MANUFACTURER'S SPECIFICATIONS
Power Output: 100 watts per channel, 8 ohm loads.
Rated THD: Less than 0.01 percent at 1 kHz and 100 Hz, and less than 0.05 percent at 10 kHz.
Frequency Response:-1 dB at 20 Hz, -0.5 dB at 20 kHz, and-3.0 dB at 50 kHz.
Signal Input Level: 0.5 V, ±0.5 dB for rated output into 8-ohm loads.
Signal Input Slew Rate Limit: 0.1 V/ pS.
Signal Input Overload: Instantaneous recovery up to +20 dB overload.
Hum & Noise: 95 dB below rated output "A" weighted, 90 dB below rated power output unweighted, 15.7 kHz measured bandwidth.
Stability: Unconditionally stable with any load under any signal.
Power Requirements: 110, 120, 130, 220, 230, 240 V, 50/60 Hz, 30 to 350 watts (depending upon signal level).
Weight: 20 lbs. (9 kg).
Dimensions: 13.4 in. (34 cm) W x 4.5 in. (11.4 cm) H x 7.7 in. (19.6 cm) D.
Price: $480.00.
Leave it to our British friends to stubbornly continue to publish performance specifications for audio products that do not conform to other internationally recognized standards. Mind you, I happen to like the sound produced by the Quad 405, and I think it is a very worthwhile and well-designed product, having innovative and meaningful circuit approaches, but I'm not altogether certain that our Federal Trade Commission would be happy about that rating of power output which appears at the beginning of this report. The makers of the Quad 405 have not specified power bandwidth in terms acceptable to the FTC. Clearly, however, what they are telling us is that the amplifier will not deliver 100 watts at the 20 kHz extreme, and in fact it does not. Rather than derate the amplifier in its most useful frequency range, they simply tell us what it will deliver at three frequencies (100 Hz, 1 kHz, and 10 kHz) and at what THD levels. For that matter, it won't quite deliver a full 100 watts per channel at the low-frequency 20 Hz extreme either, but more about that later. I know that the British audiophiles (and equipment testers) believe that we Americans and our colleagues in Japan are a bit daft with all our emphasis on wide-band design and our love affair with distortion figures approaching "absolute zero," and perhaps in this regard, they are quite correct. The slew rate spec is also one to look at closely, since it is for the input signal, rather than the amplifier's output as is usual. However, the amp's slew rate can be calculated from the figure given, 0.1 V per microsecond, since the unit's gain is 35 dB, and thus the 405's slew rate spec turns out to be 5.62 V/ NS, a moderate figure these days.
Fig 1-Simplified schematic showing Class A, dumpers, and bridge components.
Fig. 2-Schematic diagram of the Quad 405 amp.
Fig. 3-Distortion vs. power output.
Fig. 4-IHF IM products (low level peaks) generated when 14 and 15 kHz signal
mix is applied at rated output level.
(Linear sweep from 0 to 20 kHz is employed, rather than a log sweep.)
Fig. 5-IHF IM products generated when 19 and 20 kHz signal mix is applied
at rated output level.
In any case, the Quad 405, pictured here, carries no controls of any kind and may therefore be mounted out of sight, providing adequate ventilation is provided. Speaker output connections are made in the usual manner, by means of spring-loaded speaker terminals. As for the inputs, a special cable is provided which is terminated at one end in a 4-pin DIN plug and at the other with a pair of standard phono-tip plugs which can be directly connected to the output of one's associated preamplifier-control unit. If a matching Quad 33 control unit is used, that product comes equipped with a DIN-to-DIN cable, in which case the adaptor cable supplied with the Quad 405 can be put aside. A 5-ampere fuse (for 120 volt supplies) is accessible from the outside of the unit, while four additional fuses (one in each leg of the positive and negative power supply lines of each channel of the amplifier) are wired internally.
Internally, the amplifier consists of two identical circuit boards, one for each channel of amplification. Each of these boards has a pair of spring-loaded wiring terminals into which may be inserted a pair of supplied 1.8 kilohm resistors which will limit the output of the amplifier to a maximum of 20 volts rms per channel (50 watts across an 8-ohm load). The manufacturers tell the user that this "power limiting" scheme should be used if the amplifier is to be used with their well-known Quad electrostatic speakers or with any other speakers that are rated at less than 100 watts for maximum power input.
The chief circuit innovation in the Quad 405 is called a current-dumping output circuit; a simplified schematic diagram of this output circuit is shown in Fig. 1. In a current dumping amplifier there is both a low-powered, high quality amplifier which essentially operates at all times in Class A, plus a high-powered, heavy duty amplifier. The low-powered amplifier controls the loudspeakers at all times, calling upon the high powered section to provide high-power signals when required. The small amplifier is so arranged (it carries an error signal) so that when the larger power transistors (the current dumpers) get within the target area of the required output current, it will fill in the remainder of the signal current accurately and completely. According to the makers of the Quad amp, the reproduced sound quality is solely dependent upon the smaller powered amplifier which can be made very good indeed. In addition, this circuit arrangement is said to solve problems of crossover, crossover distortion, quiescent current adjustment, thermal tracking, and transistor matching. There are, in fact, no internal adjustments or alignment points within the amplifier and, once again according to the manufacturer, the choice of power transistors becomes less restrictive than in conventional Class B amplifier designs. A complete schematic diagram of the Quad 405 is reproduced in Fig. 2.
Laboratory Measurements
Harmonic distortion of the Quad 405, when measured with a 1-kHz input signal, remained extremely low at all levels up to rated output. In fact, it was necessary to calculate the actual harmonic distortion using our spectrum analyzer, since, at low levels, the THD was buried well below the noise floor. The most accurate reading we could come up with was 0.0012 percent, which is essentially the distortion of our signal source. The rated THD level of 0.01 percent (for 1 kHz) was reached at an output of 102.5 watts per channel, increasing to 0.05 percent at an output of 112.9 watts per channel, for 8 ohm loads. SMPTE IM Distortion measured approximately 0.0065 percent for all power output levels below rated output. These results are shown graphically in Fig. 3.
We decided to use the new IHF IM measurement techniques which involve analysis of difference-frequency components generated when the amplifier is subjected to signals which are 1 kHz apart. Using 14and 15kHz input signals adjusted to provide the equivalent of rated output from the amplifier, only two significant IM components were generated (at 13 and 16 kHz) as shown in the spectrum analysis photo of Fig. 4. These were some 68 dB below the input signal level, which represents an IM percentage of 0.04 percent. When the two signal frequencies were shifted up to 19 and 20 kHz, a great many more "sideband" or IM products were generated, as illustrated in Fig. 5. Calculating the rated IHF IM distortion by formula for this "worst case" condition yields a figure of approximately 0.09 percent.
Figure 6 is a plot of distortion versus frequency, with the amplifier delivering 100 watts per channel into 8-ohm loads.
At 10 kHz, we reduced input signal very slightly to find the power output that could be delivered with 0.01 percent harmonic distortion. The value of power determined was 97.3 watts. The ideal load for the Quad 405 (in terms of maximum power output attainable) seemed to be 8 ohms, unlike that of most conventional amplifiers which deliver somewhat greater power output with 4-ohm loads. With a 4-ohm load connected, maximum power output for a distortion level of 0.01 percent (harmonic) for the Quad 405 was reduced to approximately 75 watts per channel, using a 1-kHz test signal.
Damping factor at 50 Hz (referred to an 8-ohm load) measured 91.5. Dynamic headroom for short-term signal bursts measured 1.34 dB above rated continuous power output. Frequency response was flat within ±1.0 dB from 19 Hz to 26 kHz, while the-3 dB roll-off points were observed at 13.5 Hz and 48 kHz. The amplifier has a built-in, sub-sonic filter rolling off low frequencies in accordance with the curve shown in Fig. 7. Maximum power consumption of the amplifier (when delivering 100 watts per channel, both channels driven) was 450 watts.
Fig. 6-Distortion vs. frequency.
Fig. 7-Low frequency filter response.
Listening and Use Tests
At low and medium power output levels, the sound reproduced using the Quad 405 could not be faulted and, in fact, was so good that it made us wonder about the need for ultra wideband response in power amplifiers so frequently endorsed by many American and Japanese amplifier manufacturers. When we attempted to push amplifier output towards rated output, but still well below the clipping point, we did begin to notice some deterioration in sound quality, specifically in the extreme treble region and when the amplifier was subjected to program sources containing sharp musical transients.
In summary, if one needs around 50 watts per channel of maximum power (and that should be no problem with many of today's higher-efficiency speakers), the Quad 405 amplifier does very well indeed. Based upon American measurement techniques and design philosophy, however, we would be reluctant to recommend the Quad 405 for those applications where a "strong" 100 watts per channel of amplification is required.
--Leonard Feldman
Editor's Note:
Acoustical Mfg. commissioned an independent laboratory to conduct a series of blind A/B listening tests based on their belief "that it is perfectly possible to design amplifiers that do not degrade the quality of the program (and introduce) no detectable change in quality." The tests generally appear to be well designed and had, as listeners, well-known experts in this field. Both a 50 percent probability test and a Chi' test showed that the preferences expressed by the panel "were no more than would be achieved by sheer chance." Further discussion of such testing will be found in the review of the AGI 511A preamp published in June, 1978.
-E.P.
(Source: Audio magazine, Apr. 1979 )
Also see:
Quad ESL 63 Loudspeaker (Jun. 1985)
Quad Component System -- One-Brand Systems (Dec. 1983)
= = = =