Balanced Audio Technology VK-40 preamplifier Measurements

Sidebar 3: Measurements

All measurements of the BAT VK-40 were performed on serial number 040800072—the sample that had given RD some problems with the mono switch—with some spot checks on sample '71. No appreciable differences were noted. All tests of the line stage were done in full balanced mode (balanced input to balanced output), with some tests repeated for the unbalanced line input. The phono-stage measurements, of course, were performed using an unbalanced drive signal.

Starting with the preliminaries, the balanced input impedance was a very high 244k ohms, while the unbalanced input impedance was a still high 112k ohms—source components will hardly be loaded by the VK-40. The phono-stage input impedance, however, was lower than specified, at 5.2k ohms at 1kHz. The maximum line-stage voltage gain was the same for both balanced and unbalanced sources, at 21.5dB. This was with the volume control set to its maximum position of "140" on the numeric display. The unity-gain setting appeared to be "100" (actually an insertion loss of -0.06dB), and the volume-control steps were a sensible 0.5dB, except at the extreme settings.

As reviewed, the phono stage was set up for MC cartridges; its voltage gain was a very high 79.8dB including the line-stage gain. Despite that, its unweighted wideband S/N ratio was a still good 64.5dB, this improving only slightly to 68.6dB when A-weighted (both figures referred to an input level of 500µV at 1kHz and with the volume control set to "140").

The output impedance was a fairly low 404 ohms across most of the audioband, rising to 4.9k ohms at 20Hz due to the finite size of the output coupling capacitors. (The caps used by BAT in its output-coupling "Six-Pack" are of very high quality but large in physical size, which limits the maximum capacitance, hence their impedance at low frequencies.) As a result, the response rolls off prematurely into a 600 ohm load (fig.1, lower traces below 1kHz). This is admittedly an unrealistically low load, but it emphasizes the fact that the VK-40 needs to be used with power amplifiers having an input impedance of at least 50k ohms if the lows are not to sound a little lean, even with the "Six-Pack."

Fig.1 Balanced Audio Technology VK-40, frequency response (from top to bottom) at volume-control settings of "100" into 100k ohms, "100" into 600 ohms, "140" into 100k ohms (1dB/vertical div.).

To the right of fig.1 is shown the ultrasonic performance of the VK-40's line stage under three conditions. The most extended response, reaching -1dB at 80kHz, is with the volume control set to "100" (unity gain). Dropping the load to 600 ohms increases the ultrasonic rolloff to -1.5dB at 80kHz, which is a negligible change. But increasing the volume control to "140," its maximum setting, drastically curtails the HF response, the rolloff reaching -6dB at 20kHz. Fortunately, it is extremely unlikely that the preamplifier will be used above its unity-gain setting, so I don't think it's this behavior that lies behind RD's finding the preamp to sound a little veiled. The line-stage response behavior was identical for unbalanced sources.

Fig.2 shows the overall frequency response taken via the phono stage with the volume control set to "100." Any error in the RIAA correction is superbly low, and the overall response is very flat up to the measurement limit of 100kHz. Channel separation via the line inputs (fig.3) was excellent, though the increase in crosstalk with increasing frequency is due to capacitive coupling—perhaps in the volume control, given the physically separated nature of the circuits for the two channels.

Fig.2 Balanced Audio Technology VK-40, phono input RIAA error into 100k ohms at volume-control setting of "100" (0.5dB/vertical div.).

Fig.3 Balanced Audio Technology VK-40, channel separation (10dB/vertical div.).

Assuming an MC cartridge output of 500µV at 1kHz, the VK-40's phono-stage overload margins were generally excellent, at 29dB at 20Hz and 28dB at 1kHz. The margin dropped to 19dB at 20kHz, however, which is still good considering the high gain of the phono stage.

Fig.4 shows how the line-stage THD+noise percentage changes with increasing output level into two loads: 100k ohms (bottom trace above 1V) and 600 ohms. Below approximately 1.5V into 100k ohms and 750mV into 600 ohms, the measured figure is dominated by noise, which is why it increases with decreasing output level. There is a steadily increasing amount of distortion above those voltages, but it doesn't reach 1% until very high output levels that are many times what the VK-40 would be required to deliver in real-world conditions. And the preamp doesn't actually hard-clip until it's pumping out 30V or so into the higher load!

Fig.4 Balanced Audio Technology VK-40, distortion (%) vs output level (V) into (from bottom to top at 1V): 600 ohms, 100k ohms.

At 1.5V, the level at which the measured distortion is at its minimum into 100k ohms, the THD remains constant across the audioband, though the right channel was slightly worse than the left (not shown). This is due, I believe, to a slightly higher level of noise in that channel. Fig.5 shows a spectral analysis of the BAT's output while driving a 50Hz sinewave at 1.6V into 100k ohms. The only significant harmonics present are the third and the fifth, though "significant" is probably not the appropriate term to use for something almost 90dB down from the level of the fundamental (0.003%). And it should be noted that no power-supply spuriae can be seen in this graph, which is excellent.

Fig.5 Balanced Audio Technology VK-40, spectrum of 50Hz sinewave, DC-1kHz, at 1.6V into 100k ohms (linear frequency scale).

Into the punishing 600 ohm load (fig.6), the third harmonic rises to -80dB (0.01%), but the preamp otherwise maintains its aplomb. These two graphs were taken with a balanced input signal. With an unbalanced source into 100k ohms, the third harmonic lay at -80dB (not shown).

Fig.6 Balanced Audio Technology VK-40, spectrum of 50Hz sinewave, DC-1kHz, at 1.6V into 600 ohms (linear frequency scale).

Perhaps partly because of its enormous dynamic range, the VK-40 featured extremely low levels of intermodulation distortion. Even into the 600 ohm load, the 1kHz difference component was below -96dB, or 0.0014% (fig.7). However, you can see some spectral lines in this graph that are not mathematically related to the 19kHz and 20kHz input signals (19.3kHz and 9.65kHz, for example). Yes, these are extremely low in level, but perhaps they are due to breakthrough from the VK-40's control circuitry.

Fig.7 Balanced Audio Technology VK-40, HF intermodulation spectrum, DC-24kHz, 19+20kHz at 1.5V into 600 ohms (linear frequency scale).

All in all, its measurements reveal the BAT VK-40 to offer very low noise and a distortion signature that is both low in level and subjectively benign, coupled with enormous dynamic range.—John Atkinson

Balanced Audio Technology
800 First State Blvd.
Wilmington, DE 19804
(302) 999-8855