This review page is supported in part by the sponsors whose ad banners are displayed below

Good engineering extends to packaging. Here competitors could take lessons from the Germans. Once extricated from its oversize high-density cardboard box, that theme continued with the machine's innards. The full-metal remote wand with its twinned IR emitters obviously only enjoyed a quick photo op in space otherwise occupied by the single ECC81 phono board. With the external power supply, there isn't much to see since its transformer tucks beneath a shield.

The longer 6N6 valve is properly cradled to prevent the typical droop I've seen all too often in competing machines.

Here are the motorized Alps pot and infrared receiver.


Except for the off-boarded mains transformer, standard power supply duties like voltage regulation, filtering and smoothing remain on board.

Here we see the balancing transformers for the true XLR outputs segregated behind a metal partition.

The next two images show what occurs right behind the front panel on either side of the attenuator.


And here is the i/o board for the single-ended sockets on the other end of the deck.


Finally these are the two upright boards for the twinned XLR inputs. "The socketed TI INA 2134 differential summing opamps are used for cost reasons and because an input transformer represents relatively low input impedance to cause potential drive issues for certain CD players or other XLR-out sources, even more so as many deliver rather high voltages. It's far from casual to design a balanced input transformer which properly handles up to 6V at 20Hz. I thus decided against input transformers for cost and compatibility reasons, never mind that they would have taken up too much space to eliminate the phono option. Input transformers can also cause real problems with capacitor-coupled outputs which cause resonances between 5 and 50Hz depending on their capacitance value."


Back on tech for those curious about NFB and gain interaction: "High gain shows an amplification factor of x 18.5. This equates to 25dB and ~22dB of applied negative feedback. Medium gain amplifies times 8.8 which nets 19dB and applies ~28dB of NFB. For low gain the numbers are x 4, 12dB and ~35dB. The open-loop gain of the circuit is about 47dB or x 250. Its upper -3dB bandwidth is 25kHz. The following curves were generated with an Audio Precision at discrete spot measurements per frequency and with a 2V input signal to show real-world S/N behavior. The red line is the open-loop performance, the blue that of high gain. This demonstrates how the latter's noise across all frequencies is 22dB lower, i.e. directly proportionate to the 22dB of applied NFB.

"Open-loop noise between 10Hz and 16kHz is 430µV. High gain reduces this to 35µV whilst low gain is down to 7-8µV. These figures show how the circuit works perfectly within the linear region of the feedback loop as all curves track each other very closely and across the entire bandwidth. Stock the machine is fitted with a Tung Sol 12AT7 input valve whose amplification factor is about 35. One could optionally use a 12AU7/ECC82 to diminish the amplification factor to x 14. In low-gain mode—amplification factor of now just x 2—this would net a whopping -115dB of S/N unweighted, probably the absolute limit of what can be achieved with tubes. That's not a guess but actual number. The only other interesting thing would be an amplifier with built-in pot. That would simply need an additional buffer behind the volume control to offset the advantage of such a minimalist concept."