Earlier this year I purchased several of the M2Tech hiFace USB to S/PDIF converters and wrote up my thoughts for our February 2010 issue (M2Tech hiFace review can be found here). As might be expected in these times of racing technological advances the number of devices for interfacing computers to our audio systems seems to increase weekly. A few months after the original hiFace review was published I was contacted by John Kenny, a gentleman from Ireland with whom I'd exchanged emails in the past about various DIY projects. I knew John had been experimenting with modifying his hiFace with lithium polymer batteries for powering the clock. He had added a second battery to power the FPGA and USB receiver chips and installed the works in a small plastic enclosure and asked if I would be interested in reviewing the battery powered version.
John has given his product the somewhat ungainly name of MK1 boxed hiFace, so I'll just refer to it as the mod hiFace. The mod unit is housed in a simple, small but quite adequate ABS enclosure 4.5" x 3" x 1.5" thick. It addition to the existing hiFace connectors there is a power switch with a blue LED along with a charging jack. Inside the case is a hiFace module and two large lithium "Nanophosphate" batteries from A123 Systems. These cells have a 3.3V nominal output voltage, making them suitable for driving the clock, FPGA and USB receiver chips directly without the use of a voltage regulator. They reportedly have very low internal self noise but I have no way to measure this.
There was a short delay in obtaining my review unit as John had to wait for hiFace modules from M2Tech. While waiting for the mod hiFace to ship the new M2Tech hiFace Evo was announced with distributor samples set to arrive in June. A check with Steven about incorporating the Evo in this review was met with an enthusiastic yes and shortly after arranging to get one of the samples I was also offered the opportunity to review the new Halide Design Bridge. This too would be a while in coming as their initial run sold out quicker than expected. I'll describe both of these in a bit. After roughly a month I received a small box from Ireland and also a charger that was drop shipped from a vendor in the US. John does not provide a charger for his unit as he ships worldwide and has found it more cost effective to have the end user order a charger from a vendor in their region. Operation and installation of the mod unit is exactly the same as a standard hiFace with the exception of a power switch and the need to charge the batteries. John recommended eight hours of use prior to recharging the batteries. I never play my system that long and as the batteries are recharged when the power is switched off I had no problems at all with this. Some folks leave their systems powered up at all times and need to keep that in mind when considering a product such as this.
There is one additional consideration when using the mod unit and that is the requirement of a short A-A USB extension cable, something suggested even with a standard unit. A very standard USB cable from the local computer super-mart was used for listening as last I checked none of the high end cable companies were making A-A extension cables, all were A-B or A-Mini or sub-mini.
Firing up the system it was immediately apparent this was in a different league than the standard unit. My first thought was something like "where did all this bass come from?" followed quickly by the realization that my sound stage was deeper and wider with more layers on some music. Wow, so if things are this good cold out of the box, will it get even better with some time on it? I don't like the idea of "break in" but have observed it enough to know it occurs. This without listening and becoming accustomed to the sound as I make every attempt to run things in while I'm not home so that any changes that may occur are actual and not imagined. I've been fortunate enough to have two of an item a few times and compared the unused to the burnt in version and noted changes so I'm satisfied that they do happen and are often audible. Undoubtedly some will disagree but that is expected. As such, I let it play on repeat into the DAC all day while at work for a few weeks. The sound did change and became somewhat more detailed and smoother. It was not a great change but it things did seem to get a bit nicer.
Several weeks into listening to the mod hiFace John contacted me about trying some inline attenuators as the output voltage of the hiFace is about 2.8V p-p (peak to peak) including the overshoot. Discounting the overshoot the level is about 2.3V p-p, much higher than the .5V p-p standard. Some DAC's will tolerate this and some will likely have trouble. Mine both seem to be tolerant of the high level, insofar as they do not fail to operate. John's thought is that not only will the attenuators bring the output level down to normal levels, but have the additional benefit of attenuating any signal reflections. RF attenuators have been around a long time but are not typically used for this purpose as the output level of most S/PDIF sources would not allow or require their use. The logic seemed sound and I ordered up a few for the princely sum of $12 each. Not much money to risk on an experiment.
A few days later the attenuators arrived and after trying a few different combinations I ended up with a 6dB at the source and a 10dB at the DAC. These have a 75ohm BNC connection and just go inline with the cable. While this reduced the signal level to .4V p-p neither DAC had a problem with the slightly low voltage and the results were quite amazing. While I was much happier with John's mod hiFace over the stock version, the addition of the attenuators took it to a different level. Much more analog sounding, in particular how the music just seemed to flow, moving from note to note effortlessly. I don't really know of a better way to describe the effect, words like rhythm and pace are normally used but flow is what keeps sticking in my head. Additionally there was more low level detail and those little ambient cues get you a little closer to the original event. Too often increased detail comes with a forwardness and/or hardness but in this instance I noted just the opposite; a stridency which had always been present with the hiFace just went away and the sound stage became wider and deeper. Quite amazing for a $25 investment.
Employing the attenuators with a standard hiFace, the effects were much the same but the battery mod hiFace was still in a different league and in fact the system had never sounded better. I was very interested in trying to correlate the improved sound with something measurable. While not having the equipment to make jitter measurements I do have a 400MHz oscilloscope, which has no problem showing the waveform.
The hiFace was connected to the 'scope via a 75 Ohm BNC cable from the hiFace to a 75 Ohm tee adapter terminated with a precision 75 Ohm terminator and set the 'scope to 1Mohm input impedance. While not a perfect termination it was pretty close and as good as I could do. The results were quite interesting; the mod hiFace shows about a 50% faster rise time on the leading edge of the square wave. It also damps the ringing quicker after the initial overshoot. The leading edge of the square wave with the battery powered unit is nearly a straight line while the standard hiFace has a couple of kinks in the line. These can be seen pretty easily if the 'scope photos are observed side by side.
Adding the attenuators made no improvement to the rise time but did reduce the ringing. This makes sense considering the reduction in overall signal. As mentioned, I don't have a way to measure the jitter so correlating the improved signal to a reduction in jitter isn't possible in this instance but having a cleaner and faster transition from 0 to 1 certainly seems like it would make the DAC's job easier. It was nice to observe that the improvements I was hearing were at least related to measurable differences between the units.
How much did I like the battery powered hiFace? Very much, enough to buy the review sample. While very much a DIY type John's asking price is very fair and saved me the trouble of procuring the parts. The improvement of a standard hiFace, especially with the addition of the inline attenuators is considerable and worth every penny. There was concerned that having the charger connected when playing might create some additional noise or otherwise degrade the signal even though it's disconnected when playing. In the system no problems were observed. Connecting the unit back to the 'scope not difference in the output waveform was observed when connecting/disconnecting the charger.
M2Tech hiFace Evo
I was lent the initial demo unit for the US distributor for review and am very thankful for the opportunity to spend time with the Evo. When he received the Evo a power supply was not included and rather than cobble something together I called my friend Wayne Waananen at BOLDER Cable in hopes that he might have something suitable to borrow since a great deal of his business is doing modifications and building power supplies. Luck was on my side and he mailed a 9V ("standard" model, $475) unit out the next day. I also contacted Mike at Tweekgeek, the US distributor about getting the power supply that would be shipped with the production units. In short order a 9V regulated wall wart supply turned up in the mailbox. Thankfully this was a linear type instead of a switching power supply. Not quite as green but generally quieter that most switchers, something of great benefit in a audio system. One thing mentioned in the Evo manual is the option of battery power and since the battery powered hiFace was so good some of the A123 lithium cells were ordered along with an appropriate charger.
Since the Evo was brand new I wanted to put some hours on it before doing any serious listening so everything was hooked up and foobar set on repeat for a week or so. One nice thing is that the Evo uses the same driver as the hiFace, so no changes were needed to the computer. While on the subject of drivers, I am using this under Windows 7 and have also used it with Windows XP, both with no problems. There are threads in various forums about driver problems with Macintosh computers with the early drivers. I believe that most of those problems have been taken care of.
My first real listen with the Evo in the system was with the BOLDER power supply and an immediate smile crossed my face. The sound was very nice and quite similar to what I had been hearing from the JKM2hiFace though just a touch less of most characteristics, nothing specific just not quite as natural as John's modified hiFace. Once the power brick and batteries arrived I listened to the Evo with the various power sources. The supplied wall wart is OK but using a battery supply or a really good linear supply such as the one from BOLDER allows the Evo to reach its full potential. Much to my surprise I preferred the sound of the Evo when powered from the BOLDER linear supply over the LiFePo batteries. The linear supply offered a bit better dynamics, bass and presence but they were very close. Of course the linear supply avoids the maintenance and potential problems of batteries.
I was of course curious to see how much better the BOLDER linear supply was over the standard power supply. Each was connected to the oscilloscope loaded with a 100R resistor, providing a 90mA load, not high but enough to load them a bit and it was handy. The BOLDER supply had about 3.1mV RMS of ripple while the wall wart had 56.5mV RMS of ripple plus some truly ugly spikes, perhaps from the switching of the diodes. Net result being the BOLDER was about 18x quieter. Not inexpensive but the improvement in sound was reflected in the measurements.
Ultimately I preferred John's hiFace but it was very close. The Evo has much greater flexibility with regard to connections, does not require charging and vigilance over the batteries and the use of the RF attenuators. For some people the need to charge a battery powered device may not be worth the improvement in performance, in my case it is. Certainly if one is using a standard hiFace, the Evo is a significant step up and highly recommended. One caveat is that while the performance with the wall wart is OK, much improved results can be had with a better supply.
Halide Design Bridge
Sitting down to listen a few days later confirmed the first impression, still lovely and very musical. The greatest praise I can offer is there was no desire to remove it from the system and put something else in and that is exactly the case, it stayed in the system for the next few weeks and was the source of much musical enjoyment. When trying come up with some descriptions of the sound my notes are filled with items such as focused, nice bass, lovely bass, lovely background vocals, nicely delineated sound stage, great pace and flow, nice depth, great tone. I think you get the picture, the system sounded fabulous with the Halide Bridge. Unfortunately the hiFace Evo had to leave just as the Bridge arrived so there was no opportunity for comparison; I think they would have been very close in performance. Price wise it's nearly a wash, the Bridge is a little less expensive and eliminates the need for an S/PDIF cable and a USB cable. Overall I found the Bridge to be very nice and would not hesitate to recommend it, especially for those less computer savvy as almost no configuration is required.
Choices... Choices... What To
John Kenny Mk1 boxed hiFace
M2Tech hiFace Evo
Halide Design Bridge