For larger (1200 x 800) images, click any picture

INTRODUCTION

This is going to be a pretty short review by my standards (do I hear you letting a sigh of relief?).  I reviewed FiiO's TOTL Android based touch screen DAP – the X7 – in early November. At the time I gave it a 4/5 (80%), mainly due to the early firmware version, and some missing features in the UI. If I was to review the X7 today, my only con, or issue, would be the battery life.  I’ve got used to the size, and the latest updates (now at firmware version 1.8) have improved the X7 to the point that for my uses, it is practically the complete TOTL DAP I’d always thought it was going to be.
 
I can now go artist, album, track, the DAC works beautifully, the blue light can be turned off, the battery indicator seems to be a lot more accurate, and with the release of the AM2 medium power module, those with harder to drive cans now have some of the power they were looking for.
 
ABOUT FIIO
By now, most Head-Fi members should know about the FiiO Electronics Company.  If you don’t, here’s a very short summary.  FiiO was first founded in 2007.  Their first offerings were some extremely low cost portable amplifiers – which were sometimes critiqued by some seasoned Head-Fiers as being low budget “toys”.  But FiiO has spent a lot of time with the community here, and continued to listen to their potential buyers, adopt our ideas, and grow their product range.  They debuted their first DAP (the X3) in 2013, and despite some early hiccups with developing the UI, have worked with their customer base to continually develop the firmware for a better user experience. The X3 was followed by the X5, X1, X3 2nd Gen (X3ii), X5 2nd Gen (X5ii), M3 and X7.
 
FiiO’s products have followed a very simple formula since 2007 – affordable, stylish, well built, functional, measuring well, and most importantly sounding good.
 

---

DISCLAIMER
The X7 and add on AM2 module were provided to me gratis as a review samples.  I have made it clear to FiiO that I still regard any product they send me as their sole property and available for return any time at their request. But I thank them for the ability to continue use of the X7 – both for follow up comparisons and also for my own personal use. It is my intention to purchase the X7 from FiiO as soon as I can afford it
 
 
I have continued to use X7 and its modules for follow up reviews, and I recently inquired if I could purchase the devices from FiiO.  They have insisted I keep the X7 + modules for my own use. So I acknowledge now that the X7 I have is supplied and gifted completely free of any charge or obligation.  I thank FiiO for their generosity. 

---

 
PREAMBLE - 'ABOUT ME'.
 

Spoiler: Click here to get baseline info about me for use in interpreting the review

 I'm a 48 year old music lover.  I don't say audiophile – I just love my music.  Over the last couple of years, I have slowly changed from cheaper listening set-ups to my current set-up.  I vary my listening from portables (FiiO X5ii, X3ii, LP5 and iPhone 5S) to my desk-top's set-up (PC > USB > iFi iDSD).  I also use a portable set-up at work – either X5ii/X3ii/X7 > HP, or PC > E17K > HP.  My main full sized headphones at the time of writing are the Beyer T1, Sennheiser HD600, and AKG K553.  Most of my portable listening is done with IEMs, and lately it has mainly been with the Jays q-Jays, Alclair Curve2, and Adel U6. A full list of the gear I have owned (past and present is listed in my Head-Fi profile).

I have very eclectic music tastes listening to a variety from classical/opera and jazz, to grunge and general rock.   I listen to a lot of blues, jazz, folk music, classic rock, indie and alternative rock.  I am particularly fond of female vocals.  I generally tend toward cans that are relatively neutral/balanced, but I do have a fondness for clarity, and suspect I might have slight ‘treble-head’ preferences.  I am not treble sensitive (at all), and in the past have really enjoyed headphones like the K701, SR325i, and of course the T1 and DT880.

I have extensively tested myself (abx) and I find aac256 or higher to be completely transparent.  I do use exclusively redbook 16/44.1 if space is not an issue.  All of my music is legally purchased (mostly CD – the rest FLAC purchased on-line). I tend to be sceptical about audiophile ‘claims’, don’t generally believe in burn-in, have never heard a difference with different cables, and would rather test myself blind on perceived differences.  I am not a ‘golden eared listener’.  I suffer from mild tinnitus, and at 48, my hearing is less than perfect.

REGARDING THE X7
This review is essentially about the AM2 medium power amp module released by FiiO for the X7.  For a detailed look at the features of the X7, and a quick run-down on the AM1 (default) IEM module, I would recommend you read my X7 review or indeed any of the 30 something reviews on the X7 currently listed.
 
This is a purely subjective review of the AM2 module – my gear, my ears, and my experience.  Please take it all with a grain of salt - especially if it does not match your own experience.
 

THE REVIEW

PACKAGING AND ACCESSORIES
The AM2 arrived in a small black retail box measuring approximately 90 x 120 x 25mm. On the front of the sleeve is a picture of the bottom half of the X7 with AM2 module attached and some text telling you that this is the AM2 amplifier module.  I’m actually surprised it doesn’t mention the X7 specifically though. On the rear of the box are QR codes which will take you to FiiO’s website or Facebook page.

Front of the retail box

Rear of the retail box

Inner metal container

Removing the box reveals a plain tin box with a nice powder coated finish. Removing the lid reveals a black cardboard envelope, and under this is a foam cut-out with the AM2 module nestled safely inside.
 
Inside the envelope is a warranty booklet in multiple languages, a full set of stickers (which match the ones from the X7) and 2 replacement screws. I was really surprised to see the stickers, but this is FiiO definitely thinking out of the box.  If you’ve brought and applied stickers to your X7 already, the last thing you’d want is a new amp module with no adornments. Although I don’t use them – this is a nice touch.

The FiiO envelope containing the accessories	Full package including the AM2

 
As far as the AM2 goes – the other nice thing to note is the rubber dust cover/protector over the connection pins.
 
The one thing I’m surprised at that is missing though is the specifications – FiiO is usually relay good at including these.  Now it could simply be that they were omitted because this is a review sample – but worth noting anyway. The good thing is that as per usual, FiiO have already listed the spec for the AM2 in the X7 section on their website.
 
TECHNICAL SPECIFICATIONS
The table below lists most of the relevant specifications.  I have (as a comparison) also listed specifications from the default AM1 module.  To give a further idea, I’ve also listed the amplifier specs from FiiO’s X5ii DAP.
 

	AM2 Module	AM1 Module	X5ii (comparison)
Dimensions	~ 64 x 25 x 16mm	~ 64 x 25 x 16mm	N/A
Weight	32g	29g	N/A
Voltage amplification	Muses02	OPA1612	OPA1612/OPA1652
Current Drive	Buf634	AD8397	Buf634U
S/N (H/O)	≥118 dB (A-Weight)	≥115 dB (A-Weight)	≥117 dB (A-Weight)
THD+N (H/O)	<0.001% (32Ω/1 kHz)	<0.0008% (32Ω/1 kHz)	<0.001% (32Ω/1 kHz)
Output into 16 ohm	>350 mW (16Ω/1 kHz)	>200 mW (16Ω/1 kHz)	>436 mW (16Ω/1 kHz)
Output into 32 ohm	>300 mW (32Ω/1 kHz)	>100 mW (32Ω/1 kHz)	>245 mW (32Ω/1 kHz)
Output into 300 ohm	>30 mW (300Ω/1 kHz)	>10 mW (300Ω/1 kHz)	>27 mW (300Ω/1 kHz)
H/O impedance	<0.5 Ω (32Ω)	<0.2 Ω (32Ω)	<0.2 Ω (32Ω)
Peak output voltage	>8.8 Vp-p	>5.2 Vp-p	>8.2 Vp-p
Peak output current	>250 mA	>250 mA	>250 mA
Channel Separation	>72 dB (32Ω/1 kHz)	>73 dB (32Ω/1 kHz)	>75 dB (32Ω/1 kHz)
Play time	8 hours+	9 hours+	10 hours+

 
BUILD / DESIGN
 

Front and bottom of the AM5 including ports

Back and bottom of the AM5

Internal connector

 
Not much to talk about here – the AM1 and AM2 have identical dimensions, and the main differences are internally (obviously right?), but also in their exterior colouring.  Where the AM1 has a brushed titanium appearance on the front and powdered titanium appearance on the rear, the AM2 is a slightly darker shade, and powdered finish on both front and back.  Otherwise both look and feel identical.

AM5, AM2 and AM1 - with cap, hex screwdriver and screws

AM5, AM2 and AM1

AM5, AM2 and AM1

Replacing the modules is extremely easy – just a matter of using the small hex screwdriver included with the X7 – undoing two screws, sliding the AM1 out, and sliding the AM2 in. The fit on the AM2 is perfectly flush, and the only thing very apparent with the AM2 fitted is the change in colour.  This of course disappears when used with the cover.

AM1 left and AM2 right

AM2 attached to X7

AM2 attached to X7

DESIGN – INTERNALS
Although you can’t see them, it is probably a good idea to mention the internal electronics. Where the AM1 uses an OPA1612 for voltage and AD8397 for current, the AM2 uses the higher output combo of the Muses02 and Buf634.  Both have impressively distortion, SNR, and channel separation. The biggest difference is really the voltage output, and slightly higher output impedance with the AM2 (still impressively low 0.5 ohm though). As far as power goes, the AM2 is able to produce triple the output into 32 and 300 ohms. It also has a higher peak voltage output.
 
POWER OUTPUT – REAL WORLD
So the specs are listed above, bit what does that mean in the real world?  The obvious test was going to be with my HD600, so armed with a 1 kHz test tone, and my trusty calibrated SPL meter, I set about volume matching.  To aid quick swapping in this exercise, I simply undid the screws, and left them off – so I could easily slight amp module in place, and then quickly swap as I needed to.  The SPL meter was left in a fixed position, and comparative measures taken were (with the 300ohm HD600):

1:AM1 at 77/120 = AM2 at 67/120

2:AM1 at 86/120 = AM2 at 75/120
 
Both of the above were within 0.2 dB.  I then used those measurements playing actual music and recording maximum peaks – and again both were within 0.2 dB. 
 
I also checked (using similar methodology) the Oriveti Primacy IEMs I’m currently also working on (11 ohm, 107 dB/mW), and the AM1 at 34/120 = AM2 at 27/120.  To be honest, I was actually expecting there to be a bigger volume difference between IEM module (AM1) and medium power module (AM2), but as we’ll find out below, the change in volume isn’t the only thing different when paired with the HD600 – there is also the increase in peak voltage available.
 
One final note – with the very sensitive Primacy, I asked my lovely wife Tania to check for hiss or noise at high volume (with the X7 paused).  She could faintly discern some noise at above 80/120 – but she described it as a very faint crackle rather than real hiss.  For the record, I couldn’t hear a thing.  But at the level described, the noise would disappear altogether at the deafening volumes of music emitted if you were actually listening.  So the AM2 (like the AM1) is very, very clean – and good for lower impedance IEMs (in my test anyway).
 
BATTERY LIFE
Although FiiO) publishes their own real world tests with their modules, I also like to conduct my own.  So over two days, I first used the HD600 and AM1 set to a volume of 80/120, with my library set to repeat, and then kicked it off at the beginning of the day, and checked when I got home.  For most of the time, the screen was off – apart from maybe 6-8 times quickly checking progress.  The AM1 lasted for 9 hours and 42 minutes with 4% left (this was more than my initial testing of the X7 with AM1 – so I suspect FiiO have worked on power management with some of their firmware releases).  I had Bluetooth and Wi-Fi both turned off, and the pulsar light off as well – so this would have helped.
 
Using the same scenario with the AM2, but the volume on 70/120 (to mimic an approximate volume match), I repeated the exercise.  AT the 8 hour 49 minute mark I had 5% of battery life left, so I was actually pretty happy with performance.
 
Of course this will also be dependent on what else you have running, what headphones you are driving, and also if you are using Wi-Fi, Bluetooth, or using the screen a lot.
 
SONICS (subjective)
OK – so we’ve covered the specs, build, power and effect on battery life.  I know the bit everyone is waiting for is any sonic changes.  Before we start, all I’ll say is that my ears are probably not as sensitive as many of you, I volume match very closely, and I’m subject to the same amounts of potential placebo as all humans.  The swapping for the comparisons were as quick as I could make them to preserve auditory memory (same procedure as before – screws undone – swap units, adjust volumes to the pre-set levels, and listen). I varied between rapid swapping (portions of a track about 10-15 seconds) and longer listening periods (a full track at a time).
I used a mix of my usual test tracks - http://www.head-fi.org/a/brookos-test-tracks and concentrated mainly on tracks exposing detail, dynamic contrast, soundstage, bass quantity and vocal quality.
 
With the HD600
My first test was sound staging (Amber Rubarth – Tundra), and I really couldn’t discern any change in soundstage width or depth at all.  In fact both the AM1 and AM2 sounded excellent with this track, and the obvious difference was the amount of headroom – as it is a pretty quiet recording.
 
With the bass test I got my first inkling that there was a bit of a difference.  It was very subtle, but mid-bass impact appeared a touch cleaner and quicker with the AM2.  It was very subtle though and took a lot of time before I was sure it was there.
 
This was repeated when I briefly tried the Beyer T1 with the AM2 (it was actually a pretty nice combo).
 
Listening to vocal tracks and my go-to detail track (Sultans of Swing), the other difference was in dynamic contrast – I guess what you would all energy or vibrancy. The AM1 + HD600 was a little bit more laidback, while the AM2 seemed to have a little more engagement, just a little more life (if that makes sense).  Again – to me the differences were very subtle, and both the AM1 and AM2 were excellent with the HD600, but if I was exclusively listening with the Senn, then I would use the AM2 as default.
 
With the Oriveti Primacy
I stuck with the Primacy because I’ve got to know it really well recently and if my paid work will ever slow down, I’ll be able to finally finish my review. It is a triple driver hybrid – so although a relatively easy load to drive, it can quickly show contrast particularly between bass and mid-range.
 
There was no change at all in staging, and overall tonality was virtually unchanged.  The only really small difference for me was simply in overall presentation, and this was more a feeling from listening to complete songs and switching rather than fast A/Bing. So this could merely be all in my head – but I’ll relate it anyway.  Like the HD600 – the AM1 with IEMs just seems a little more laid back / gentle / smoother overall. The AM2 is a little livelier, a little more there.  I know it’s not volume because I consistently checked throughout.
 
So would I use the AM2 with IEMs as well?  This is a tough one because both modules again sound excellent. Yes there is a bit more battery life with the AM1 and that is a big factor – but the reality is that we’re not talking hours.  For ease of use (not having to swap), I think I’ll probably leave the AM2 in place – it simply gives me options if I ever want to swap to full sized headphones.  But for those driving easy loads and primarily using IEMs, the AM1 really may be all you need. Just be aware that the changes between AM 1 and AM2 (to me anyway) are pretty subtle.
 
GRAPHS
The last thing I did before the final edit was to hook the X7 up as DAC on my PC, and record the output under loopback to see how well the overall measurements were on frequency.  Now one thing I need to stress here is the roll-off in the top end is the filter set-up on the X7 DAC section.
 

Frequency graph	Graph enlarged

 
You’ll see the two curves for AM1 and AM2 are practically identical, so no, the Muses2 is not rolled off (unless the OPA1612 + AD8397 is too).  What it does show is an approximate 6dB difference at the same input volume (between AM1 and AM2), and overall a slow roll of about 1 dB between 3 kHz and 10 Khz.  I guess this is what gives the X7 some of its excellent “analogue” tonality. The rest of the curve is very linear. Oh and in case you are wondering, the most excellent LP% and L5 Pro both also show similar roll offs when measured as DACs – so this isn’t a fault – it’s part of why they sound so good.  

CONCLUSION

I’ve pretty much already summarised everything – but to put it in a couple of sentences …..
 
The AM2 module has a great build, is easy to fit, and sounds fantastic.  I do think there are benefits for higher impedance cans sonically, and if you are running out of volume headroom with the AM1, the AM2 will definitely help.
 
The cost in lower battery life is pretty small, and although I’ve listed it as a con – it was really expected anyway.  It’s hard to give the AM2 anything other than top marks. At an approximate release price of USD 99.00 there is nothing I can think of which makes me hesitate at all in recommending it.
 
(Edit - price adjusted, Sunny informs me the AM2 MSRP is actually $99 so even better deal!)
 
I'm expecting some people to comment on the colour difference - but as you can see, in the case it is not at all noticeable.
 
  
 
FINAL THANKS
Once again thanks to Sunny over at FiiO for giving me a chance to try the AM2 before its global release.
 
ADDENDUM
George asked me in the comments section about the roll-off I measured in the DAC section with both AM1 and AM2. I included the graphs originally because I'd had the comment that the Muses2 chip was rolled off and smoother than the OP1612 in the AM1, and I wanted to show this wasn't the case.  So tonight I reset the loopback tests, and remeasured the X7 (AM2) as well as the X3ii, X5ii, X5 original, FiiO's E17K dac/amp and also the Luxury & Precision LP5 (a TOTL player that really does sound incredible).  I'm expecting these graphs may surprise a lot of people - and especially George as he's told me he doesn't like anything rolled off - yet I know he really enjoyed both the X5 and X5ii.  The graphs were all produced with ARTA and a loopback measuring a 16/48 signal (16/44.1 with the LP5 due to driver limitations).  You'll see a very similar pattern on all the DAPs except for the E17K - and I've thrown that in there simply to show that my measuring equipment is working as intended.  These are provided simply as a matter of interest - and have no real bearing on the review.