MassKobo 465 Headphone Amplifier Measurements

Content and analysis made possible thanks to support from https://headphones.com, and https://patreon.com/goldensound supporters.

This unit was loaned to me by Elise Audio for review

Design:

The MassKobo 465 is an amplifier that by many in recent months has been regarded as ‘the pinnacle’ of sound for a headphone amplifier, and whilst this post is focused on objective data (full subjective video review is coming), in most ways, I’d be inclined to agree. This amplifier sounds absolutely incredible, and many would assume that it may have clear colourations that are contributing to that, but as we will see….it doesn’t….

I’ll have a full video review coming describing my subjective impressions, but here’s the data!

The 465 internally is a class A ‘hybrid discrete’ approach, using discrete transistors to provide the main output but with opamp modules to facilitate feedback. This product is hand-built by Mr Masuda in Japan, and whilst the design may at first appear simple, there are reasons as to why the high pricing is possibly more justifiable.

The modules in the centre contain MassKobo’s ‘Driver Units’, and on the right, a balanced input unit. It’s a little unclear what the balanced input unit is doing. It appears to be a balanced to single ended conversion circuit, possibly to feed a split single ended signal to the two driver units so that the single opamp in each can still facilitate a feedback loop. The actual output of the amplifier is balanced/differentially driven.

These units are encased in resin, and a better view of them can be seen below:

One of the biggest reasons for the price of this unit is simply that there is extensive time taken to bin/select components based on performance. Mr Masuda takes large numbers of components such as opamps, testing all individually and selecting only the best performing individual components to use in the 465. This takes a fair bit of time, and cost, given as most of the components purchased do not actually get used. The bill-of-materials for this amp is therefore higher than the combined costs of the components in the unit you actually receive. This is similar to some other amplifiers such as the Zaehl HM1.

The actual output itself is driven by discrete transistors, with the opamps facilitating feedback. The front of the amplifier has single ended 6.35mm, balanced XLR-4 and balanced 2x XLR-3 outputs, each of which the user can select between. As well as three gain settings and selectable inputs.

This is of course an incredibly expensive amplifier, and whilst there are reasons as to why this will cost more than others such as the time taken to get components ready through the binning process, or the cost of materials given as only a small number make it into the final product, many will still find it hard to justify spending around $15,000 on a headphone amp.

But regardless of whether you feel it’s justified or not, you do have to appreciate the care that’s gone into producing it! So, let’s look at the measurements:

Measurements:

Test Setup:

– Audio Precision APx555 B-Series Analyzer with 200kOhm input impedance set unless otherwise specified
– Neurochrome HP-LOAD Dummy Load
– All measurements shown are with the amplifier being fed XLR unless otherwise specified
– Measurement setup and device under test are running on regulated 230V power from a Furman SPR-16-Ei
– Audioquest Mackenzie XLR and RCA interconnects
– Exact analyzer/filter configurations for each measurement are detailed in the full reports
– Measurements shown are from the XLR-4 Output unless otherwise specified
– All tests shown are using a 32 Ohm load unless otherwise specified

Full Measurement Reports:

Reports available here:

SNR (from 4V): 105dB

IMD SMPTE: -103dB

DC Offset: 0.5mV active, 0.6mV idle

CMRR: 47dB (same value at 100hz, 1khz, and 10khz)

Gain RCA Input: 1.6dB (Low) 14dB (Med), 30dB (High)

Gain XLR Input: -13dB (Low) -1dB (Med), 15dB (High)
NOTE: The XLR inputs have an optional switch which provides approximately 13dB attenuation. With it set to “-10dBm”, the gain values for XLR input are 13dB higher. It seems happy with full line level input on either option, with almost no performance impact, so choose the one that provides you the most comfortable gain/volume adjustment.

Output Impedance: 10 Ohm (Bal) 5 Ohm (SE)
This may alter the frequency response of IEMs, and to a smaller degree, over-ear dynamic driver headphones. Don’t be surprised if many headphones and IEMs are a bit bassier on this amp.

1khz Sine, 4V XLR in, Amp set to Med gain, Max Vol, XLR Out (300 Ohm):

Pretty good performance, interestingly, not limited by actual harmonic distortion, but instead by some static products at about 2.3khz, 4.6khz and multiples. These are present regardless of the signal output, possibly something oscillating slightly? Though low enough in level not to worry. Actual harmonics are all below -130dB.

1khz Sine, 4V XLR in, Amp set to Med gain, Max Vol, XLR Out (32 Ohm):

Nearly identical performance, nice. Slightly lower output voltage due to the 10 Ohm output impedance, but otherwise, the amp is pretty much load invariant at reasonable levels.

1khz Sine, 4V XLR in, 700mV XLR Out (32 Ohm):

At headphone levels the amp is doing great and is still close to 16 bit resolution.

1khz Sine, 4V XLR in, 50mV XLR out (IEM Level, 12 Ohm):

Quite good SNR for IEMs. >80dB is about the same as a Chord Mojo 2 for example. Some really sensitive IEMs will hiss but the majority of IEMs can be driven fine. The main thing to consider is that the 10 Ohm output impedance will change the FR of many IEMs.

Power On/Off Behaviour:

This test shows the output of the device when turned on/off, to check whether the output may put connected headphones at risk. Regardless of what this test may show, please ALWAYS disconnect your headphones when turning any amplifier on/off unless the manufacturer explicitly instructs you to do otherwise.
This test is conducted with the amplifier in medium gain, and volume pot turned down.

When powering on, the amp has almost no unusual output. When powering off, there is a small spike to 7mV, unlikely to be of any concern, but maybe don’t turn it off with sensitive IEMs still in your ears.

Power:

Maximum power is about 4.5W @ 32 Ohm, however it is fairly load invariant, with delivery upto 1W before any load-dependent change in performance, making it quite ideal for hard to drive headphones. For many difficult headphones like planars, the actual distortion vs level is much more important in determining whether the headphones will be ‘properly driven’ than the max power spec itself.

Volume Channel Matching:

1dB channel difference reached at -50dB attenuation in low gain (fantastic!)

At 50mV output for IEMs, low gain, channel imbalance was 0.2dB

For the majority of range of the potentiometer, matching was within 0.3dB (excellent!). Past 90% it would deviate more.

Volume was very well matched overall.

IMD SMPTE (Intermodulation Distortion):

IMD is overall low and consistent vs output level, same goes for DIM shown below.

DIM (Dynamic Intermodulation Distortion):

Crosstalk:

Crosstalk is a little high, and with more leakage from right channel into left vs the other way around.

THD+N vs Frequency:

THD+N is flat vs frequency, good stuff

Multitone:

Square Wave:

Amplifier is AC coupled and will filter out infrasonic content. As also shown by the frequency response under 20hz:

Wideband Noise/Distortion:

The 465 uses a switchmode PSU which does appear to produce some switching noise above 100khz, though very low in level.

Conclusion:

Overall, the 465 performs quite well in all areas, there’s not really any clear issues at all, all areas are good to great.

This is a discrete, class-A amplifier with plenty of power, low noise and distortion (and CONSISTENT behaviour vs frequency and level as well), that is also quite load invariant. If I had one small complaint it’d be the output impedance, but 10 Ohm is not likely to have a significant effect on anything other than IEMs.