Columns Retired Columns & Blogs |
Antique Sound Lab Explorer 805 DT monoblock power amplifier Measurements
Sidebar 3: Measurements
Footnote 1: Every musical peak will be accompanied by a instantaneous burst of second-harmonic distortion that might well improve the sense of dynamics.—John Atkinson
Ah, I see Art Dudley is reviewing a single-ended tube amplifier. And he expects my measurements to trigger something along the lines of "This is the most perversely distorted amplifier I've seen in all my life, and if Art Dudley thinks it's good, he needs therapy." Okay, let's see if Mr. D. is in need of, if not therapy, then at least what professional amplifier user Keith Richard once called "heavy trank."
I ramped up the output tube bias as recommended by the Explorer 805 DT's manual, with the measurements taken at 1388mA bias current and after the amplifier been turned on for several hours. (I used the tubes that had been marked for each sample.) The Explorer offers a high 29.96dB of gain from its 16 ohm output transformer tap, this dropping to 28.8dB and 27.2dB from the 8 and 4 ohm taps, respectively (all figures measured at 1kHz into 8 ohms). All three transformer taps preserved absolute polarity, and the amplifier's input impedance remained close to the specified 100k ohms across the audioband. The wideband, unweighted signal/noise ratio (ref. 1W into 8 ohms from the 8 ohm tap) was a disappointing 54dB due to the presence of hum, which I could not eliminate with the top-panel trim pot. The S/N ratio improved to 69.6dB when A-weighted.
As expected, the Explorer's output impedances were high, at 7.2 ohms (16 ohm tap), 3.65 ohms (8 ohm tap), and 2 ohms (4 ohm tap). However, these figures were taken in the midrange and high treble; at 20Hz, the Explorer's source impedance was much lower, and varied according to whether I was loading the amplifier with 8 or 4 ohms. This interdependence suggests some sort of positive feedback arrangement, at least at low frequencies.
When we look at the Explorer's frequency response, therefore, we see not only the usual changes due to the Ohm's Law interaction between the load and the amp's source impedance—fig.1 shows the response from the 4 ohm tap, fig.2 the 16 ohm behavior—but a degree of peaking in the low bass. This ranges from 2dB to 6dB at 18Hz, depending on the output transformer tap and the load impedance. While too low in frequency to make the amplifier sound bass-heavy—"full" was how Art described it—this boost will make the speaker sound uncontrolled in the bass, as Art noted, and make the system prone to low-frequency feedback, again as he found.
Fig.1 Antique Sound Lab Explorer 805 DT (4 ohm tap), frequency response at (from top to bottom at 1.5kHz): 2.83V into simulated loudspeaker load, 0.5W into 16 ohms, 1W into 8 ohms, 2W into 4 ohms, 4W into 2 ohms (2dB/vertical div.).
Fig.2 Antique Sound Lab Explorer 805 DT (8 ohm tap), frequency response at (from top to bottom at 1.5kHz): 2.83V into simulated loudspeaker load, 0.5W into 16 ohms, 1W into 8 ohms, 2W into 4 ohms, 4W into 2 ohms (2dB/vertical div.).
Note also that the Explorer's high source impedance means that the response changes with my simulated loudspeaker load in these two graphs are large, ranging from ±1.5dB from the 4 ohm output (fig.1) to ±3dB from the 16 ohm output (fig.3). Such changes will be very audible, meaning that the amplifier's tonal balance will change according to the speaker with which it is used.
Fig.3 Antique Sound Lab Explorer 805 DT (16 ohm tap), frequency response at (from top to bottom at 1.5kHz): 2.83V into simulated loudspeaker load, 0.5W into 16 ohms, 1W into 8 ohms, 2W into 4 ohms, 4W into 2 ohms (2dB/vertical div.).
The amplifier's response starts to roll out above 16kHz, but what should have been a smooth rolloff is disturbed by the presence of ultrasonic resonances at 53kHz and 150kHz. The peaking due to these resonances depends on the output tap used and the load impedance, but is worst from the 16 ohm tap into low impedances. The subjective effects of these resonances are unpredictable, but they do mar what would have been excellent 1kHz (fig.4) and 10kHz (fig.5) squarewave responses.
Fig.4 Antique Sound Lab Explorer 805 DT, small-signal 1kHz squarewave into 8 ohms.
Fig.5 Antique Sound Lab Explorer 805 DT, small-signal 10kHz squarewave into 8 ohms.
The big specification for a single-ended amplifier is its output power. Antique Sound Lab claims 50W for the Explorer, at <8% distortion, presumably with the output transformer tap matched to the load. Figs.6, 7, and 8 show how the amplifier's THD+noise percentage varies with output power into 16, 8, 4, and 2 ohms. The amplifier's output distortion remains below our usual 1% limit between 100mW and 1W, with the lowest figure—0.5%—obtained from the 4 ohm tap into 16 ohms. It's fair to note that the measured percentage will include a contribution from the hum I noted earlier, and that the amplifier will most often be used in this power region, which goes some of the way toward explaining why Art wasn't driven from the room screaming. But I had to relax the distortion limit to 10% for the amplifier to deliver its specified power of 50W (17dBW), which it did when the load was half the nominal output tap.
Fig.6 Antique Sound Lab Explorer 805 DT (4 ohm tap), distortion (%) vs 1kHz continuous output power into (from bottom to top): 16 ohms, 8 ohms, 4 ohms, 2 ohms.
Fig.7 Antique Sound Lab Explorer 805 DT (8 ohm tap), distortion (%) vs 1kHz continuous output power into (from bottom to top): 16 ohms, 8 ohms, 4 ohms, 2 ohms.
Fig.8 Antique Sound Lab Explorer 805 DT (16 ohm tap), distortion (%) vs 1kHz continuous output power into (from bottom to top): 16 ohms, 8 ohms, 4 ohms, 2 ohms.
Regarding how the measured distortion varied with frequency, the curves for the 4 and 8 ohm taps are shown in fig.9 and 10. (The 16 ohm graph, not shown, is very similar, other than having correspondingly higher percentages of distortion.) The expected dependence of the THD figure on decreasing load impedance can be seen, but note also the catastrophically rising distortion below 100Hz. Like the reduced source impedance in the low bass and the corresponding LF boost in the frequency response, this indicates that the Explorer has something weird like positive feedback going on at low frequencies. I am surprised that AD was not bothered more by the quality of the Explorer's bass.
Fig.9 Antique Sound Lab Explorer 805 DT (4 ohm tap), THD+N (%) vs frequency (from bottom to top): 2.83V into 16 ohms, 8 ohms, 4 ohms, 2 ohms.
Fig.10 Antique Sound Lab Explorer 805 DT (8 ohm tap), THD+N (%) vs frequency (from bottom to top): 2.83V into 16 ohms, 8 ohms, 4 ohms, 2 ohms.
That he wasn't bothered more by the amplifier's generally poor linearity is due to the distortion being heavily second-harmonic in nature (fig.11). (The second harmonic lies an exact octave above the fundamental tone and so is musically consonant with it.) But as figs.12 and 13 show, a regular series of higher harmonics is also present in the amplifier's output, these decreasing more or less linearly with increasing order. Fig.12 shows that the residual 120Hz power-supply component lies at -56dB (0.15%), though the 60Hz hum component is at a more respectable -79dB (0.011%). As I noted earlier, I couldn't reduce the level of this hum any further. You can see in fig.12 that it results in each spectral component being surrounded by 120Hz-spaced sidebands.
Fig.11 Antique Sound Lab Explorer 805 DT (4 ohm tap), 1kHz waveform at 1W into 8 ohms (top), 1.1% THD+N; distortion and noise waveform with fundamental notched out (bottom, not to scale).
Fig.12 Antique Sound Lab Explorer 805 DT (4 ohm tap), spectrum of 50Hz sinewave, DC-1kHz, at 1W into 8 ohms (linear frequency scale).
Fig.13 Antique Sound Lab Explorer 805 DT (4 ohm tap), spectrum of 1kHz sinewave, DC-10kHz, at 1W into 8 ohms (linear frequency scale).
The low harmonic nature of the Explorer's distortion will be subjectively benign only if it is not accompanied by high levels of musically dissonant intermodulation distortion. Unfortunately, as is revealed by fig.14, the Explorer does have relatively high IM distortion. The 1kHz difference component that results from an equal mix of 19 and 20kHz tones driven at 1W into 8 ohms from the 4 ohm tap lying at -40dB (1%).
Fig.14 Antique Sound Lab Explorer 805 DT (4 ohm tap), HF intermodulation spectrum, DC-24kHz, 19+20kHz at 1W into 8 ohms (linear frequency scale).
I recommend that this amp be used from its 4 ohm output transformer tap with sensitive speakers, but even then, "broken" is the word most engineers would use to describe an amplifier that measures as poorly as did the Antique Sound Lab Explorer 805 DT. All I can assume is that the mainly second-harmonic nature of its distortion and the lowish powers that it would have been asked to deliver with Art's chosen speakers didn't get in the way of the music. Unless, of course, they enhanced the music (footnote 1). Pass me the heavy trank.—John Atkinson
Footnote 1: Every musical peak will be accompanied by a instantaneous burst of second-harmonic distortion that might well improve the sense of dynamics.—John Atkinson
- Log in or register to post comments