Blue plot = W3-1364. Red plot = DX20BF00-04
Fig 2 is the DX20BF00-04 crossed at 2.6kHz (24dB/oct) with the Tang Band W3-1364SA. As can be seen, the summation is excellent as there are no cancellations on either side of the crossover frequency.
SuperMicro-DX Frequency Response
Fig 3 – Summed Response of W3-1364 and DX20BF00-04 at 2.6kHz (24dB/oct)
For clarity, Fig 3 is the summed response of the SuperMicro-DX only. The overall response stayed within a +/- 2.5dB window.
Cumulative Spectral Decay of SuperMicro-DX
Fig 4 – Waterfall of SuperMicro-DX. Courtesy of REW.
Fig 4 is the Waterfall plot of the SuperMicro-DX. The Time Window is at 1.0ms, so this is a highly magnified view. The DX20BF00-04 is exceptionally clean. There are virtually no artifacts above 5kHz.
Toneburst Energy Storage of SuperMicro-DX
Fig 5 – Toneburst Energy Storage of SuperMicro-DX
The Toneburst Energy Storage in Fig 5 looks fabulous. It’s not often that there’s so little storage energy (light blue slices) from 1kHz~20kHz.
Spectrogram of SuperMicro-DX
Fig 6 – SuperMicro-DX Spectrogram. Courtesy of REW.
The Spectrogram in Fig 6 is a 2D representation of Fig 5. The blue slices from 1.5kHz~2kHz can be clearly seen on the left of the Spectrogram. The very minor slices from 4kHz~5kHz are also depicted.
We can conclude from these measurements that the SuperMicro-DX will reproduce music accurately and without coloration.
Harmonic Distortion of SuperMicro-DX
Fig 7 – Red plot = 2nd Harmonics. Violet plot = 3rd Harmonics
The Harmonic Distortion in Fig 7 doesn’t display any anomaly. The SuperMicro-DX is dominated by 2nd harmonics, which averages about -50dB.
Integrating the Bass
This is where I ran into some unexpected problems. Initially, I crossed the SuperMicro-DX with a direct radiating Usher 8137A at 125Hz. It is just passable, nothing outstanding. I then swapped the Usher with the Toucan, a bandpass sub using a Dayton SD215A-88. That didn’t work either. I replaced the Toucan with the Albatross, a Dayton DA270 bandpass sub. Again, something was amiss. No matter what crossover frequency I used, from 90Hz~250Hz, it just didn’t sound right.
I eventually threw caution into the wind. I re-adjusted my electronic crossover from a 3-way crossover to 2-way. In doing so, the W3 bass now rolls off naturally in a Closed Box alignment instead of 125Hz at 24dB/oct.
I then merged the Albatross using only the bandpass acoustic low pass. In other words, no electronic crossover is used to integrate the Albatross to the SuperMicro-DX. And what do you know, now it sounds right.
Fig 8 – SuperMicro-DX Closed Box Response
The plot in Fig 8 is that of the SuperMicro-DX in a active 2-way configuration. We can see the reflections in my room causing nodes, one peak at 200Hz and a deep notch at 150Hz. These are unavoidable.
What is more important is the acoustic roll-off. Since this is a Closed Box, the bass is rolling off gentler at 12dB/oct.
Fig 9 – Summed Response of Albatross Bandpass Subwoofer with SuperMicro-DX
Fig 9 is the full response with the Albatross added to the SuperMicro-DX. Though the notch and the peak are still there, the bass now extends down to 30Hz. But that is the easy part. The difficulty was getting enough SPL from 100Hz~150Hz. When I tried crossing the Albatross with an active crossover at 125Hz, I ended up with an ugly suck out. Because of that, I lost a huge amount of detail in the bass. It appears that for best integration, using the natural acoustic roll-off of the Albatross and the SuperMicro-DX is the way to go.
Peerless DX20BF00-04 Performance
As far as sound quality is concerned, I cannot find any fault with the DX20. The treble is not coarse or brittle. When used with the Tang Band W3-1364SA, the resulting sound is neutral – accurate. There’s no coloration, no excessive sibilance, no honking and no grating highs.
But this evaluation is not simply about how the DX20 sounds like. It is also about how well it integrates with the woofer. There are very few dome tweeters that rolls of nicely like the DX20BF00-04. The frequency responses that you see in the datasheets are all derived with the tweeters mounted onto a IEC panel which measures 1.35 meters x 1.65 meters. When the same tweeter is mounted onto a real speaker box with a baffle of 8 inches wide, you’ll find the response doesn’t look anything like in the datasheet.
Almost all dome tweeters suffer from this. But not the DX20BF00-04. Because of the short horn construction, the roll-off is exceptionally smooth. Actually, it looks very similar to horn loaded compression drivers.
Now that I’ve ascertain the DX20BF00-04 can be crossed at 2.5kHz without difficulty, I will be designing new speakers around the DX20BF00-04.
Note: Unless otherwise stated, all measurements were made with the mic at 1m, tweeter axis. Impulse Window=5ms. No smoothing applied.