What can I say? A great little amplifier! So cool! And so different from any other, current or former. Brought to the market in 1984, the new Musical Fidelity A1 is only 68 mm high, but as such a far cry from the 1960s/70s flat style of B&O’s Jacob Jensen. The A1 was designed by the Pentagram design studio, founded by Alan Fletcher, Colin Forbes and Kenneth Grange. Grange is remembered, for example, as the designer of B&W’s futuristic loudspeakers. Obsessed with aluminum gray, B&O would never ever have made its amp black and blue, or left the cover plate looking like an unfinished waffle iron, not even if it had a clear technical function, as it does in this amplifier.
The stripped-down front panel features a power switch, input selector and volume control. As it was, no tone controls, no channel balance control, no headphone output. No point, other things being equal, in recycling the audio signal through control circuits. The input selector and volume control are recessed into the front panel so that they can only be turned from below – a la QUAD 33 preamp. There is a small slanted tab on the front panel for that.
IMITATION OR ORIGINAL
Musical Fidelity claims to have copied the technical performance values and circuitry directly from the 1980s version, so that “even those who didn’t get to join in the first round can now enjoy the famous amplifier”. In this particular case, tinkering with the circuitry designed by Tim de Paravicini could have meant financial suicide. Moral suicide for sure.
When Leak (IAG) released remodeled Stereo 130 and 230 amps with a DAC and Bluetooth, it also turned out that the original germanium-based 2 x 15 Watt output stage had been turned into 45/75 W power packs with a new output stage topology. If Musical Fidelity (Audio Tuning) had gone the way of Leak with A1, it would have taken away everything that makes the A1 A1, and ultimately the reason for re-releasing the amp. Instead Musical Fidelity has made some changes especially to the A1’s power supply, but says it has kept the input and output stage, as well as the circuit layout, historical, which is enough to preserve the amplifier’s original identity.
Which version of the original? Musical Fidelity is a bit vague on this. It refers to the 1985 Mk I version, but elsewhere mentions the 1988 version. The 1990 MkII or Mark 3 would also have been suitable as models, but the 1991 Final Edition with its power MOSFETs not, let alone the 2008 “all-new” A1.
FINE-TUNING THE POTENTIOMETER’S OPERATION RANGE
The new A1 uses Texas Instrument TL071 op amps for the input stage, up from a TL084 in 1984. There are two of them per channel, meaning that the first operational amplifier before the ALPs RK series potentiometer (originally Alps Blue), not only sets the input impedance, but also amplifies the signal. The gain can be bypassed with the front panel normal/direct switch, which originally served as a tape/monitor switch on a tape recorder.
Direct mode cuts the input voltage gain by -10dB (32dB), making it easier to match the potentiometer’s operating range with a high-output digital source or sensitive speakers. Normal mode (42dB) is better suited for a turntable, FM tuner, or insensitive speakers. There is still a line-level tape out for the recorder, but unlike the pre-out, it is fixed: the preamp’s volume control has no effect.
In addition to the five 300mV/25kOhm RCA line inputs, A1 sports a RIAA input for MM and MC cartridges. According to the manufacturer, the signal first meets a discrete current-form input stage (low input impedance and high output impedance), and a current-to-voltage converter as a voltage source, which also takes care of the RIAA frequency EQ. The voltage gain on the MC side is now higher (70dB), and the input impedance (nominal 1 kOhm) is automatically matched. A short test with Columbia and Bang & Olofsen test records, a Lenco L75/Ortofon/Denon DL-103 as the testbench, proved that the A1’s MC input is technically and sonically a solid performer.
Originally, all four OP amps of the A1’s preamp stage were powered by the ±12 V rails of the right channel power amp stage. The new A1 has a new, more efficient, dual mono-wound mains transformer, and with it completely independent, twice as efficient left and right channel power supplies for the output stages. The filtering and regulation of the main power supplies have all been further improved.
NOT THE POWER LOSSES, BUT THE IDLE POWER
The symmetrical and discrete bipolar topology of the original A1 output stage has been retained unchanged. Two identical, common-emitter Motorola 2N3055(npn)/MJ2955(pnp) output stage pairs per channel have been replaced by Planar Audio output devices. The type is not disclosed, but according to Musical Fidelity, the technical values are close to that of the original. The output power has remained the same: 25 W in pure Class A into 8 ohms (36 W into 4 ohms), and a hefty 25 A (pk-pk). Other sources give 28 W and 20 A.
In pure Class A? Well, that’s printed on the front panel. The sceptics have come up with figures like 8 watts in Class A into 8 Ω (quiescent current of 700 – 800 mA), after which the amp, then, with a higher power, or with a lower load impedance, slides into Class AB/B. Tim de Paravicini famously disagreed. In his mind the total current was double, and the output stage only worked when the amplifier continuously drew a certain power from the wall (80 W) regardless of whether it was at idle or driven at full power, which strongly suggests genuine Class A biasing.
The amp’s manual states that the amplifier is “heavily biased into Class A”, and calls A1’s operating class “dynamic” Class A, meaning the design of the output stage allows for temporary automatic additional current consumption when the Class A bias is exceeded, but this is said to happen not by suddenly switching into Class B. Instead the amount of bias current varies dynamically according to changes in the load. The amp automatically adjusts the amount of bias current to keep the amplifier operating in Class A, but does not use more current than necessary. This resembles Threshold’s famous Stasis amp and its non-switching output stage design, where biasing the output stage forces both transistors of the pair to stay on, even if the continuous bias current is lower.
The history of solid-state amplifiers knows only a few single-ended Class A implementations, such as the Sugden A21 and some by Nelson Pass. Class A phase shifters are more common, but only a handful exist that have genuinely sought different biasing methods to solve the problems related to transistor switching on and off, or have seriously tried to reconcile high power losses and high bias. Whether Class A biasing is a shortcut to happiness in terms of sound quality is another matter.
PROPER VENTILATION
Either way, Class A biasing produces a hot-running amplifier, and this latest version of the A1 is no exception, despite the fact that heat dissipation has been enhanced by increasing the size of the cover plate and by careful component selection. When the amp is on, the cover plate is barely touchable. Otherwise, operating the A1 was like coming home. There are five 300mV/25kOhm line inputs to choose from. Moving the knobs from below requires a little practice. Despite the manufacturer’s countermeasures, the switching peak is still there, so it is wise to turn the volume down before switching the power on/off. The same applies to the MC/MM input selector on the rear panel. The manual recommends leaving 50 cm space behind and 25 cm on top, which given the heat, is a recommendation worth taking seriously.
Speakers? Despite its modest output power by today’s standards, the A1 has a reputation as an efficient amplifier that can handle even more difficult speaker loads. The claim is not entirely unfounded. The A1 showed no clear signs of weakness when I hooked it to a 91 dB/8 ohm full-range speakers nor with a 87 dB/6 ohm two-way bookshelf speakers, and, somewhat surprisingly, not with Quad’s 85 dB/8 ohm ESL63 electrostatics.
SMOOTH BUT NOT SOFT
Did I say that I had an A1 in my system in the 1990s? The A1 MkII, to be more specific. My loudspeakers, at that time, were 86 dB/4 ohm Gradient 1.3, which are famous for not spitting on the output power. For some reason, if my auditory memory serves me well, this new A1 gave the impression of being even stronger than the one I had.
This amp, unlike many others, knows how to make music rhythmic and lively, and reveal music’s dynamics almost regardless of the genre, and note: also at a slower tempo, which I find important. Very pulsating and physical, indeed. The difference is like between a good and excellent drummer. No delays, no hesitation with the attacks. The Direct mode seemed to add some focus to the sound, sharpen the contours and make it easier to locate sound sources within the sound stage.
Tim de Pavaricini is known to have said that, with his Class A circuitry he aimed for a tube amplifier kind of sound. But how should this be understood? That the sound of the A1 is somehow warm or colored? In my opinion, the A1 did not sound particularly warm (nor did the MkII in the 1990s), not at least in the sense that the frequency response would have appeared to roll off too early or because of a woolly mid-bass. Yes, there was something velvety about its tone, but the smoothness or softness, so to speak, did not prevent the A1 from sounding bright, clear and open when such was required. When I replaced my A1 with VTL’s 50-watt push-pull Class A tube amp, but the tone of the sound remained surprisingly similar, which I thought was due to VTL’s KT90 output tubes ( KT90 pentode cares as much about the output power as about tonal subtleties), so perhaps the A1’s sounded like a KT90 tube amp.
MUSICAL FIDELITY LS3/5
That there is something special about the A1’s sound was this time perhaps best revealed when, towards the end of the review, I asked the A1 drive Musical Fidelity’s new BBC LS3/5 (or should I say Falcon) loudspeaker. The tweeter is a 19 mm Mylar dome, and the woofer a 5 inch Bextrene unit. The announced frequency response is 80 Hz – 20 kHz +/-3dB, sensitivity 82,5dB (W/m) and rated impedance 15 ohms. It’s not notoriously hard to drive a loudspeaker but only gives 83dB at one watt, with max SPL 95dB, and that means that it surely benefits from amplifier power. The A1 provided the required kick for the LS3/5, but more importantly, it made the LS3/5’s midrange bloom in a rare fine way, of which eg. vocal music greatly rejoiced. It sounded as if both were born to be together from birth.
Technical specifications
Musical Fidelity A1
Type: integrated stereo amplifier
Output power: 2 x 28 W into 8 ohms, 2 x 36 W into 4 ohms
Frequency response: 10Hz – 40 kHz (+ 0, –1dB)
THD + N: < 0.05% (< 0.01% A-weighting)
Channel separation: 85 dB
SNR: 82 dB
Input impedance (line): 25 kOhm
Input sensitivity (nominal): 300 mV rms
Analog inputs: 5 x rca, 1 x phono
Analog outputs: 1 x tape out, 1 x pre out
Dimensions (W x D x H): 440 x 283 x 68 mm
Weight: 10.5 kg
Price: €1599
