
The benefits of controlling resonant break-up modes are obvious and has been Scan-Speak’s “mantra” for years, leading to the development of the Revelator sliced cone technology and the oval voice coil Ellipticor technology. Now, the company is participating in the Composite Sound’s metamodal development program leading to the release of the new D3004/606200 1” TPCD dome tweeter featured in the July issue of Voice Coil and this month’s Test Bench explication — the new D3004/666000 AirCirc motor tweeter.
It is important to note that both of the new TPCD metamodal Scan-Speak tweeters are replacing Scan-Speak’s previous Beryllium versions. This is due to the skyrocketing cost of Beryllium, mostly caused by defense applications of Beryllium taking priority. For this reason, the US government has been stockpiling Beryllium for defense purposes in the event of a national emergency since 2016. Other than being toxic and carcinogenic, Beryllium is still an outstanding material for tweeter and compression driver diaphragms, so it is sad to see go by the wayside.
The feature set for the Scan-Speak D3004/666000 tweeter (shown in Photos 1-3) includes a new 26mm (1”) TPCD thin-ply carbon metamodel designed diaphragm, an AirCirc motor design with six neodymium slug magnets, a black anodized machined aluminum faceplate that can be custom laser engraved (company logo or system model number, etc.), a patented Symmetrical drive (SD-2) neodymium motor structure, a large roll coated cloth surround, 90W IEC 18.4 power handing (with 2.5kHz second-order Butterworth HP network), long-term 130W IEC 18.2 power handing, 91.5dB 2.83V/1m sensitivity, plus gold-plated terminals.



After completing the impedance measurements, I recess mounted the Scan-Speak D3004 TPCD diaphragm tweeter in a bookshelf-size enclosure that had a baffle area of about 11”×6” and measured the horizontal on- and off-axis at 2.0V/0.5m (normalized to 2.83V/1m) from 0° on-axis to 45° off-axis using the Loudsoft FINE R+D analyzer and the GRAS 46BE microphone (supplied courtesy of Loudsoft and GRAS Sound & Vibration). Figure 2 shows the on-axis response for the D3004, which exhibited a ±1.63dB response from 2.5kHz to 22kHz with response out to 40kHz (±1.93dB from 2kHz to 40kHz).




Last, Figure 6 gives the two-sample SPL comparison showing the two Scan-Speak D3004 samples to be closely matched within ≤1dB throughout the driver’s operating range from 3kHz to 18kHz.

Next, I used the Listen, Inc. SoundCheck V21 software and AudioConnect analyzer and SCM ¼” microphone to measure the impulse response with the tweeter recess mounted on the same test baffle. Importing this data into the Listen, Inc. SoundMap software produced the cumulative spectral decay (CSD) waterfall plot given in Figure 7. Figure 8 depicts the Short Time Fourier Transform (STFT) displayed as a color-variegated surface plot.


For the final test procedure, I set the 1m SPL to 94dB (4.4V for the D3004/666000), using a pink noise stimulus, and measured the second and third harmonic distortion at 10cm. The SoundCheck distortion graph is depicted in Figure 9, demonstrating very low third harmonic content.

One of the current problems for high-end driver manufacturers is the ridiculously high cost of Beryllium diaphragms. All the OEMs I have talked to recently tell me that they have gone from about $30 for a 1” Beryllium tweeter dome to $160 for the same dome, which in the long run is probably unsustainable. Focal is the only speaker manufacturer actually fabricating its own diaphragms, and since it is processing raw material, it makes the company less dependent on Materion Corp. While the timbre of TPCD and Beryllium are not identical, the comparison is pretty favorable, so it is no surprise that more and more high-end driver OEMs are releasing the new metamodal TPCD diaphragm transducers. This new incarnation from Scan-Speak is an excellent example of the new high-resolution diaphragm format and exhibits Scan-Speak’s attention to detail and build quality.
For more information, visit www.scan-speak.dk. VC
This article was originally published in Voice Coil, January 2025