The Yulong D100 is an integrated DAC/amp unit that costs $449, plus $40 fixed shipping; that makes it realistically a $500 product. Mine is from eBay seller wsz0304 who has been excellent to deal with. The D100 is roughly 10 inches wide, 7 inches deep, and 2 inches high, and weights about 5 pounds. It has a built in switch for selecting 110/120v or 220/240v operation, so it works pretty much everywhere, and has a standard detachable IEC power cable. It accepts most types of digital signals including USB, optical, coaxial SPDIF, and AES/EBU. It accepts signals up to 24/192 over the AES/EBU, optical, and coaxial digital inputs, and accepts 24/96 over USB. Once the digital to analog conversion is complete, that signal can be sent out through a standard RCA output, or via balanced XLR output. It also has a built in headphone amp which features separate ¼ inch jacks; one for low impedance headphones and one for high impedance headphones. They don’t specify the cutoff ranges, but overall the unit claims to handle headphones with impedances ranging from 16 to 600 ohms. They encourage you to try both jacks with your headphones and see for yourself which works best. It features a built in display with some useful info, and a simple front panel button that cycles through the various inputs. There is a “sound mode” button that allows you to cycle between mode 1 and mode 2. Mode 1 is full range, and mode 2 is supposedly darker. With the help of a translator friend on this forum (you know who you are, thanks again!) I was able to find out that mode 2 has “a minor reduction of 0-3dB between 15-20kHz range resulting in slightly darker highs. Sound mode only applicable for the headphone out.” I wish this mode dug deeper, perhaps down to 10kHz, as that might have made it more noticeable. As it is I don’t think I can reliably tell them apart in a blind test.
Thus far most of my general description could be any old DAC/amp combo. The magic here is in the design, in the combination of parts and implementation. Standard digital signals are accepted by a Cirrus Logic CS8416, and USB input is handled by a Texas Instruments TAS1020B. From there the signal is sent in I2S form to the Analog Devices AD1896 asynchronous sample rate converter, where it is processed to a 110kHz sample rate using a high precision system clock, and word lengths are padded to 24-bit. That signal then goes to an AD1955 where the actual digital to analog conversion happens. I/V conversion, filtering, and buffering are then handled by the combo of AD4075 and OPA275. At that point the signal is ready to go when using the D100 as a DAC. When using it directly with headphones, the signal is sent through the headphone amp section which features an OPA2134.
The key portion of this design is the ASRC process. I’m no expert, but I’m told that ASRC can be tricky to do well, yet is also well worth it when properly implemented. The main enemy in the digital world is jitter, and it must be dealt with. There are other methods of reducing jitter, such as running in Asynchronous USB mode like Ayre and Wavelength, but those are considerably more expensive. By using off the shelf solutions in a well designed circuit, the D100 is able to achieve excellent results in a manner similar to Benchmark with their DAC-1.
Looking at the parts list above, many of the choices seem familiar. The AD1955 is used in the Berkeley Alpha DAC among others. The CS8416 is used in the PS Audio Digital Link III and the Grace Designs m902. The Lavry DA10 actually uses the same AD1955/AD1896 combo as the D100. The USB receiver (TAS1020B) can also be found in products from Wavelength, Ayre, Empirical, and of course the Benchmark DAC1 USB. Some may scoff at the OPA2134, but it is used in the Matrix M-Stage (aka Lehmann Black Cube Linear) with excellent results, so again we see that good products are more than just the sum of their parts.
The point here is not just to list a pile of parts and somehow claim high quality. If that were the goal I think ASRC could be avoided in favor of more exotic capacitors and such. Yulong claims they designed this unit to show that they can compete above the level of your standard eBay type DACs, and they have put in the effort with prototyping, testing, and even redesigning. And we can even find a bit of proof for that: the original info stated that they had tested various frequencies and found that 256fs/132.3kHz was about as high as the AD1896 could run, and provided the best performance. The master clock ran at 33.8688MHz. The AD1896 absolute max MCLK is 34MHz under specific conditions, so everything fit together. But somehow after the first 50 units were made, it was discovered that the lead-free AD1955 in the production version could not run reliably at 256fs/132.3kHz. There were minor changes to the circuit for an end result of 110kHz resampling. The sound signature and technical measurement supposedly remains unchanged, with a claimed slightly better imaging for the 110kHz version. I can’t verify this as my unit is serial number 92 and thus already has the redesign. But the point is there is some real engineering happening here, not just throwing things together and churning out an impressive spec sheet. They could have gone with 192kHz upsampling, but claim that the 110kHz gives a distinct filter performance benefit (Benchmark makes the same claim).
Much has been said of the benefits of Asynchronous USB mode versus Adaptive USB mode. You can read all day and not find a consensus on the issue. While it is true that Asynchronous mode (not to be confused with the ASRC process) seems to have more potential, it is also generally expensive, and in practice there are plenty of excellent options (like Wadia, and the above mentioned Lavry and Benchmark units) that utilize Isochronous Adaptive mode and achieve excellent results. Ultimately it is the sound that counts, not the method used to produce it.
DSC_0253.jpg
DSC_0256.jpg
PACKAGE
The D100 arrived very quickly considering its international journey to my house. The box is fairly plain, but the packing inside was sturdy and did the job. Inside I found the D100 itself, a power cable, a USB cable, and a user manual. Nothing too exciting but enough to get you started.
The manual is particularly confusing, with multiple typos and odd wording. Most likely it was translated poorly from the original language. There’s not really much info in there anyway; it still talks about the 132kHz ASRC. I talked directly to Yulong to get the updated info above. About the only interesting thing in the manual is a printout of a frequency response test, which looks nice but is too small and hard to read to be of much use. You can see it on the eBay listing from wsz0304 as well.
BUILD QUALITY
The D100 is very nice looking. It features a thick brushed aluminum silver faceplate, and even though the rest of the unit is black, it has a nice textured smooth feel to it. The volume knob turns with tiny clicks that are felt but not heard, and the buttons on the front panel have a nice action to them.
The LCD display lights up in a bluish color, and tells you a bit of useful info: which input you have selected, what the sample rate is of the incoming signal, and which sound mode you have selected. I found the sample rate info to be consistently accurate, which is helpful when dealing with a variety of signal types. It allowed me to figure out that I needed to switch a setting on my Rotel player to output at 96kHz instead of downsampling to 48kHz.
One thing I like is that it remembers which input you last used before you turned the unit off. Apparently the old Yulong DAH models had an issue with resetting to a certain input after every power down, and that has been fixed on this model.
Overall the unit looks and feels like a high end piece of gear, and I don’t think anyone would be disappointed with it. One odd thing I noticed: when I first took some pictures for this review, I found that the flash made visible a slight haziness to the front panel finish. It was not something I could see directly, but only showed up with flash photography. Check my pictures for examples. I ended up wiping it down with a damp cloth and it went away completely, so it must have been some type of remnant from the manufacturing or shipping process. In any case, if the front looks odd in the pictures, that is the explanation, and it is not a real issue to be concerned about.
DSC_0258.jpg