Review: Balancing Mojo with MST Audio’s Mojo-Kai

Sound

In the absence of a great-performing one-box balanced solution there is Mojo-Kai. Let’s look at a some single-ended RMAA results:

Noise – Mojo-Kai: -115,9dB – – Mojo: -116,5dB
DR – Mojo-Kai: 115,6dB – – Mojo: 116,4dB
THD – Mojo-Kai: 0,0005% – – Mojo: 0,0002%
IMD – Mojo-Kai: 0,0009% – – Mojo: 0,0006%
Stereo crosstalk – Mojo-Kai: -114,9dB – – Mojo: -116,5dB

My regular Mojo measures better.

Loaded:

Noise – Mojo-Kai: -115,8dB – – Mojo: —116,6dB
DR – Mojo-Kai: 115,5dB – – Mojo: 116,4dB
THD – Mojo-Kai: 0,056% – – Mojo: 0,062%
IMD – Mojo-Kai: 0,101% – – Mojo: 0,110%
Stereo crosstalk – Mojo-Kai: -93,8dB – – Mojo: -93,7dB

Loaded, it is a wash. I can explain every result except for the narrow rift in unloaded stereo crosstalk:

Every Mojo I’ve used has tested differently. The first I tested returned results similar to my Mojo, but hissed like the Dickens. The second tested within 1,5dB of my current Mojo. Ryuzoh’s outputs 0,5dB less volume at the same maximum stable voltage setting. We should expect its results to stray from my unit by the same number. By and large they do. Only stereo cross talk challenges that assertion. At first I thought the anomalies came from my Lynx HILO. But, after extensive testing, re-testing, both on the day, and months later, my iPhone and iPod, for instance, reveals measured numbers that change by much smaller fractions, if at all. So if Lynx contributes at all to the differences, it is microscopically.

Back to Mojo and Mojo-Kai:

Loaded, the story is much the same, though unfavourable discrepancies appear in DR and Noise; favourable discrepancies appear in THD and IMD. For all intents and purposes, stereo crosstalk is the same.

Honestly, I can detect no audible difference between the two. If you’re not a balanced-out listener, stick with Mojo.

But Mojo-Kai isn’t about its single-ended output. If you’re parting with 45.000¥ for it, you’re a balanced headphone user. And in balanced, this mod really struts its stuff.

NOTE: Because Mojo-Kai’s balanced output gain has been lowered by 14dB, its maximum stere-stable voltage is lower; in this case, it tops out about where an iPhone 4s’s maximum volume does.

Noise – Mojo-Kai: -115,3dB – – Mojo: -111,4dB
DR – Mojo-Kai: 115,3dB – – Mojo: 111,1dB
THD – Mojo-Kai: 0,0003% – – Mojo: 0,0004%
IMD – Mojo-Kai: 0,0083% – – Mojo: 0,033%
Stereo crosstalk – Mojo-Kai: -115,4dB – – Mojo: -112,3dB

And here are loaded test results:

Noise – Mojo-Kai: -115,3dB – – Mojo: -111,3dB
DR – Mojo-Kai: 115,2dB – – Mojo: 111,1dB
THD – Mojo-Kai: 0,0029% – – Mojo: 0,014%
IMD – Mojo-Kai: 0,0083% – – Mojo: 0,033%
Stereo crosstalk – Mojo-Kai: -114,4dB – – Mojo: -93,8dB

Measurable gains show under every metric, both loaded and unloaded. Most of these I consider to be inaudible. A gain of ~30dB in stereo crosstalk, however, is something else. Of course, Mojo spitting 93dB into a notoriously hard-to-drive earphone, is impressive. But Mojo-Kai’s 114,4dB is insane.

Straight numbers don’t tell a great story. Mojo-Kai’s DR in frequencies from 10Hz to 70Hz, expands by 30dB to 10dB, respectively. And driving the SM2, Mojo-Kai achieves stereo separation in excess of -100dB is achieved from 100Hz to 1kHz. From there it runs straight and gentle to its nadir, at -90dB, where the regular Mojo clods off at -64dB.

Also, the strange frequency dip at the very top of the frequency spectrum when either Mojo or Mojo-Kai drives the SM2 is gone. That dip is roughly -1dB by 20kHz, and so, inaudible. But it’s gone.

Listening

One thing to note: because Mojo-Kai’s balanced gain has been lowered, you must remember a few things:

1. turn down the volume before switching to single-ended
2. despite ample current overhead, it doesn’t have the voltage to drive insensitive headphones to insane volume levels

Ryuzoh has a thing for FitEar earphones. So do I. What I love about the FitEar MH335DW is that it is sensitive but not overly prone to hiss, and only minimally accented toward the lower midrange. I don’t know if it is Ryuzoh’s Mojo, or his mod, but the Mojo-Kai he lent me hisses very slightly more than a regular Mojo. The difference is small but noticeable through earphones too sensitive for most of the market: Ultrasone IQ and Shure SE846. MH335DW hisses only barely through either Mojo.

At the ear, I reckon that the MH335DW sounds warmer single-ended than it does balanced. I don’t know if this has to do with its drivers and crossovers being wired for single-ended. Perhaps, it is because when balanced, its load is effectively lightened. Whatever the case, to my ears, balanced MH335DW is lighter, sprier, and in my opinion, more neutrally mated to trance and EDM where tidiness and space are personally paramount.

I notice the same thing through Ultrasone’s IQ. Drier, wider, and tidier. I would accept the charge of placebo as I have no way to properly blind ABX earphones, especially when such a decibel gap exists between the two outputs. But that is what I hear.

End words

Whatever the case, Mojo-Kai is especially well suited to earphones. If you’re a balanced earphone user worried by the shameful output of your DP-X1 and AK380, not to mention Opus#1, Mojo-Kai is your best bet. It plays nice with any device with USB protocol, and thanks to transformer coupling, is less susceptible to RF interference through coaxial input (tested with Lynx HILO with nearby iPhone 6). It is a good design. It is, in fact, a design that makes more sense to me than the stock Mojo, whose second output is redundant.

And, whether or not Mojo-Kai is unseated by upcoming mods (doubtful), it speaks volumes about Mojo’s modularity. By no means is Mojo an analogue to the mid-1990s’ Honda’ Civic, a veritable modding champ. It is a high-end portable DAC. But it can be modded. Ryuzoh says that he can plug a heat sink onto its FPGA to further stabilise its outputs.

Stock, Mojo is world’s better than most of the competition. If you’re okay to pay for balanced, Ryuzoh’s mods take it to another level.