Boulder 2050 monoblock amplifier Measurements
Following its 1/3-power, 1-hour preconditioning test, the Boulder 2050's heatsinks were hot but could be touched for 2-3 seconds without discomfort. All measurements were made in the Boulder's balanced configuration; there is no single-ended input, but the amplifier can be driven single-ended with external adapters. The voltage gain switch was left in the high position, as J-10 had left it.
The Boulder's input impedance measured 220k ohms, its output impedance 0.017 ohms at either 20Hz or 1kHz, this increasing to 0.021 ohms at 20kHz. The 2050's voltage gain is 25.7dB. The signal/noise ratio ref. 1W into 8 ohms measured 100.1dB over a 22Hz-22kHz bandwidth, unweighted. Over a 10Hz-500kHz bandwidth this worsened to 89.1dB unweighted (-103.3dB A-weighted). Pin 3 of the balanced input is connected as positive, not pin 2 as specified. DC offset measured approximately 4.4mV; though it fluctuated slowly, the voltage drift was never significant.
Fig.1 shows the 2050's frequency responses at 1W into 8 ohms, 2W into 4 ohms, and 2W into a simulated real load. The responses are virtually identical—within 0.05dB. The 1kHz squarewave response (not shown) is almost textbook-perfect. The 10kHz squarewave, shown in fig.2, is nearly so, modified only by a very slight rounding of the leading edge and a visibly finite risetime.
Fig.1 Boulder 2050, frequency response at (from top to bottom at 6kHz): 1W into 8 ohms, 2W into 4 ohms, and 2.828V into simulated loudspeaker load (0.5dB/vertical div.).
Fig.2 Boulder 2050, small-signal 10kHz squarewave into 8 ohms.
The THD+noise vs frequency curves are plotted in fig.3. Except for a common rise at higher frequencies—to percentage levels that are still insignificant—this is outstanding performance. The waveform of the distortion imposed on a 1kHz sinewave, shown in fig.4. Taken at 200W (!) into 2 ohms, this is dominated by a noisy third-harmonic content. Note that the sensitivity of the measuring equipment had to be set at maximum to take this reading, even at this high power output. Into higher impedances of 4 and 8 ohms, the waveforms (not shown) were barely recognizable, even at a 100W output. This latter result makes it clear that the THD+noise shown in fig.3 is dominated by noise, at least below 1kHz.
Fig.3 Boulder 2050, THD+noise vs frequency at (from top to bottom at 5kHz): 4W into 2 ohms, 2W into 4 ohms, 2.83V into simulated loudspeaker load, and 1W into 8 ohms.
Fig.4 Boulder 2050, 1kHz waveform at 200W into 2 ohms (top), distortion and noise waveform with fundamental notched out (bottom, not to scale).
The spectrum of the 2050's output reproducing 50Hz into a 4 ohm load is shown in fig.5. The output power here is 840W (!), and all of the artifacts are below -80dB (0.01%); in fact, the distortion is lower in level than the noise.
Fig.5 Boulder 2050, spectrum of 50Hz sinewave, DC-1kHz, at 840W into 4 ohms (linear frequency scale).
Fig.6 shows the amplifier's spectral response to a combined 19+20kHz signal—the intermodulation products resulting from an input signal consisting of an equal combination of these two frequencies—at a power level of 599W into 4 ohms. (The amplifier clips with this input signal above this level.) This is also a superb result, with, again, no artifacts higher than -80dB (0.01%). The result into 8 ohms at 378W (not shown) is as good or better.
Fig.6 Boulder 2050, HF intermodulation spectrum, DC-24kHz, 19+20kHz at 599W into 4 ohms (linear frequency scale).
The 1kHz THD+noise vs output power curve for the 2050 into 8 ohms is shown in fig.7. This requires a brief explanation. The clipping level shown here is under 700W. The actual clipping level was measured at 677W (28.3dBW). Why is this, when the amplifier is rated at 1000W? The limiting factor is not the amplifier, but the test procedure. We do not measure equipment with a rigidly controlled line voltage. (Those who do usually use a Variac.) Instead, we prefer to perform the tests the same way you would use the equipment: direct from the wall outlet.
Fig.7 Boulder 2050, distortion (%) vs continuous output power into 8 ohms. (The wall voltage was 103V, explaining why the Boulder didn't reach its specified power on this test.)
But the Boulder's current draw is so high that it caused our line voltage to sag from a measured 111V with no load to 103V with a single 2050 at 677W output into 8 ohms (1kHz). That's why the reading here is lower than the spec, and why we made no attempt to measure the maximum output into lower impedances. The limitations of our test equipment were also factors—at 677W into 8 ohms, the output voltage is nearly 74V. Take seriously Boulder's precaution on staying away from the loudspeaker terminals during heavy use!
I have no doubt that given a regulated wall supply, the 2050 will meet Boulder's power specifications. The 2050 is nothing less than a brute of an amplifier, but a brute with surprisingly refined performance. I have never had any other amplifier on the bench that measured better. It is definitely one of those products for which measurements are interesting but superfluous.
Still, I wonder about that class-A rating. While the 2050 got quite hot during the 1/3-power pretest, at no other time during the testing was it more than pleasantly warm to the touch. A thousand watts of class-A power should run hot all the time, even at idle. But even in the heat of a Santa Fe summer, I wasn't complaining.
Because of the 2050's huge need for power input, however, anyone contemplating buying a pair should very seriously consider setting them for 220V and driving them from a 220V outlet. No, make that two 220V outlets—one for each of the two monoblocks, each of them on completely separate circuits and wired for enough current capacity to make your electrician question your sanity.
Boulder does mention in their owner's manual that the 2050's performance may be limited by the power line, and with a 120V line this is certainly the case—at least measurably. (Whether or not you'll actually hear the difference between 677W and 1000W is a good question. Perhaps you will, briefly.) And look carefully to your loudspeaker's power-handling capability. High power with music shouldn't normally be a problem with any loudspeakers likely to be used with this amplifier, but be paranoid about all the usual speaker-busters, including plugging and unplugging leads with the amplifiers on. This isn't as big a problem with balanced connections, but just the same, be careful.
With the exception of the lower-than-expected wide-bandwidth S/N measurements—which were apparently not a factor in the listening tests—this is a superb set of test-bench results.—Thomas J. Norton