- Published on 06 July 2009
- Pass Labs XA30.5 Stereo Power Amplifier
- Page 2: Design of the Pass Labs XA30.5 Stereo Power Amplifier
- Page 3: The Sound of the Pass Labs XA30.5 Stereo Power Amplifier
- Page 4: The Pass Labs XA30.5 Stereo Power Amplifier On the Bench
- Page 5: Conclusions About the Pass Labs XA30.5 Stereo Power Amplifier
- All Pages
What do a 30 watt stereo solid-state amplifier and a 500 watt solid-state monoblock have in common? A lot, it turns out. The seemingly diminutive Pass Labs XA30.5 weighs 75 pounds, has an enormous toroidal power transformer, and uses 20 output transistors to deliver 30 watts of pure class A power. It turns out these specs are similar, or in some cases, better than my recently reviewed Emotiva XPA-1 monoblocks. The XA30.5 proves that rated power is by no means everything. In fact, if you have speakers of even moderately high sensitivity, the XA30.5 renders any discussion of power output irrelevant. When paired with the right speakers, this may be one of the best amplifiers there is, regardless of price.
- Design: Two Channel Power Amplifier
- Power Output: 30 Watts RMS x 2 into 8 Ohms, Class A
- MFR: 1.5 Hz - 100 kHz
- Sensitivity: 0.77 V for Full Output
- THD+N: 1% at 1 kHz, Full Output
- Input Impedance: 30 kOhms
- Damping Factor: 150
- Slew Rate: 50 V/uS
- Power Consumption: 200 Watts
- Dimensions: 7" H x 19" W x 19" D
- Weight: 75 Pounds
- MSRP: $5,500
- Pass Labs
Nelson Pass is a legend in the field of audio electronics. He started Threshold in the 1970's and has been responsible for dozens of landmark audio amplifiers since then. He holds over half a dozen patents for audio amplifier circuits. He also is active in the DIY audio arena, publishing many of his designs for audio hobbyists to build on their own. Pass has always been a proponent of class-A amplifiers, where the bias current through the output device is high enough that the transistor is always conducting. These designs operate the transistors in their optimal, most linear operating regime, but require a tremendous amount of power and dissipate a large amount of heat for a given power output. This means the whole amplifier has to be massively overbuilt as opposed to a typical class AB design where a much smaller bias current is used.
In a sense, a class AB amplifier and a class A push-pull amplifier like the XA30.5 are the same. A class AB amp is designed to operate in class A for the first few watts of output power, then transition to class B where one transistor handles the positive part of the waveform, and another the negative. The "class A" portion is the small overlap between the two. Class A amplifiers, if of a push-pull design with a pair of transistors (or several pairs) handling the positive and negative portions of the waveform, simply have a much larger overlap region. They're designed to operate with both transistors fully capable of amplifying the entire signal, both positive and negative voltage swings, over a large part of their power output.
Pass Labs, the latest company headed by Nelson Pass, has built both single ended class A amplifers and push-pull amplifers of both high bias class AB design and class A design (similar except for the level of transistor bias). The single ended amps begin to clip when they reach their class A limit, but the push pull class A amps can keep going. Pass Labs currently offers both a series of high bias class AB designs (the X series, with rated powers of 150W to 1000W per channel) and the XA series amps, which deliver their rated power output entirely in class A.
Both series of amps make use of at least three of Nelson Pass' patents. Both use the "supersymmetry" input state, the "Aleph" output stage and a single ended class A constant current bias supply. The supersymmetry input state is a patented fully differential amplifier topology that uses the cancellation offered by the balanced design to eliminate the need for large amounts of feedback to reduce noise and maintain stability. This is a key Pass technology that allows the design of low feedback amps with lots of power. The supersymmetry input stage drives a pair of Aleph output stages, one for each leg of the balanced signal. Each Aleph output stage also has a single ended constant current bias supply, so that the first watt or so of output power is delivered in single ended class A. This topology is an absolutely zero compromise approach to audio amplifier design. All of Pass' current designs use this topology. They differ in the number of output transistors needed to achieve the rated power output, and the amount of bias on the transistors.
The XA amps are all biased to dissipate three times their rated power output at idle. The X series amps are biased to dissipate about 1.5 times their rated power output. All XA amps are rated by the power they deliver into class A, not their power at 1% THD. The XA amps all deliver about 20% the power of the similarly sized X series amp (30W vs. 150W, 60W vs. 300W all the way up to 200W vs. 1000W. This is somewhat misleading, though as all the XA amps will deliver about three times their class A power in class AB mode at 1% THD. The XA30.5 is the smallest of the XA series amps, and is outwardly very much the same as the X150.5 amp. The XA30.5 is biased to deliver 30W of class A power, while the X150.5 is biased to deliver the first 5W RMS in class A. The XA30.5 (as we'll later see) hits the 1% THD point at 100W RMS, while the X150.5 can make it to 150W RMS. In this sense you trade more class A power for less maximum power.
One point here that Nelson Pass made when I spoke with him at the beginning of the review: There is no "clunk" or any other audible sign when any of these amps transition from class A to class AB. It is a completely seamless transition both in terms of measurement and audibility. Once the amp does transition into class AB, the distortion spectrum will begin to change subtly, with higher order distortion harmonics beginning to increase. Even this change is not sudden, though. In all cases, the Pass low feedback designs have their distortion biased toward low order harmonics. When operating in class A, high order distortion is further reduced. Unlike many high feedback amps, the distortion versus power output curve rises very gently as the 1% THD point is reached, rather than rocketing up as the amp begins to clip like with high feedback designs. This allows Pass amps to operate much closer to their rated power output while maintaining excellent sound quality.
The XA30.5 is quite an amazing piece of equipment. The amp weighs 75 pounds, even more than one of my 500 W Emotiva XPA-1s. The toroidal power transformer is huge. I'm not sure what its power rating is, but it looks as large as the 1200VA transformer in the Emotiva. The XA30.5 uses 20 output transistors, 5 for each leg of the balanced output. The 500W Emotiva uses 24. The biggest XA amp, the XA200.5 monoblock uses 80, yes eighty, output transistors per channel. The transistors run at only about 20% of their rated output power, so they never are pushed hard. Inside the amp, an input board located at the rear of the amp contains the supersymmetry input stage. The Aleph output stages (4 per XA30.5) are mounted on boards on either side of the chassis, bolted to enormous extruded aluminum heat sinks. The floor of the amp houses the power supply PC board and the toroidal transformer. A blue-lit current meter that measures the bias current through the devices dominates the beautifully machined front panel. While operating in class A, this meter doesn't move. When the amp transitions into class AB, it will begin to wag a bit as more current is drawn.
The rear panel houses two pairs of binding posts for the speakers (that do not accept banana plugs, by the way). An third binding post pair is used for a 12V trigger input. Power is supplied through a standard IEC jack, with a main power toggle switch. Thankfully, a pair of very sturdy handles are mounted on either side of the rear panel for carrying the amp. A front panel switch toggles the amp between run and standby. It takes about 30 minutes for the amp to warm up to the point that the bias current stabilizes. The heat sinks stabilize at about 50 degrees Celsius in about 1 hour. Every bit of this amp is top notch, with no compromises in design or execution. It certainly looks like a $5500 product.
The XA30.5 is the finest amplifier I have had in my listening room. This was actually quite surprising to me. While I expected the amp to be fabulous, I thought the low power output would handicap it when deployed in my room with my speakers. The Gallo Reference 3.1s have 88 dB/W/m sensitivity. While this is not low sensitivity, it's not particularly high either. From the first instant I listened to the XA30.5, it was clear this assumption was incorrect. In retrospect, this should not have been a surprise. Most music (even at pretty high volume) happens at well under 30W. Only bass transient peaks when listening at high volume caused the current meter on the Pass amp to move at all. When the needle did move, it only moved by a small amount. This says that the Pass amp was almost always operating in class A, tipping into class AB only occasionally.
The Pass amp has stupendous current delivery capability, with its huge transformer and loafing output devices. Large-scale dynamics, when compared to the 500W Emotiva XPA-1 monoblocks, were only very slightly diminished and then only on hardcore, bass heavy electronica and rock. This was the only place where the XA30.5 wasn't clearly superior to the XPA-1s (which are admittedly less than half the price of the XA30.5). On bass heavy electronica like Daft Punk's Human After All and Crystal Method's Vegas, the extension, control and power was there, but the sharp leading edges of transients produced by the more powerful amplifiers were rounded off a bit. In terms of microdynamics, the XA30.5 was superior. The aliveness of the XA30.5 was unmatched in my experience. The XA30.5 retained the energy and pace of the Emotivas on every album I listened to, but did so with superior smoothness and imaging.
The reality delivered by the Pass amp was its greatest asset. This was due to a combination of the great microdynamics delivering that "alive" sound, combined with stupendously three dimensional and well-defined images. Each image in the soundstage was amazingly round and deep when compared to the typically flat delivery from most amps. Overall soundstage width, and particularly depth, was also the best I've heard from the Gallos in my room. Width was about 1 foot wider on each side than the XPA-1s. The depth, which is mainly a result of spatial acoustic cues being clearly extracted from recordings, was absolutely insane. With recordings that had the information in them, like several of my Acoustic Sounds purist blues recordings, the depth was such that some of the soundstage sounded to be so far behind the speakers that it was in my front yard. On highly engineered recordings, the flatness was still there. An amp can't put in information that has already been engineered out. Even with these flat recordings (which are most of them), the three-dimensional nature of the images was still there, they were just all arranged in a single plane. An example of this was Gomez's Bring it On, with beautifully round images, but all in one plane just behind the speakers.
One thing class A amplifiers are known for is an incredibly liquid, clear and smooth midrange. The XA30.5 did not fail to deliver. The absolute clarity of the midrange combined with a complete absence of grain, glare or hardness was addictive. Unlike some tube class A amps, there was no trace of added warmness, just complete clarity and transparency, but delivered with a smooth-as-silk lack of grain. My favorite examples were the simple vocals of Allison Krauss and Ralph Stanley on the SACD recording of the O Brother Where Art Thou soundtrack. With the XA30.5, these simple unaccompanied vocals were just stunning. Some class A amps also have limited high frequency extension, but not this amp. I think some of these preconceptions are the result of single ended class A vacuum tube amps, which have some major technical limitations.
The solid state XA30.5 has no such limitations, with high frequency extension flat to well past 50 kHz. The high end also has stupendous smoothness and liquidity, giving every recording a much-improved sense of smoothness and "ease," meaning a lack of grain and glare at the high end. This is the stuff that can give you a headache after listening for a long time, or feels like a "pressure" on your ears. You'll only know it was there when you can take it away. The XPA-1s could not match this smoothness and ease. This was another area of the XA30.5 performance that benefited every recording. While they had the extension, accuracy and precision, the low order distortion spectrum of the XA30.5 gave it a major edge in the reproduction of the midrange and treble.
In the bass, the XA30.5 delivered admirably well-controlled and tonally accurate bass, but lacked a bit of the sharp edged impact of the more than 15 times more powerful XPA-1. This rounding off of the dynamic impact in the bass was the only limitation of the XA30.5, and this was compared to an amp that excels in this area. With even a slightly more sensitive speaker (I would say anything over 90 dB/W/m), the XA30.5 would catch up. For less sensitive speakers, there are plenty of XA amps with higher powers, from 60W to 200W class A power. These are all monoblocks, and much more expensive than the XA30.5, though (starting at over $11k).
On the Bench
I measured the XA30.5 using my Roland Edirol UA-101 24 bit 192 kHz USB soundcard with Spectra Plus FFT analysis software. Measurements were made using the balanced input of the XA30.5, running into both 8 ohm and 4 ohm loads. Measurements presented here are into the 8 ohm load for simplicity. Results were qualitatively identical into the 4 ohm load.
The first three plots show the distortion spectrum of a 1 kHz sine wave at 1 watt, 15 watts and 30 watts RMS power outputs. The THD+N levels are measured at 0.08%, 0.04% and 0.13% respectively. More important is the distortion spectrum. Particularly at the 15 watts and 30 watts output levels, the distortion is dominated by the 2nd and 3rd harmonics. High-level harmonics are at least 10 dB down from the low order harmonics. This is the key to the midrange magic of class A amps. As I've said before, THD is an almost meaningless measurement. What's important is how that distortion is distributed in frequency space. Another clear indication of the XA30.5's overbuilt design: At the rated power output of 30 watts, the amp is nowhere near the 1% THD level where power output is usually defined. In fact, the XA30.5 put out about 110 watts of power into 8 ohms before it hit 1% THD+N.
THD+N versus frequency measured extremely flat, rising only moderately up to 48 kHz. This measurement was made at 15 watts RMS power output.
Frequency response was perfectly flat to 48 kHz as far as I could tell. The gentle rise towards 8 kHz then the gradual falloff towards 40 kHz is the frequency response of the soundcard, not the amp.
Intermodulation distortion products between 1 kHz and 100 Hz sine waves are more than 70 dB below the fundamental tones.
Time domain response is shown in this plot of a 1 kHz square wave, sampled at 96 kHz. Ringing is minimal, with good shape reproduction. I could only look at the 10 kHz sine wave on my standalone oscilloscope; the spectra plus software is not fast enough to resolve the square wave. On the scope, the performance was qualitatively similar to the 1 kHz square wave, but with a bit more ringing, and more slope on the tops of the voltage plateaus. Slew rate was about 50 V/uS as measured with the scope with a 1 kHz square wave.
Overall, the measurements of the XA30.5 were exceptional, and back up the real test, how the amp sounds.
If you have a speaker with 90 dB/W/m sensitivity or greater, the XA30.5 may be the best amplifier available. It has everything you need, without excess. Some people may not be convinced. They might think they NEED more than 30 W of class A power. They would be wrong. The XA30.5 might just be the perfect amp for any moderately sensitive speaker. They deliver incredible smoothness and liquidity throughout the frequency band, with no limitations in dynamics or frequency extension. For my Gallos, I think I'd want the XA60.5, but they are more than twice as expensive as the XA30.5. With the Gallos, only the bass macrodynamics see any limitation. For many owners, this may not be an issue. This amp is definitely a spectacular value. If you have the right speakers, the XA30.5 will deliver world-class performance for a fairly reasonable price, at least when compared to any competition that could equal them. The question is, will your ego allow you to buy this amp and not have "size" issues? If you let your brain drive your decision rather than other equipment, this amp will be where you end up.