LABORATORY TEST REPORT
Newport Test Labs measured the frequency response of the Castle Avon 4 Loudspeakers as 40Hz to 30kHz ±3.5dB which, although it’s just slightly outside the dB tolerance of Castle’s specification (which is 45Hz~20kHz ±3dB) more than compensates by extending both lower and higher than Castle’s claimed extension. Looking at the response itself (shown in Graph 1) you can see that were it not for the tweeter rolling off a little above 15kHz, the ±dB tolerance could have been much tighter, because the response between 45Hz and 16kHz is within ±2.5dB. (Note that the trace shown on this graph is the result of two different measurements being spliced together at 1kHz. The response below 1kHz is the averaged result of nine individual frequency sweeps measured at three metres using a pink noise test stimulus. The response above 1kHz was measured using a gated sinus technique that simulates the response that would be obtained in an anechoic chamber.)
The director of Acoustic Design at IAG Group Ltd, Peter Comeau, and his team have managed to get the frequency response of the Castle Avon 4 extraordinarily flat across the midrange, so that from 150Hz up to 4kHz, the frequency response is within ±1dB. As for that roll-off above 15kHz shown in Graph 1 (and in Graph 2), that could have been a microphone/ speaker positioning issue, because ribbon tweeters are highly directional. However, the roll-off occurs above the limit of hearing for most adult listeners in any case.
The high-frequency response of the Castle Avon 4 is shown in Graph 2 both with the grille fitted (red trace) and without it (black trace). You can see that the grille is truly acoustically transparent. The response is very slightly more linear without the grille, but the differences between the two traces are so slight that
I’d recommend listening with the grilles on all the time, to ensure additional mechanical protection for the drivers.
The low-frequency response of the Castle Avon 4 is shown in Graph 3. Note that the level of the transmission line’s output (red trace) has not been adjusted to compensate for the difference in radiating area between it and the bass driver. You can see that it starts delivering bass at 150Hz, and keeps delivering bass within 5dB of this level right down to 30Hz and is 7.5dB down at 20Hz. This makes it fairly obvious where the Avon 4’s bass extension comes from, and seems proof enough that Castle’s ‘twin drive transmission line’ is a genuine advance in transmission line technology.
The dual bass drivers have a very smooth response across their operating range, and their low-frequency output is extended, holding at almost reference level down to 70Hz before rolling off at around 12dB/octave. From 70Hz up to 260Hz the response is within ±1.5dB. The roll-off at 260Hz is due to the crossover point being at 280Hz. You can see the midrange driver kicking in at this frequency, so the electrical and acoustic crossover points coincide very neatly, suggesting that the drivers were either designed specifically for this design, or just extremely well-chosen.
Newport Test Labs’ measurement of the Castle Avon 4’s impedance is shown in Graph 4. As you’d expect, the impedance curve at low frequencies is atypical, due to the transmission line loading, with a single resonant peak at around 72Hz (12Ω) with a small ‘shelf’ at 28Hz at around 6Ω, after which the impedance drops to its minimum of 4Ω at 10Hz. In fact, except for the resonant peak, the impedance stays well below 8Ω right across the audio band until at around 12kHz it rises above this value and continues to rise thereafter—which makes the speaker very ‘amplifier-friendly’. This rise in impedance is due to the transformer required for the ribbon tweeter, so you can see the ribbon starts working at around 5kHz. Although Castle specifies the ‘nominal’ impedance at 8Ω, it’s able to do this because it also specs the minimum impedance (which it puts at 3.5Ω), which is kind of a loophole in IEC2685. Newport Test Labs’ tests put the minimum impedance at 4Ω and I would personally regard the nominal impedance of this design as being more like 4Ω, than 8Ω.
Newport Test Labs’ measurements put the sensitivity of the Castle Avon 4 loudspeakers at 87dBSPL at one metre using its standard test methodology. This is 2dB lower than Castle’s own specification (of 89dBSPL), but Newport Test Lab’s technique is tougher on speakers than the method used by speaker manufacturers and almost always results in lower figures, so 87dBSPL is a very good result and means you will be able to achieve very respectable sound pressure levels in your listening room without requiring too much amplifier power.
The Castle Avon 4s didn’t put a foot wrong in Newport Test Labs’ measurements, returning a superbly flat and extended frequency response, good sensitivity, and an easily manageable impedance modulus.
Overall, an outstanding example of good speaker design. Steve Holding