ABX Double Blind Test Results: Power Amplifiers

ABX Home ABX Hardware ABX Data ABX Literature
SMWTMS Tests PSACS Tests

SMWTMS ABX Test Results
Power Amplifier ComparisonResultCorrect p less thanListeners
10 Watt Tubes vs. Dyna 400* Different 63 / 80 = 79% 0.000510
Paoli 60M vs. Dyna 400* Different 78 / 126 = 62%0.0059
Swartz 40 vs. Dyna 400* Same 41 / 84 = 49%-6
Tiger B vs. Dyna 400 Same 43 / 86 = 50%-9
Dyna Mk III vs. Bose 1800 Same 55 / 107 = 51%0.42311
Crown SA-2 vs Phase Linear 400 Same 113 / 253 = 45%-17
ARC D120 vs CM Labs CM914a Different 82 / 90 = 91% 0.00053

* These were the first three audio double blind tests ever done.
For the first three comparisons of Power Amps, the 10 Watt Tubes vs. Dyna 400, the Paoli 60M vs. Dyna 400, and the Swartz 40 vs. the Dyna 400, the speakers were RTR Electrotstatics.
The Swartz 40 is a 20 Watt per channel power amplifier with quasi-complimentary ouputs built exactly to the five transistor per channel design in the RCA Transistor Manual. By the espoused standards of subjectivists, nothing sould sound worse than this amplifier.
The Tiger B vs. Dyna 400 and Dyna Mk III vs. Bose 1800 were compared on Ohm F speakers. Source equipment included a Shure V-15-III in an SME arm on a Thornes TD-125 turntable, and a Heath AP-1615 pre-amp.
For the first four tests, the musical selection was the Harry James Sheffield Direct Disc, The King James Version. On the Dyna Mk III vs. Bose 1800 the music was from Cat Stevens. The Crown SA-2 vs Phase Linear 400 comparison used music from the first Sheffield Direct Disc by Thelma Houston, the Telarc Firebird, the Sheffield Lincoln Majorka volume II, Paul Simon's Rhymin' Simon, Loggins & Messina Best of Friends, and Leo Kottke.
The smallest Magnepans with KEF B139 Subwoofers on a passive crossover were used for the Crown SA-2 vs. Phase Linear 400 test. The source equipment for this test was Shure V-15-IV, Dual 1019, and a Dyna PAT-5 with a Walt Jung Mod.
For the ARC vs. CM Labs, a Janus cartridge in a Magnepan Unitrac arm on a Linn Sondek LP12 turntable fead an MX10a Audio Standards headamp driving a CM Labs CM901a preamp. The speakers were Acoustat MK121-2 full range electrostatics. These speakers required a great deal of power and the Audio Research D120 was unstable when clipped, which proved audible. For more detail on this test see the Carlstrom, Krueger, and Greenhill article listed on the ABX periodicals page.
In the comparison of the 10 Watt tube amp vs. a Dyna 400, two mono non-identical 6V6 push-pull tube amps were paired for left and right channels. The better tube amp was a home brew with an honest 10 Watts and no controls. Its mate for the day in the second channel was a Heathkit that was always shy of its rated 7 watts and had tone controls which were set as flat as possible. Its frequency response curve was not bad but less than flat:
Graph: Heathkit A7 Frequency Response
In spec talk it might be called flat +/- 0.9 dB over the 20 Hz to 20 kHz range. Passable, but this is not good enough to pass an ABX test.
Different The different symbol means it is unlikely the percent correct score occurred by chance and thus the null hypothesis was disproven, which substantiates a real difference in sound quality.
Same The same symbol means that it is likely the correct score occurred by chance, the null hypothesis is not rejected, which means no difference was found. It does remain possible a difference may be substantianted with further testing.

PSACS ABX Test Results
Power Amplifier ComparisonResultCorrect p less thanListeners
B & K ST140 with all film caps
vs. Parasound HCA 800II		 Same 31 / 66 = 47% -4
Sumo Andromeda vs. Parasound HCA 800II Same 49 / 102 = 48% -10
Sumo Andromeda vs. Parasound HCA 800II
Long Term		 Same 5 / 10 = 50% -1

The 100 watt per channel B & K ST140 had all its electrolytic capacitors, except the power supply filters, replaced with film types, plus other modifications.
The Parasound HCA 800II is a totally stock 80 watt per channel amplifier.
The long term comparison of the Sumo Andromeda to the Parasound was done by the Sumo's owner after a five week warm up.
Send questions or comments on the PSACS tests to Tom Nousaine.
Different The different symbol means it is unlikely the percent correct score occurred by chance and thus the null hypothesis was disproven, which substantiates a real difference in sound quality.
Same The same symbol means that it is likely the correct score occurred by chance, the null hypothesis is not rejected, which means no difference was found. It does remain possible a difference may be substantianted with further testing.
ABX Home ABX Hardware ABX Data ABX Literature
Send questions or comments on the ABX Web Page to David Carlstrom.
Web Page Created by David Carlstrom. Last Modified: 1/26/2011