- Written by John E. Johnson, Jr.
- Published on 04 April 2011
- Pass Labs XA-100.5 Pure Class A Monoblock Power Amplifier
- Page 2: Design of the Pass Labs XA-100.5 Power Amplifier
- Page 3: The Pass Labs XA-100.5 Power Amplifier In Use
- Page 4: The Pass Labs XA-100.5 Power Amplifier On the Bench
- Page 5: Conclusions About the Pass Labs XA-100.5 Power Amplifier
- All Pages
Most amplifier manufacturers produce power amplifiers that are biased as Class AB, meaning that they run a few watts in Class A and the rest in AB. This is usually a push-pull configuration, and the AB moniker refers to the slight overlap in the + portion of the waveform that is delivered by one section of the amplifier with the - portion handled by the other amplifier section. This keeps some current flowing during the transition from the + to - and - to + portions without there being a point where both halves are completely off, which would produce a lot of distortion. During the Class A activity, both the + and - sections of the amplifier are producing the entire waveform.
The problem is that Class A operation produces a lot of heat dissipation when it is idle, and it is only about 30% efficient, so a 100 watt Class A amplifier is dissipating about 300 watts of heat. Because of this, the 100 watt Class A amplifier requires a massive power supply, and that transfers to EXPENSIVE. In a world where most consumers want maximum power for the least amount of money, Pass Labs bucks the trend by building Pure Class A power amplifiers. They are heavy, they are inefficient, they get hot, and they are mucho dollars, but there are customers who are willing to pay a premium for such a product. Pass builds a number of models, including the XA-100.5 monoblock reviewed here. Does it deliver a sound worth its price for just 100 watts? Read on . . . .
- Design: Pure Class A Solid State Monoblock Power Amplifier
- Power: 100 Watts RMS Pure Class A; 175 Watts Class AB (into 8 Ohms)
- MFR: 1.5 Hz - 100 kHz
- THD+N: 1% at Full Output
- Input Impedance: 30 kOhms
- Inputs: XLR and RCA
- Dimensions: 9" H x 19" W x 21.5" D
- Weight: 100 Pounds
- MSRP: $16,500/pair USA
- Pass Labs
The XA-100.5 is as unique in appearance as it is in sound. The front panel has a standby/on switch (main on/off toggle is on the rear panel), and a meter that reads straight up when the amplifier is on. This is a measure of the current flowing, not watts. It indicates that the amplifier is drawing current even though you may not be playing any music. As long as the meter points straight up, you are running in Class A. If you are playing music loud, and there is a demanding transient, the meter will move a bit to the right, indicating Class AB operation (up to 175 watts output into 8 ohms).
The rear panel has a pair of handles to lift the amplifier with (I would seriously suggest two people), an XLR balanced and RCA unbalanced input (pins 1 and 3 are shorted with a U pin that you remove if you want to use the XLR jack), two sets of speaker binding posts, a 12 volt trigger connection, and a grounding post. The amplifier acts as a bridged circuit, so both speaker jacks are hot, and you should never ground the - post. A diagram of the output is shown below.
The heat sink fins are mounted in an unusual manner, horizontally, angled upward. More typical fins are mounted vertically, which act like chimneys for the heat.
The amplifier uses 40 Mosfet output devices (transistors). That's a lot of output devices for just 100 watts. In a mass market receiver, you might find its 100 watt power amplifier (one of the five or seven channels) having only 4 output devices.
Most power amplifiers these days use bipolar output transistors, and there is a long story as to why Nelson Pass uses Mosfets, which you can read in some of his white papers. They are well written and easy to understand, so I recommend them even if you don't care why he uses Mosfets, because there are good explanations of Class A, distortion, negative feedback, and other things.
There are only two gain stages in the XA-100.5 and there is no global negative feedback. Pass operates under the premise that the less gain stages, the better, because each gain stage adds distortion, and then one usually adds negative feedback to reduce that distortion. Pass also designs the circuits to produce mostly odd-ordered distortion (predominantly 3rd order) rather than even-ordered distortion. This probably surprises most of you. It surprised me. But, it turns out that when double blind studies are performed, using amplifiers that produce either mostly 2nd order or 3rd order distortion, about 1/3 prefer 2nd order, 1/3 prefer 3rd order, and the rest either like neither or both.
Here is a photo of the inside of the chassis. You can see all the output devices lined up against the inside of the heat sinks on the sides.
It became very obvious that the XA-100.5 has huge overhead when I played the Telarc version of Copland's Fanfare for the Common Man. I could get the needle on the meter to move a little to the right, meaning that it was going into Class AB operation, but I didn't hear any clipping. The thunderous bass drum beats probably were drawing several hundred watts from the Pass amp during those short transients. But just think about it for a moment. The amp weighs 100 pounds and has 40 output transistors. That's 2-1/2 watts per transistor for the 100 watt output specification. If there is a short transient high demand, the Pass XA-100.5 can handle it. The bench test THD+N vs. Output is for continuous output, not peak demands. I suspect that this amplifier can easily deliver 400 watts for a short transient.
Stringed instruments work well for detecting IM distortion, which can make them sound very congested. This Telarc SACD of Grieg, Dvorak, and Elgar has plenty of violins, and through the XA-100.5, they were crystal clear. Easily distinguishable in their complex harmony. No harshness at all. That's a characteristic of Class A sound.
Mozart's Requiem in D Minor was unfinished (later completed by Franz Sussmayr), but the composition that Mozart did complete is still a masterpiece. It has a choir and soloists, and multiple human voices are another good test for congestion caused by harmonic distortion, because the voice is located in the lower midrange frequencies, so there can be several orders of audible harmonics. Nonetheless, the XA-100.5 withstood the onslaught of so many people singing. Soloists stood out from the choir's soundstage rather than being buried in mush.
Enough of this Requiem stuff (a Mass for the dead). Here is something for the living, and The Manhattan Transfer know how to live, or in this case, swing. Along with John Pizzarelli and Janis Siegel, The Cincinnati Pops have their hands full, but the XA-100.5's power meter remained rock steady in spite of a full orchestra and jazz musicians doin' their thing. This amplifier delivers tremendous dynamics and is lightning fast.
Over the course of listening to the XA-100.5, it became obvious to me that having plenty of Class A instead of just a few watts, really makes a difference when the music needs to be cranked up. There are many small Class A amplifiers, usually tubes, often using 300B triodes, and the sound is breathtaking, but you can't party without more power.
On the Bench
Distortion measurements were performed within an 80 kHz bandwidth, using a Carver Mark IV ribbon speaker as the load, except where indicated.
At 1 kHz, the 3rd an 5th harmonics were predominant, along with a small 7th harmonic. This is all odd-ordered distortion, which was the designer's (Nelson Pass) intention. I had always thought that odd-ordered harmonics would sound bad, but that is not the case, as long as they are primarily the lower numbered harmonics (3rd and 5th). In any case, the total amount of distortion was less than 0.01%.
Using 19 kHz and 20 kHz sine waves, the B-A peak at 1 kHz was 102 dB below the fundamentals. This is excellent.
IMD, using the SMPTE standard of 60 Hz and 7 kHz sine waves, was only 0.009%.
THD+N vs. Frequency was less than 0.1% out to 8 kHz, then rose to 0.2% at 20 kHz. The Impedance/Phase graph for the Carver Mark IV ribbon speaker is shown in the second illustration.
The measured frequency response was 20 Hz - 60 kHz, - 1 dB.
Power output at 8 Ohms was far above the specification. The sharp knee was at 150 watts, and clipping (1% THD+N) was at 175 watts. From 0 to 100 watts, operation is in Class A, and above that, it goes into Class AB operation.
At 4 Ohms, the sharp knee occurred at 250 watts, with clipping at 290 watts.
A marketing department would definitely have specified this amplifier as a 175 watt output amp (maybe even stretching it to 200 watts). Nelson Pass chose to remain conservative and specify it according to the amount of Class A ouput it has, which is 100 watts. Most of the time, we only use about 10-15 watts. It is only with the transients that a lot more power is needed, and the XA-100.5 has that; it's just not all Class A. The amplifier is characterized by low-odd-ordered harmonics, which to my surprise, sounded great. That is because Pass designed it with this in mind, instead of a poorly designed amplifier that has a lot of negative feedback which may flatten the frequency response but also results in higher-odd-ordered harmonics (9th, 11th, etc.), and those would degrade the sound quality. Let's call it "controlled distortion". A fine product from a brilliant designer.