Head-Fi.org › Forums › Equipment Forums › Dedicated Source Components › ami MUSIK DDH-1: XMOS+PCM1795 USB & S/PDIF DAC
New Posts  All Forums:Forum Nav:

ami MUSIK DDH-1: XMOS+PCM1795 USB & S/PDIF DAC

post #1 of 9
Thread Starter 

Hi there, got a DDH-1 loaner thanks to my local reseller and I felt like this product would well deserve its own thread here on head-fi together with a quick review while I'm at it :happy_face1:

 

CREATOR: gd-jpeg v1.0 (using IJG JPEG v62), quality = 90

 

CREATOR: gd-jpeg v1.0 (using IJG JPEG v62), quality = 90

 

The official product page can be found at http://amimuse.com/ami-musik-ddh-1/

 

Its user manual is available in five different languages at http://amimuse.com/support/downloads/

 

6moons already posted a thorough review at http://www.6moons.com/audioreviews/ami/1.html

 

 

So on the input side, we have coax / Toslink / analog line-in & USB.

 

Concerning the outputs, we have RCA line-out, Toslink and both 1/4" & 1/8" headphones jack.

 

As you can see, it's using a USB3 B connector even though it's a USB2 XMOS audio controller, I asked my reseller about this and he told me that it's a good way to be future-proof, it provides a more secure fit than a USB2 B connector and USB3 cables have to be seriously built and heavily shielded due to the ten times higher speed rate compared to USB2. It's the first time I see this trick being used and that's a really great idea IMHO, it's also fully backward compatible with USB2 B connectors of course.

 

Here's what it looks like in use:

 

 

 

So let's see what's inside, shall we:

 

imagebam.com

 

 

2 boards, one for the USB conversion and one for the DAC circuitry.

 

Let's start by the USB board:

 

imagebam.com

 

imagebam.com

 

-XMOS chipset, supporting up to 32bit 192kHz resolution. Fully UAC2 compliant and only requiring drivers on Windows that also provide bit-transparent 32bit ASIO capabilities. These drivers are rock-solid and they come with the latest 1.63 update that's fully W8 compatible of course :normal_smile :

 

imagebam.com


-Two 1.0 ppm TCXO clocks for 44.1/48kHz multiples, some manufacturers like to go cheap and use a PLL-based single clock synthesis..this will add at least 50ps of jitter when two discrete 1.0 ppm clocks will reach as low as 2.5ps timing period jitter....über-low jitter, here I come:popcorn:

 

imagebam.com

 

-IL1117 3.3V high efficiency linear regulator to feed the XMOS chip. For the record, the XMOS datasheet implementation is bus powered and uses a Semtech SC4626A switching-mode voltage regulator(with a working freq of 2.5MHz). Many/most XMOS "by-the-book" designs rely on this architecture, the Hiface 2 comes to mind. AMI took power management very seriously.


-Not USB bus powered, I can use a USB cable with the +5V pin broken. All the rage in "audiophile" USB cables is to ensure that its power lines won't pollute its data lines(some companies even sell grossly overpriced USB computer controller boards that come with a switch that can disable the +5V USB bus power). Well fear not, you can either tape or break the +5V pin and you just got yourself an "audiophile" grade USB cable :bigsmile_face:

 

All this to say that IMHO this board uses the best of what USB audio currently has to offer and it could sell for +$500 if it came in a small box with RCA/BNC connectors for S/PDIF. I've got nothing to complain about, that 's good =)

 

 

Now for the DAC board:

 

imagebam.com

 

-PCM1795 123dB SNR 32bit 192kHz DAC chip


-TPA6120 headamp IC


-A whole bunch of swappable opamps, whose description can be found at http://amimuse.com/support/faq/

 

-Nichicon audio grade caps


-Opamps are fed 20V(±10V) which a far better choice than the usual 12V and would appear to be a good compromise between SQ improvement due to the higher voltage swing(for instance most discrete opamps require at least 17V to sound their best) and an excessive amount of heat to dissipate.

 

imagebam.com


-An LM78L05 5V linear regulator to feed the DAC chip & headamp IC, nice touch again!

 

imagebam.com

 

As you can see, I've rolled the opamps.....6moons has pictures of the stock board. I decided to use some of my favorite "sociable" opamps, OPA2132P as I/V, OPA627BP(genuine for sure as they were sourced from digikey) as buffer in the very same setup as in the Eximus DP1, and two OPA132U on an adapter for the line-out.

 

I used dual opamps for the I/V as they are taking care of both polarities of the same channel and single opamps on adapters for stereo signals. Eventually I will solder four more OPA132U's and hear whether they'll make any difference over OPA2132P as I/V(TI make very clear in the OPA1612 datasheet that single opamps provide better signal separation than regular dual opamps)...I will also try OPA827 and OPA1611 as the former is supposedly great for I/V and I really used to enjoy the latter(TI advise using them in high-end BD players).

 

Opamps are just really fun and mostly harmless if you stick to unity gain stable chips and always make triple sure that everything's been soldered and inserted in the right direction.....it's just sheer fun to solder and seek any audible difference/improvement, it really is lego for grown-up OCD'ed audiophiles ^^

 

So I did quite a lot of listening but after a short while, I felt like the stock SMPS was holding back SQ......opamps do have a spec called PSRR and the cleaner the PSU the higher the SQ really.

 

Here's the stock SMPS brick:

imagebam.com imagebam.com imagebam.com

 

So I plugged a linear 12V linear PSU(lucky me it was the same barrel connector size), as you can see there is no question that it's no SMPS due to the much higher weight and fixed input voltage:

 

imagebam.com imagebam.com

 

Well, now we're talking! After PSU & opamp rolling, this little shiny box becomes hard to believe.....I A/B'ed it against the built-in DAC of the Eximus DP1(2.5ppm TCXO driven XMOS/dual-mono PCM1794/OPA2132P/OPA627AP):

 

imagebam.com

 

Well, the similarities in design were very hard to miss. The DP1 DAC section goes through some discrete "diamond" buffering so the sound is sweeter and more forgiving when OTOH the pimped DDH1 presents the naked truth and a less colored but just as 3D sounding soundstage. I seriously did not expect this and I'm honestly floored by how the DDH1 is now performing after those minor modifications :ph34r:

 

Surely, OPA627 has been a long time favorite of audiophiles for a good reason....its musicality and effortless -yet very detailed- presentation are the reason why it's still being used in new designs. Some ppl claim OPA827 to be an improvement but that doesn't match my experience at all. Also many people buy cheap 627's from ebay, too bad it's the most counterfeit opamp..

 

You can read in this review that OPA627 is a real treat through headphones, and it's also undoubtedly part of the reason why the DP1 sounds so great to many ppl(including me ^^).

 

This highly skilled DIY designer ran an opamps shoot-out and wrote the following:

Quote:

OPA132 [..] (2132) in DIP packages, [..] A very good chip to start with. Indeed, you may be so happy with it that you stop looking at other chips!

 

OPA627 [..] The first thing I noticed is the cleanliness of the sound. With this chip in the test amplifier, I heard known problems in a low-end portable source more clearly [..]

 

The 627 also seems to do better on recordings with room ambience: it reveals details about the acoustic space that the OPA132/134 chips will hide, making them sound "flat" in comparison. But these two chips are more alike than different. Both have the characteristic laid-back and dark Burr-Brown sound, and both are very tolerant, stable chips.

 

He of course made sure that all opamps were stable and used within a proper surrounding design. Implementation is everything and rolled blindly in an unfit circuit OPA132 can sound "grainy" and OPA627 dark and veiled.....let me tell you that this is not the case here, it's working just as well as it does in the DP1 to my ears:atsmile:

 

I ran those tests through the widely acclaimed headamp of the DP1, so I also tried if off an old SONY AVR:

 

imagebam.com


Well, it still sounds really fantastic tbh.....of course the headamp of this AVR is based on NJM4565(even worse than 4580 apparently)  and its volume control is digitally attenuated so it's not quite audiophile grade, yet the 3D SS, clarity and thick deep bass of the DDH1 keep impressing me spin.gif

 

Whenever I'll have the opportunity, I will try the DDH1 with <$500 serious headamps and will report back.

 

Of course it also sounds good stock but NJM2114(an improved 5532) and LM49860(a higher grade LM4562) on swappable sockets really begged to be rolled for more expansive chips. Actually, manufacturers do have to make a living so when they use OPA627 or AD797 you are basically paying a lot more than if you bought them yourself and rolled them on your own.

 

My original intent was to find an XMOS DAC with serious clocking, swappable opamps, a fixed volume line output and great power regulation....the DDH1 does just that and then some :beerchug:

 

I also tried the built-in headamp but of course it's no competition with the DP1 headamp or any other $1K headamp I played around with in the past......we're on head-fi and a TPA6120 headamp without discrete buffering or a more elaborated surrounding design will have a hard time impressing the most demanding amongst us. This said, it's very capable considering that it's thrown-in with the DAC and I got solid bass response out of it. It really is a vanilla TPA6120 implementation as far as my eyes and my ears can tell.

 

Of course you can use the line input together with the pot as a preamp and it also comes with a Toslink output.

 

There's a "variable / fixed" switch at the back of the unit that toggles the front pot onto the line-out, you can see a relay right behind the pot and can hear it clicking when activating this switch.

 

I didn't try its S/PDIF capabilities because again this protocol is broken by design(the galvanic isolation of Toslink costing at least 1000 ps of jitter) and it would be a shame to leave this high-end 1ppm XMOS based asynchronous USB input unused, 6moons used the Toslink input in their review if you feel like checking it out. This said, S/PDIF inputs are great to plug a TV tuner, game console or CD player when you can't be hassled to turn on your computer.

 

All this to say that once slightly modified, it's really hard to believe that this little monster sells for $550....again, last year you were lucky to get a bus powered USB to S/PDIF transport for this kind of money, in 2013 you get a most impressive sounding and cleverly well-built self-powered standalone DAC for the same price....and you can even play around with rolling as much as you like :bigsmile_face:

 

I see that AMI is a young company and that this would be their first product, my reseller told me that he will also stock their forthcoming products and I can hardly wait to hear what that'll be all about.

 

Hope you had as much fun reading this short review as I had writing it, [:loomy] this little thing and personally I'm definitely pulling the trigger on it[:alexsilvio] 


Edited by leeperry - 9/26/13 at 3:40am
post #2 of 9
Thread Starter 

One week of break-in later, this thing still sounds as glorious as it did a week earlier :happy_face1:

 

I've had the opportunity to try a few headamps but they all were a far cry from the DP1's...I would really like to hear this thing: Yamaha A-S300 - Integrated Amplifier

 

Killer vintage look, low price, a remote control and made by yamie, amazon carry it so I could always try it for free for a month :cool:

 

Here's what's in the tuner version(R-S300).

 

I realize that a remote control means that compromises will have to be made, but hopefully this 2012 receiver will work in +24bit(and possibly 32bit) so the added value of a remote might win over a -maybe inaudible- SQ decrease. Apparently, orthos come with a more or less dead-flat impedance curve and would be very easy to drive due to the fact that they are seen as a pure resistive load by the amp.


Edited by leeperry - 10/7/13 at 10:24am
post #3 of 9
Thread Starter 

As it turns out, an A-S300 owner has confirmed me that its volume knob is motorized and would appear to be working in the analog domain. I do remember that ages ago I had a japanese AVR that worked in the exact same way.

 

OTOH, in the audiophile world digital attenuation in 24bit means that dynamics get butchered: http://www.6moons.com/audioreviews/resonessence/2.html

 

Only the 48bit volume control of the ES9018 would save the day and I can't seem to be able to find a single serious headamp with a motorized analog volume control and a remote, huh..God forbid, a remote control? the audiophilism gods say that you're grounded :evil: 


Edited by leeperry - 10/1/13 at 12:39pm
post #4 of 9
Thread Starter 

Either way, still having fun with soldering & rolling like back in the days...discrete is nice but opamps are just a sheer lot of fun:popcorn:

 

imagebam.com

 

Actually I think the line-out buffer and DAC LPF play the same role in coloring the sound, I can swap them and the sound is exactly the same. OPA827 is supposedly a very high-end chip and I've always enjoyed the natural sound of OPA1641(TI advise it for use in high-end BD players). Fun times :happy_face1: 

post #5 of 9

Nice review/summary here. I just covered this same device at InnerFidelity and came to more or less the same conclusions - this little guy is the real deal. The headphone stage is the weakest link for sure, but it's not without a certain charm when mated with the right headphones. I've definitely heard worse. 

 

In the time between writing that article and publishing it, I did find some opamps I really enjoyed - the Muses 01 or 02 rolled into all spots of the DDH-1. The 02 kept a similarly warm sound while improving stock in almost every way, while the O1 changed it into a more neutral sound. But I still don't feel comfortable recommending hundreds of dollars worth of upgrades on a $550 device. Plus I'm just not quite the opamp nut that leeperry is :wink_face:

post #6 of 9
Thread Starter 

Nicely done :normal_smile :

 

I agree that the variable Vs fixed line-out volume control doesn't change much if anything but it's still nice to be able to bypass the pot anyway.

 

It's also undeniable that the XMOS drivers are as flawless as it's gonna get, their chipset even allowing the use of two discrete clocks so their solution really offers everything we could hope for.

 

I listen to my DDH1 through the DP1 headamp on a daily basis and I like how it's really true to the source, barely colored and its very tight clocking is a real treat to listen to :happy_face1:

post #7 of 9
Quote:
Originally Posted by leeperry View Post
 

Nicely done :normal_smile :

 

I agree that the variable Vs fixed line-out volume control doesn't change much if anything but it's still nice to be able to bypass the pot anyway.

 

It's also undeniable that the XMOS drivers are as flawless as it's gonna get, their chipset even allowing the use of two discrete clocks so their solution really offers everything we could hope for.

 

I listen to my DDH1 through the DP1 headamp on a daily basis and I like how it's really true to the source, barely colored and its very tight clocking is a real treat to listen to :happy_face1:

 

Remember when a USB stage of this caliber was pretty much unheard of? Wasn't so long ago that you couldn't get a USB implementation this good, at any price? They simply didn't exist. Then for a while you could get them but only by spending mega dollars on some Wavelength gear (not cheap). Now we get it crammed in this $550 unit. Technology is great (when used properly). 

post #8 of 9
Thread Starter 

Well, now that high grade XMOS inputs don't cost a fortune anymore, I belive companies like Wavelength like to explain that when everyone's using ES9018 in ASRC mode, they are the only ones who know how to disable this feature completely, making it sound less colored and far more "natural". The sales pitch had to adapt I guess.


Edited by leeperry - 11/11/13 at 2:30pm
post #9 of 9

My DDH1 is for sale if anyone wants it. The best small DAC amp I've tried yet. Really impressive and I'm tempted to keep it, but have no real need for it. Bought it just to try and that's it. 

New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Dedicated Source Components
Head-Fi.org › Forums › Equipment Forums › Dedicated Source Components › ami MUSIK DDH-1: XMOS+PCM1795 USB & S/PDIF DAC