Emotiva XPA-2 Measurements and Analysis

I did some quick spot-checking on the XPA-2 amplifier gain structure to ensure it could be properly driven with a wide assortment of preamps or receivers. My personal criteria for amplifier gain structure is that it should be able to hit full power when driven with 2Vrms. The XPA-2 blew this notion out of the water with an unusually high voltage gain of 32dB (un-bridged) and 38dB (bridged). After confirming this twice via my measurements, I pinged Emotiva technical support and they informed me this was purposely done for end users using a budget receiver many of which tend to clip above 1Vrms output. I can certainly appreciate this as I have found this to be the case with many receivers even in the $1k price range. The XPA-2 reached its full power with only 1.2Vrms. Although the penalty of having such high gain is also higher noise, the XPA-2 remained very quite and yielded a SNR of 105dB at 319 watts (unweighted) which translates down to about 80dB at 1 watt. Emotiva informed me that part of the reason for such good SNR measurements is that all of the voltage and current gain is done in last stage of the amplifier.

I also suspect Emotiva boosted the gain to make their amplifiers sound more impressive (remember the predilection for people to associate louder with better) when a newbie hooks up their amplifier for the first time. Personally, I’d prefer to see a +3, 0, -3dB toggle switch on the backpanel for those who have preamps with plenty of drive to further lower the noise floor.

Frequency Response

Figure 1. XPA-2 Frequency Response

The frequency response was very linear and extended to a -3dB point of about 130kHz. In fact at full power the response was still ruler flat validating their claim of +- 0.15dB from 20Hz to 20kHz at all power levels.

Power Bandwidth vs Distortion

I experienced a rather odd test glitch when running power tests on the XPA-2. Anytime I swept 20kHz to 20Hz, the amp would instantly shut off if I drove it to more than 100 watts. I later discovered that sweeping from 20Hz to 20kHz eliminated this problem. Upon further investigation it turned out that the power protection circuitry built into the XPA-2 was being tripped by my test gear when it ramped up to high frequency testing. This is NOT a real world scenario, NOR did this ever occur in my listening tests. In fact, it has taught me a lesson to always sweep low to high despite the fact that my test scripts default the other way around.

Figure 2. XPA-2 Power Bandwidth vs Distortion

Amplifier Power & Efficiency

Using a full power (0.1% THD + N) continuous power test at 1kHz, I measured about 65% efficiency which is about the best you can get from a linear amplifier design. The idle power of this amplifier was around 45 watts and at no point during my strenuous power tests did I find the XPA-2 to heat up to any degree other than a slight warming which didn’t prompt me to pull my hand away from if as I plopped it down directly on the top of the chassis. You can thank the generously massive heatsink area for this amps cool running characteristic.

Full Bandwidth (20Hz to 20kHz) Power Measurements at 0.1% THD + N

** not tested during review

Editorial Note from Emotiva on Power Measurements

According to Lonnie at Emotiva, they achieved the following power measurements on an XPA-2 in their test lab with has a dedicated 20A line of less than 20 ft running to their mains (much shorter than my power feed from the mains). It is possible that under their testing scenario, they achieved less line sag than I did which explains their higher power measurements.

Stereo 4 ohm load, both channel driven, 0.1THD: 500.1 WRMS.

Stereo 4 ohm load, both channels driven, 1.0% THD: 527.0 WRMS.

Bridged 8 ohm load, 0.1%THD: 1.04 KWRMS

Bridged 8 ohm load, 1.0%THD: 1.105KWRMS

Not only did the XPA-2 stomp the RPA-1 in terms of sheer output power, but it also bested my $7k 10 channel Denon POA-A1HDCI by a considerable margin. For example, with 1 CH driven into 8 ohms the XPA-2 delivered a whopping 312 watts while the POA-A1HDCI delivered around 185 watts. In 4 ohm loads, the XPA-2 delivered 515 watts while the POA-A1HDCI delivered 300 watts. In bridged mode it was especially interesting how the XPA-2 delivered nearly 2X the power of the POA-A1HDCI belting out 860 watts and sagging my 20A line from 119Vrms to 116Vrms in the process. This amplifier was so powerful that I had to create an elaborate series-parallel wiring scheme for my power resistors to avoid them turning into a grilled cheese sandwich. I was unable to achieve the 1 kwatt power that the XPA-2 is speced for in bridged mode but I am sure its achievable (at least momentarily) into hard clipping when using a variac to keep the line voltage at a constant 120Vrms. This is NOT a real world test scenario and one I don’t subscribe to when doing power tests.

I didn’t test the XPA-2 into a 4 ohm load bridged since Emotiva didn’t spec this, and I noted the power line sag was becoming quite a factor due to the sheer power of this amp. But, I would suspect based on the 1.2kVa transformer and the 65% measured efficiency that this amp would likely deliver around 900 watts or so into a 4 ohm load.

Keep in mind most review publications test at clipping and don’t do continuous power measurements so our power numbers are usually a lot more conservative than what you typically find from other reviewers.

For more info, see: The All Channels Driven (ACD) Test

That being said, the XPA-2 exceeded its power specifications by considerable margins. Into 8 ohms it was rated to 250 watts at 1% THD , yet I measured 312 watts at 0.1% THD. Into 4 ohm loads it was rated to 500 watts at 1% THD while again my measurements exceeded this rating producing 512 watts at 0.1% THD. Make no mistake folks this is one POWERFUL amplifier and most definitely the most power for the money amplifier that has ever come across my test bench since I launched Audioholics nearly 10 years ago!

FFT Distortion Analysis

  

Figure 3. XPA-2 FFT Distortion Analysis at 1 watt (XPA-2 left, RPA-1 right)

I ran FFT distortion plots at various power levels to determine how clean this amplifier really is. At 1 watt, the spectral distortion was good with the second order harmonic being (9.007 + 67.533)dB being 76.54dB down from the fundamental or 100*alog^-1(-76.54/20) = .015%. While this is a respectable measurement, it is no way near as pristine as what I’ve seen on their RPA-1 demonstrating a measurable trade off between higher sustained power of the XPA-2 versus measurably cleaner low power of the RPA-1. If you tend to favor low listening levels, you may opt for an RPA-1/ RPA-2 over the XPA-2.

Figure 4. FFT Distortion Analysis at Rated Power (unbridged left, bridged right)

At near full rated power (250wpc) into 8-ohms, spectral distortion of the XPA-2 was again good but not stellar, with the second order harmonic being (32.354+40.681)dBv = 73.05dBv or 100*alog^-1(-73.05/20) = 0.02% This was not as good as I’ve seen on any other Emotiva amplifiers I’ve previously measured, none of which have come close to being the powerhouse that the XPA-2 demonstrated to be.

In bridged mode at the same power levels, the even order harmonics of the FFT spectral distortion profile appeared about 10dB better than unbridged mode, bringing it to near equal footing with the RPA-1 but able to belt out significantly higher power levels (nearly 4X the power of the RPA-1) when driven to its limits.

Crosstalk

Figure 5. XPA-2 Crosstalk at Rated Power

Running a full range frequency sweep through the XPA-2 amplifier at full rated power (250wpc @ 8-ohms), I measured channel to channel crosstalk between the two channels where one was the disturber and the other was the Device Under Test (DUT). The Audio Precision plotted crosstalk of both channels over frequency by varying the Distruber/DUT channels. The XPA-2 produced surprisingly excellent crosstalk measurements (>140dB at 1kHz) with only a gradual rise with increasing frequency because of capacitive coupling. This is the best crosstalk measurement I’ve ever measured in an amplifier and despite the XPA-2 having a singular centralized power supply, the channel to channel isolation was on par with the best monoblock amplifiers I’ve ever seen!

Output Impedance and Damping Factor

Figure 6. XPA-2 Amplifier Output Impedance vs Frequency

Amplifier output impedance when the XPA-2 was driving 8-ohm and 4-hom loads at 1 watt was within our preferred 100 mohm range for the entire audio bandwidth. We like to see amplifier maintain 150 mohm output impedance or less so that it will sound consistently good for a larger variety of speaker and cable loads with rather complex impedance profiles. This is a very good measurement and yet more confirmation that the XPA-2 amplifier design was well executed.

Figure 7. XPA-2 Amplifier Damping Factor vs Frequency

Again the XPA-2 did NOT disappoint. An amplifier with a low output impedance will also maintain excellent damping characteristics since these are related metrics. The XPA-2 maintained a damping factor above 100 for 8 ohm loads and ½ that as expected for 4-ohm loads. This fell right within our benchmark criteria.