Solid State Power Amp Reviews

To perform the measurements on the MBL Reference 9011, I used Stereophile's loan sample of the top-of-the-line Audio Precision SYS2722 system (see the January 2008 "As We See It" and www.ap.com). Before doing the testing, I ran the MBL, serial number 40431, at one-third its rated power for 30 minutes, which thermally is the worst case for an amplifier with a class-A/B output stage. (I usually do this test for 60 minutes, but my time with the review sample, serial no.40431, was limited.) At the end of that period, the chassis was only mildly warm. However, the third section of heatsinks from the front on the amplifier's right-hand side was hottish, at 115°F (46°C). (The other heatsink sections were cooler, at 90°F/32°C.)

The amplifier was tested as a bridged monoblock, as that is how MF used the pair of 9011s in his auditioning, with all three AC transformers active. The voltage gain into 8 ohms from the balanced input, assessed between the two positive terminal posts (one for each half of the balanced amplifier), was 27.1dB. The 9011 preserved absolute polarity when the positive speaker lead was connected to the uppermost binding post. The balanced input impedance was the specified 20k ohms in the bass and midrange, dropping slightly to 16k ohms at the top of the audioband.

The output impedance was very low, at 0.05 ohm at 1kHz and below, rising to 0.1 ohm at 20kHz. (This measurement includes 6' of speaker cable.) As a result, the variation in frequency response with our standard simulated loudspeaker was minimal (fig.1, gray trace). This graph also shows that, into higher impedances, the MBL 9011 has a very wide small-signal bandwidth, its response into 8 ohms (blue trace) dropping by just 1.65dB at 200kHz. The amplifier's reproduction of a 10kHz squarewave therefore has a superbly square shape with very low risetimes (fig.2).

The amplifier's ultrasonic output decreases into lower impedances, however, the response being –3dB at 85kHz into 2 ohms (fig.1, red trace). Even so, with an amplifier with a wide bandwidth like this, it is important to use interconnects that are correctly shielded. The amplifier might not go into overload, as happened with MF's TARA Labs and ZenSati interconnects, but the presence of RF interference can still have an unpredictable effect on sound quality.

When I measured the Ypsilon BC1 transformers Michael used to convert the VST-100 Mk.II preamplifier's single-ended output to balanced, these both broke the ground connection between preamp and power amps and had a frequency response that rose significantly above the audioband (+12.5dB at 181kHz). Both of these factors exacerbated the MBL 9011's possible susceptibility to picking up RF interference, particularly with the unshielded ZenSati interconnects.

Even with its wide small-signal bandwidth, the Reference 9011 was a quiet amplifier. Its unweighted, wideband signal/noise ratio, ref. 1W into 8 ohms, was 83.7dB, with all AC-supply–related components lying below 100dB. The A-weighted ratio, again ref. 2.828V into 8 ohms, was a superb 102.7dBA, equivalent to 130dBA ref. the measured clipping power into 8 ohms (1% THD+noise), which was 540W (27.3dBW, fig.3), almost 1dB greater than the specified 440W (26.4dBW). This amplifier has true 20-bit dynamic range. The amplifier clipped at 870W into 4 ohms (26.4dBW, fig.4), compared with the specified 840W (26.2dBW), but fell below its specified power of 1390W into 2 ohms (25.4dBW), at a measured 1190W (24.7dBW, fig.5). This is most likely because I don't hold the AC wall voltage steady during these tests. It was 122.1V with the amplifier idling, 121V with the amplifier clipping into 8 ohms, but 117.4V at clipping into 2 ohms.

To ensure that I was measuring actual distortion rather than noise, I examined how the THD+N percentage changed with frequency at 25V output. Even so, as shown in fig.6, the MBL 9011 was extremely linear. Into 8 and 4 ohms (blue and magenta traces, respectively), the rise in THD in the treble was minuscule, though it was more pronounced into 2 ohms (red trace). The distortion was predominantly third harmonic in nature (fig.7), though this was still very low in level (fig.8). Intermodulation distortion with the punishing equal mix of 19 and 20kHz tones, tested at a level about 3dB below visual clipping on the oscilloscope, was also low (fig.9). The difference product at 1kHz was at –120dB (0.0001%), though the higher-order products were up to 20dB higher in level.

The MBL Reference 9011 measures superbly, as I have come to expect from this German manufacturer. There doesn't appear to be any technical reasons why the 9011 would sound different from the Reference 9007, which, from my experience, sounds very much like Michael described it in his 2006 review. But owners should both take their dealers' advice on what cables are best to use with this very–wide-bandwidth amplifier and use it with preamplifiers having a true balanced output.—John Atkinson