From: "Saved by Internet Explorer 11" Subject: USB AUDIO | H i F i D U I N O Date: Thu, 2 Jul 2015 14:38:38 -0700 MIME-Version: 1.0 Content-Type: multipart/related; type="text/html"; boundary="----=_NextPart_000_0000_01D0B4D4.C856C2E0" X-MimeOLE: Produced By Microsoft MimeOLE V6.1.7601.17609 This is a multi-part message in MIME format. ------=_NextPart_000_0000_01D0B4D4.C856C2E0 Content-Type: text/html; charset="utf-8" Content-Transfer-Encoding: quoted-printable Content-Location: https://hifiduino.wordpress.com/category/usb-audio/ =EF=BB=BF
=20DIYINH= K=20 has developed a new version of the XMOS-based USB to I2S Audio = Interface. This=20 version provides isolated I2S signal lines. Plus it uses superior=20 oscillators.
You can read about the older version here [link]
ISOLATOR
The Isolator chip is the Silicon Labs 8661 Low Power 6-Channel = Digital=20 Isolator [link].=20 This is the newest device from Silicon Labs. It is based on RF coupling = [link].=20 Apparently the Si-Labs isolators are the lowest cost ones [link]
Si86xx enhancements include:
The =E2=80=9Cclean=E2=80=9D side of the isolator requires a separate = power supply which must=20 be provided for proper functioning. You need to provide a 5-16V supply. = This=20 will be further regulated by a low noise regulator.
OSCILLATORS
According to DIYINHK, this version uses NDK NZ2520SD Ultralow phase = noise=20 oscillators from Japan. There are 3 oscillators: the 48 MHz for the = digital=20 circuitry of the XMOS device, the 45.158MHz and 49.152 MHz for the audio = frequencies.
NDK has some general information on their oscillators, including = jitter data=20 here: [link]. Below=20 is the phase noise plot of the NDK oscillators. Note the phase noise = curve of=20 the =E2=80=9CSD=E2=80=9D models.
NDK=E2=80=99s NZ2520SD low-phase-noise crystal oscillator has phase = noise=20 characteristics superior to those of the general NZ2520SB model. The = company=E2=80=99s=20 high-accuracy crystal oscillator (OCXO) has even better phase noise=20 characteristics near the reference frequency.
Here is the phase noise plot of the Xpresso oscillators used in the = previous=20 version
And here is the phase noise plot of the CCHD-957. Note the phase = noise value=20 of the CCHD-957 at 10 Hz (-97) vs the phase noise value of the NDK = oscillator at=20 10 HZ (-113)
MASTER CLOCK
The Master Clock is available straight from the NDK oscillators or at = the=20 appropriate scaled-down frequency through the output pins.
Straight from the oscillators (45.158MHz and 49.152 MHz) through a = U.FL=20 socket:
Through the output pins (22.5792Mhz/24.576Mhz).
The =E2=80=9CC7424Z=E2=80=9D device is used as a clock divider to = generate the 22.xx and=20 24.xx frequencies used by XMOS device and also used by the downstream = device=20 (the DAC). Seems the clock line to the XMOS device is also isolated in = order to=20 prevent any kind of noise leakage from the XMOS/USB side to the = Clock/Clean=20 side.
Note that there is no =E2=80=9Cre-clocking=E2=80=9D of any of the I2S = signals. However, Xout=20 can be used to control an external re-clocking board for the I2S = lines.
COMPLETE ISOLATION
As we=E2=80=99ve seen, the =E2=80=9Cclock section=E2=80=9D including = the I2S signal lines is=20 isolated from the =E2=80=9CXMOS/USB section=E2=80=9D. The master clock = line generated by the=20 local clocks that connects to the XMOS device is also isolated. The two = sections=20 are powered independently.
On the backside of the board, we can see that the ground planes are = also=20 separated between the =E2=80=9Cclean side=E2=80=9D and the = =E2=80=9Cdirty side=E2=80=9D:
USB POWER
The XMOS side (or =E2=80=9Cdirty side=E2=80=9D) is powered by USB by = default. Just like in=20 the earlier version, the USB power can be bypassed by removing the FB1 = ferrite=20 and external 5V power can be applied through the external power = connectors.
REPROGRAMABLE
For the advanced DIY=E2=80=99er, the XMOS device can be reprogrammed = with different=20 firmware. This can be accomplished through the CN6 connector (which is = the JTAG=20 interface for the XMOS chip):
The following device from XMOS, the xTAG-2 [link]=20 can used to upload new firmware to the XMOS device (updating firmware = through=20 USB is not enabled in the default firmware). The xTAG-2 provides a high = speed,=20 low latency bridge between USB on the host and=20 fast JTAG connection (up to 10Mbps) on the target.
JTAG enables the transfer of data into non-volatile device memory = (uploading=20 firmware).
There are 20 pins in this device because the device is also a = debugger. The=20 JTAG interface pins are specified in the hardware manual [link]
While testing a CM6631A interface, I discovered that for 44.1K/48K = sample=20 rate material, the bitclock was running at 128FS (rather than the more = common=20 64FS). Thus when connected to a Sabre 32 DAC, it reported a sample rate = of=20 88.1K/96K. I described the behavior as =E2=80=9C2X bitclock=E2=80=9D [link].
2X BITCLOCK
BLCK is the same for both 44/48K and 88/96K. This means that for = 44/48K=20 sample rate, the data is running at 128fs and for the rest of the = sample=20 rates, it is running at 64fs.
Even though the Sabre32 DAC specifies BCLK to be 64 fs, it appears = that it=20 also supports 128fs. Why? because when I play 44.1KHz material, = the DAC=20 reports 88.2KHz sample rate and it sounds perfectly fine. (The sample = rate=20 reported by the Sabre32 DAC is based on the frequency of the = bitclock,=20 the DAC reports 88.2KHz for 44.1K material)
C-Media, announced the availability of a firmware tool that allows = the=20 selection of the bitclock frequency and other parameters [link]
It creates a hex file when you click =E2=80=9CGenerate FW=E2=80=9D. = The destination folder=20 must be the folder where the file =E2=80=9CUsbDac.bin=E2=80=9D (which = comes with the config=20 tool) is located
In order to =E2=80=9Cfix=E2=80=9D the =E2=80=9C2X bit clock=E2=80=9D = issue, you need to select x64 BLCK. For=20 the DIYinhk board, you also need to select =E2=80=9CSpeaker=E2=80=9D for = Terminal Type and =E2=80=9CI2S=20 Format=E2=80=9D
Then you need the =E2=80=9CFirmware update tool=E2=80=9D to upload = the s/w to the=20 CM6631A-based interface.
There is instructions included with the FWUpdate tool. If you need a = copy,=20 you can ask user tdtsai at diyaudio [link]
Once you update the FW with the x64 BCLK rate, the 44.1K/48K sample = rate will=20 correctly display in the Sabre32 DAC.
Diyinhk has a new version of this board that also takes advantage of = the=20 config tool to generate different versions of the firmware [link]
I have been contemplating modding the Airport Express since at least = 2009 [link],=20 but never got around doing it.
Now, adding an additional =E2=80=9CAirplay=E2=80=9D input to the = audio system is a desirable=20 option, especially for an =E2=80=9CiOS household=E2=80=9D like mine. = With multiple built-in=20 switchable inputs available in the Sabre-32 DACs which are easily = implementable=20 under software control [link],=20 there is even no need for an external switching device.
Certainly no modding is necessary since the AE already has a Toslink = output.=20 But for diyers, there is always something to mod :-)
I2S Output?
For this model of AE, tapping into the I2S signals is not feasible = since it=20 is based on the PCM2705 which only supports SPDIF output. (for I2S = output, you=20 would need a different PCM chip (PCM2706 or PCM2707) or use a newer = version of=20 Airport Express that uses an I2S based DAC such as the second and third=20 generation AE (the first generation =E2=80=9CN=E2=80=9D and second = generation =E2=80=9CN=E2=80=9D. The original=20 was 802.11 =E2=80=9CG=E2=80=9D).
Best Model for = Audio?
Reports seem to indicate that the original AE is probably the best = one to use=20 for audio and for modding=E2=80=A6
The original Apple Airport Express was introduced in 2004. According = to=20 Wikipedia,
The original version (M9470LL/A, model A1084) was introduced by Apple on June 7, 2004, and includes = an analog=E2=80=93optical= audio=20 mini-jack output, a USB=20 port for remote printing or charging the iPod (iPod Shuffle only), and = a=20 single Ethernet port. The = USB port=20 cannot be used to connect a hard disk or other storage = device.
The original AE (=E2=80=9CG=E2=80=9D version as in 802.11g) seems to = have the =E2=80=9Cbest=E2=80=9D audio=20 performance of all the AE models.
1- Tests published at CA shows that the =E2=80=9CFirst Gen N=E2=80=9D = version (the follow-on=20 version to the =E2=80=9CG=E2=80=9D version) has better audio than the = current =E2=80=9CSecond Gen N=E2=80=9D=20 version of AE [link]
=E2=80=A6To make a long story short, yes, the AE First Gen will = have a lower jitter=20 and better sound with your outboard DAC, than the AE Second Gen. An = outboard=20 DAC can only suppress incoming jitter, but can never = remove.
2- The =E2=80=9CG=E2=80=9D version in turn has better audio = performance than the =E2=80=9CFirst Gen=20 N=E2=80=9D version of AE:
I have both a wireless G model & a wireless N model. My = DacMagic=20 frequently lost lock when used with the N but is rock solid with the = G. At the=20 time I was struggling with this issue, forums and such indicated that = the=20 jitter in the N model was much higher than the = G=E2=80=A6
OPENING THE AIRPORT EXPRESS
For reference here is the =E2=80=9Coriginal=E2=80=9D Airport Express = take-apart post: [link= ]
The two halves of the Airport Express are bonded together with some = kind of=20 crazy glue. The only option is to cut it apart. But rather than brute = force, I=20 preferred an approach with some finesse=E2=80=A6:
Several passes were required in order to gauge the appropriate depth = and=20 avoid cutting into any internal components. In particular the power = supply wires=20 are butting against the casing.
The case can still be used after cutting the the case apart. Apple is = a=20 master in industrial design. We=E2=80=99ll try to save the case.
Inside the Airport Express.
Certainly a cleaner approach than what iFixit did :-)=20 [link]
The switching supply was made by Samsung in their China Dongguan = factory=20 (Dongguan is about an hour drive from ShenZhen). Nowadays, Samsung is = likely not=20 building =E2=80=9Clow-end=E2=80=9D components anymore=E2=80=A6
THE AUDIO BOARD
The AE is partitioned into two parts: the network board and the audio = board.=20 The audio board is a daughter card that plugs into the network = board.
Removing the RF shield:
Removing the audio board:
The audio board in detail. Even for a simple DAC implementation such = as this,=20 the board is a 4-layer PCB.
The PCM2705 USB DAC
NETWORK BOARD
That is a lot of electronics in a $99 package!
COMPARING WITH OTHER AE
Here is the 2nd Generation (also knows as the first generation = =E2=80=9CN=E2=80=9D) AE. This=20 version connects to the on-board DAC chip with I2S
The DAC chip used in the second generation AE is the CS4344 (the = 10-pin chip=20 in the photo below)
Here is the 3rd generation AE (current generation) courtesy of Rogue=20 Amoeba
SOME EXPERIMENTS
Functionality without the = audio=20 board
I tried powered up the AE without the audio board in place. It turned = out=20 that the networking capabilities of the device work fine, but its = Airplay=20 capabilities were disabled. Seems the AE performs a power-on = self-test=20 and disables the Airplay capability if it finds the audio section not=20 functioning.
PS Wiring
According to iFixit, [link]=20 the 3.3v and 5v supply wires are as follows:
Functionality without the = 5V=20 supply
I removed the red wire from the connector to test whether the device = would=20 operate without 5V. (If this is the case, then replacing the power = supply would=20 be easier since we would only have to worry about a single supply -the = newer AE=20 only require 3.3v for operation)
Result:
The network part of AE works without the 5V supply. Now, lets see if = putting=20 back the audio board will enable Airplay=E2=80=A6 Airplay does not = work without=20 the 5V supply.
WHAT TO MOD?
Power supply = replacement
The standard mod has been the power supply mod, replacing the = switching dual=20 supply with a linear dual supply. It is the most popular mod on = the AE. In=20 fact, there are commercial versions Apple Express with vastly enhanced = power=20 supplies such as La Rosita Alphas:
=E2=80=A6and the Micromega Streamer products:
Notice the unique design of the Micromega: the antenna of the AE = extends=20 outside of the case for better reception (I suppose if you totally = enclose the=20 AE the WIFI reception would be very poor)
=E2=80=9CMONSTER MOD=E2=80=9D
I found this project in the Vietnam AV Network. The user replaced the = clock,=20 power supplies, and output capacitors, plus a isolation transformer in = the spdif=20 output [link]
USB AUDIO OUTPUT MOD
I have not seen another reference to this mod. So I think it is the = first=20 :-)
What if we can take the USB output, and feed it to an external USB = interface?=20 We can then try different devices such as the PCM2706 with both SPDIF = and I2S=20 output, or the TENOR TE7022 with similar capabilities. We can add=20 isolation. Isolating the USB will greatly clean up the digital = signal and=20 isolate the more sensitive audio part from any digital noise in the AE,=20 including noise generated by the network/wifi chips and related = parts.
No need to replace the = original=20 PS?
Since we tapping into the USB Audio stream, a purely digital signal, = there=20 may be not much to gain in replacing the original PS, especially since = we cam=20 add a USB isolator for full isolation. In addition, being a = communication=20 device, we can be assured that the designers have done their work in = ensuring no=20 (at least unrecoverable) errors in the digital domain.
Undoubtedly, a PS mod would benefit the analog performance and even = perhaps=20 the conversion to digital audio, but for this mod, I think am going to = leave the=20 original PS in place.
Looking closely to the input power connector, we can see that the = designers=20 have taken good care in filtering out the noise from the switching = supply. Aside=20 from the EMI shield covering the components, we find that the power = lines and=20 ground are further isolated with ferrites:
Tapping to the USB = lines
The USB data from the AE audio board is =E2=80=9Creadily = available=E2=80=9D. There are 4 vias=20 where the the 4 USB signals can be tapped:
We can even leverage the built-in USB connector by first cutting the = existing=20 lines (which are only good for connecting to a printer) and tapping the = USB=20 signal from the Audio board. 5V and GND are already there, so only two = wires are=20 needed:
Here is the mod. I tried to move the wires away from any active=20 circuitry. Zero cost and completely and easily reversible.
Cut the original signals to the USB connector (which we won=E2=80=99t = use anymore=20 since this is for printer sharing)
The metal shield covers everything including the USB wires
TESTING THE MOD
I used the iBasso D2 USB Portable DAC/AMP to test my mod. The USB = interface=20 is the PCM2706 and connects to the Wolfson WM8740 DAC. The PCM 2706 is = part of=20 the PCM 270x family, so it should work=E2=80=A6
Results:
Airplay complains that there are no speaker connected to the AE. The = device=20 is just too smart. It senses if there is a plug in the output connector = and=20 prevents Airplay from connecting if there is nothing plugged in=E2=80=A6 = No problem. Use=20 any mini-plug into the output port.
Now Airport connects, I first try to to connect a portable speaker to = ensure=20 that everything is in working order. I get sound in the = speaker=E2=80=A6So far so=20 good.
Now I connect the iBasso into the USB port of the AE=E2=80=A6 = Nothing=E2=80=A6 The iBasso has=20 a =E2=80=9Clink=E2=80=9D LED and it is not lit. No connection to the = iBasso=E2=80=A6
OK, perhaps I need to reboot the AE with the iBasso = connected=E2=80=A6 Reboot=E2=80=A6=20 Airplay does not show up in iTunes=E2=80=A6 This AE is just too smart. I = figure that=20 since there are two USB interfaces connected, the AE can=E2=80=99t = figure out what is=20 going on (there is no handshake/link to any of the two DAC chips) then = the self=20 test fails and the AE does export itself as a functional = device=E2=80=A6
Back to bench.
I theorize that the firmware in the AE is looking to one USB DAC to = link up=20 with. I would make sense to turn off the on-board DAC chip and allow a = single=20 device to hang off the =E2=80=9CUSB port=E2=80=9D.
Pin 7 is the VDD (digital power) for the PCM2705. We can just lift = the=20 pin.
I didn=E2=80=99t even have to use the soldering iron. Just an exacto = knife and a bit=20 of pressure. There is also analog power, but I won=E2=80=99t bother = lifting those.
Result: Yes! it works. After booting up the AE with the iBasso=20 connected, immediately the link LED lights up. Airplay shows up in = iTunes=E2=80=A6 Play=20 a song=E2=80=A6 Ahhh=E2=80=A6 Music!
Took advantage of the new beta firmware for 384K support [link]=20 and tried in the diyinhk CM6631A USB interface.
According to tdtsai,
CM6631 do not have this capability to support 384k/352.8K. So we = can=E2=80=99t=20 change firmware to support it. By the way if you want CM6631A to = support=20 384k/352.8K you need use 45/49 pairs external clock.
The diyinhk interface has both the CM6631A chip and also the higher = speed=20 clock pair.
UPDATING FIRMWARE
I received both firmware and update tool from Mr. tdtsai (see the = diyaudio=20 link above)
Current firmware
New Firmware
To play 384KHz material I used quanghao=E2=80=99s DAC (since it = requires 100 MHz=20 clock and my Buffalo DAC has an 80 MHz clock)
BTW, users have started to receive the =E2=80=9Cproduction=E2=80=9D = model of the group buy=20 [link]
RESULTS
Well, it plays as advertised. More and more interfaces are moving up = the=20 sample rate capability=E2=80=A6
Tried both ASIO and WASAPI. Here is the WASAPI setting in foobar:
Here is the ASIO setting in foobar:
It is a great time to be an audio diy=E2=80=99er. There is currently = great=20 availability of quality boards aiming at providing the greatest fidelity = with=20 incredible VALUE.
Here are some side by side photos of the two USB-I2S interfaces I = own.
Worthy of mention is the upcoming next generation Wave IO board. Mr. = Lorien=20 posted a sneak peek at his next generation board [link]
From the look of the layout, this board has electrical isolation of = the=20 outputs and flip-flop reclocking after the isolator.
Updated 4/30/13
The new Teac UD-501 DSD capable DAC has received good reviews = everywhere. Was=20 curious about its USB interface. Here is a photo taken from a French = site,=20 qobuz.com [link]
There is a TMS320 chip and an unidentified chip. It is possible the=20 unidentified chip is the USB receiver or the TMS320 is doing receiver = function=20 and the unidentified chip is a microprocessor to control the TMS320 = chip.
The TMS320 is a DSP chip [link]. It is a =E2=80=9CC67=E2=80=B3 = series, an up to 1GFLOP 32-bit=20 floating point DSP (the XMOS are rated at 500-1500 MIPS, likely = equivalent=20 performance). This family of DSP products was introduced in 1983 and new = models=20 have been introduced along the way.
The =E2=80=9CC67=E2=80=B3 series is a current product and an EVAL = kit, in particular for the=20 C6748 chip, is available from TI [link],=20 [link]. According to = the=20 documentation:
=E2=80=A6The
TMS320C6748, a low-power dual-core applications = processor based on=20 a fixed-point C64x+=E2=84=A2 instruction set and the floating-point = C67x+=E2=84=A2 instruction=20 set. It provides significantly lower power than other members of the=20 TMS320C6000=E2=84=A2 platform of DSPs and provides both floating-point = precision and=20 fixed-point performance in the same device. With a wide variety of = standard=20 interfaces for connectivity and storage, the C6748 development kit = enables=20 developers to easily bring audio, video and other signals onto the = board.=E2=80=A6 Included in the C6748 development kit is all the hardware = and software=20 needed for two demonstrations, a fingerprint-recognition demo and a=20 face-detection demo.
Here is the block diagram for the eval kit:
Looks like a very capable and general purpose processor for not just = audio=20 but a lot of other things not even related to audio. You can take a look = at the=20 WIKI for all the available libraries for the DSP [link].=20 The DSP can even implement audio decoding and filtering.
In addition, the USB-2 interface is provided by the FT232 chip (and = thus the=20 unidentified chip is likely not the USB interface chip; however, = there are=20 no FT232 chips in 48-pin package).
Further, the DAC has an upsampling feature and is provided by the = Cirrus=20 CS8422 chip on the main taking the I2S output from the USB board.
Thus the USB board, I believe, it is just moving the bits to = the DACs.=20 It looks kind of overkill; perhaps in some future Teac will add other=20 capabilities such such as PCM to DSD conversion and different upsampling = algorithm on the DSP chip. It is also likely that Teac is just reusing = hardware=20 from their TASCAM proline. A device such as the TASCAM=20 US-366 is a =E2=80=9CUSB interface with DSP Mixer=E2=80=9D
This design from TEAC gives credibility to what Musiland is doing in = their=20 upcoming =E2=80=9CSuperDSP=E2=80=9D chip. The good thing about the = Musiland product is that the=20 DSP is dedicated to audio, and the chip will have native support for USB = 2 and=20 USB 3 interfaces.
UPDATE
Appreciate Mr. Daussin, the reviewer of the TEAC DAC, providing = additional=20 information in the comments section. The unidentified chip is the = TPS65070.=20 Readers would recognize =E2=80=9CTPS=E2=80=9D being power chips. The = TPS65070 is a single-chip=20 with multiple voltage outputs. Here it provides the different voltages = required=20 by the DSP chip. It is a convenient, integrated solution to provide the = power to=20 the main chip instead of using separate regulators.
The TMS320 DSP chip has implemented USB 2.0 capability in s/w and = there is no=20 need for a FT USB receiver chip (In the block diagram the USB 2.0 = receiving=20 capability is provided by a separate chip because (I think) the DSP chip = is=20 already burdened with many different functions).
Updated 4.21.13
Follow the discussion of this board and XMOS technology in general in = the=20 diyaudio thread: [link].
The new board
The new chip from XMOS
Cannot tell from the chip markings what part is it. It is = =E2=80=9Cmissing=E2=80=9D the part=20 number. But it is definitely a =E2=80=9CU=E2=80=9D part as there is no = separate USB controller=20 chip. Here is the datasheet of one version of the part: [link]
The previous generation XMOS devices required a separate USB = interface chip=20 (foto from here: [link]):
The audio clocks. According to the datasheet, pin 1 is the enable pin = and pin=20 3 is the output pin. Looks like the enable pin is controlled by the XMOS = chip,=20 enabling/disabling the clocks for the corresponding sample rate.
The local ultra low noise LDO (excuse the left over cotton fibers = from my=20 cleaning :-)). You can bypass the USB power by removing FB1 and provide = external=20 5V.
The 2 LC filters for filtering the output from the built in DC-DC = converters.=20 Here is place to add larger caps and increase the filtering.
Solid ground plane. CN1 is for external powering (instead of using = USB=20 power).
If you read the datasheet and compare the topology of this board you = will=20 find that the manufacturer has followed the build and layout = recommendations=20 from XMOS. In fact it is so simple, that not following the = manufacturer=E2=80=99s=20 recommendations is kind of hard :-)
The new XMOS chip has several advantages over the old device
Apparently 384KHz support was added by diyinhk according to this = lively=20 discussion at diyaudio [link]
FEATURES OF THE MODULE (From the manufacturer)
Comparing the Xpresso clocks with Crystek.
DEVICE DRIVERS
I first checked it by plugging it to a Macintosh computer. Native = MacOS=20 supports up to 352K and 384K sample rate.
On the PC, it requires device drivers.
Version 1.63.0 is the latest driver from Thesycon.
No indication of 352K/384K capability for WASAPI shared mode. This = may in=20 fact a limitation of Windows mixer as it expects to resample every audio = stream=20 to the selected sample rate. However, in WASAPI exclusive mode, this = part (the=20 mixer) is totally bypassed and the output sample rate from the player=20 application is passed directly to the hardware.
The driver also supports ASIO
CONNECTING TO DAC
Connected the board to my old Buffalo II DAC (80 MHz clock) and = tested all=20 sample rates from 44.1KHz all the way to 384KHz. With the 80MHz version = of the=20 Buffalo II DAC, 352.8KHz and 384KHz work fine with oversampling turned = off. All=20 sample rates work as advertised. Modern designs are pretty robust, = especially=20 this version of the XMOS chip which requires only a few external = components.
The socket in the XMOS device matches the socket pin arrangement for = the=20 latest version of the $99 ES9018 board and a ribbon cable can be used. = For the=20 Buffalo simple cat-5 twisted pairs can be used.
Here is an implementation from a reader with diyinhk $99 DAC = board:
The XMOS is hiding=E2=80=A6
Straight from the eBay page [link]
I have to congratulate this Mr. diyinhk. He can non only produce = designs with=20 such speed, but can somehow price them so amazingly low. This new module = is=20 priced at $59.95 -Can=E2=80=99t resist not getting it!
FEATURES
WINDOWS DRIVERS PROVIDED (NO DRIVER REQUIRED FOR=20
MACS)
Looks like the first USB interface using the latest generation XMOS = chip (XCORE=20 USB) with built-in USB.
This is an update to this post on using foobar to play both PCM and = DSD files=20 [link]
Playing PCM and DSD files with foobar:
After installing foobar2000,
From the foobar2000=20 website, download:
UPDATES
Get to the =E2=80=9CSuper Audio CD Decoder=E2=80=9D repository [link]
When you extract the zip files you will find:
foo_input_sacd
This plug in shows up as =E2=80=9CSuper Audio CD Decoder=E2=80=9D and = allows you to playback=20 the following formats:
To install or update foo_input_sacd, just drag it to the components = window in=20 foobar and click =E2=80=9Capply=E2=80=9D
The control panel looks like this: you can select DSD output or PCM = output=20 (when converting DSD to PCM)
More on the different options: [link]
I tested the conversion of DSD to PCM, using both DSD64 and DSD128 = files. The=20 results were as expected:
PCM=20
Conversion |
Output Sample =
Rate for=20
DSD64 Input |
Output Sample =
Rate for=20
DSD128 Input |
44.1KHz | 44.1KHz | 44.1KHz |
88.2KHz | 88.2KHz | 88.2KHz |
176.4 | 176.4 | 176.4 |
352.8KHz | 352.8KHz | 352.8KHz |
foo_dsd_asio
foo_dsd_asio is kind of a =E2=80=9Cmeta-asio=E2=80=9D driver. It is = used by the sacd plug-in=20 above as the output device and in turn foo_dsd_asio outputs to an ASIO = driver=20 for an actual device. Apparently the sacd plugin does not output = directly to an=20 ASIO driver, it must go through the meta-asio driver. (Actually, ASIO = itself is=20 a meta-device driver because it talks to the actual device driver)
For one thing, foo_dsd_asio handles the =E2=80=9CDSD playback = method=E2=80=9D: for example,=20 =E2=80=9CDoP Marker 0=C3=9705/0xFA=E2=80=9D. =E2=80=9CDoP=E2=80=9D means = DSD over PCM. Marker 0=C3=9705/0xFA means use the=20 marker (for DSD) as specified in the proposed =E2=80=9CUSB Link for DSD = Audio via PCM=20 Frames=E2=80=9D open standard [link][link]. This = =E2=80=9Cmarker=E2=80=9D method is=20 predominantly driven by the MacOS since its built-in USB2 audio driver = only=20 supports PCM. On the PC side, there is no native support for USB2 audio = so=20 people use ASIO and ASIO can support both PCM and DSD streams. The = Amanero board=20 handles =E2=80=9CASIO native=E2=80=9D, but other boards may require a = marker on the DSD=20 stream.
Thus, foo_dsd_asio has the following functionality:
To install or update foo_dsd_asio, run the ASIOProxyInstall-0.6.x.exe = program.
I tested the PCM to DSD conversion. The=20
latest foo_dsd_asio plug-in handles all sample rates as shown in =
the=20
following results:
PCM Sample=20
Rate |
DSD64=20
Conversion |
DSD128=20
Conversion |
44.1KHz | 2.82=20 MHz | 5.6MHz |
88.2KHz | 2.82=20 MHz | 5.6MHz |
352.8KHz | 2.82=20 MHz | 5.6MHz |
96KHz | 3.07=20 MHz | 6.14MHz |
384KHz | 3.07=20 MHz | 6.14MHz |
AMANERO FIRMWARE/DRIVER UPDATE
Recall that previously ASIO4ALL was used because the Amanero lacked = native=20 ASIO support. Amanero updated the firmware and released ASIO = drivers.
For instructions on enabling native ASIO playback, you can follow = this=20 excellent post:
Amanero=20 ASIO or you can get the ASIO drivers from the Amenero site [link].=20 Note that the ASIO drivers compliment the standard drivers and you need = both for=20 ASIO support.
After updating the Amanero device with ASIO support you do the=20 following:
Fist set the output to be foo_dsd_asio (in order to support DSD = output, and=20 also PCM since this is the output device)
And then in the foo_dsd_asio configuration screen, select the Amanero = board.
TIPS FOR UPDATING FIRMWARE
Erase=20
firmware
1- Plug board into USB port
2- Short the pads as shown for at least 1 second (I used a paper = clip)
3- Unplug the board
Download and Install =
Atmel device=20
driver
1- Download and unzip the update tools from the Amanero web site [link].=20 You will get several files: a device driver for the ATMEL chip, a = configuration=20 application for updating the firmware and some other files.
2-Plug the board into a USB port. At this point the device is = completely=20 unidentified. You may get a prompt to install a driver or you may not. = Following=20 is the manual installation of the ATMEL device driver:
3- Open the Device Manger under the System control panel. The device = shows up=20 as =E2=80=9Cunknown device=E2=80=9D as shown below
4- Right click on the unknown device and select Update Driver = Software,=20 indicating the location of the driver which is in the folder you = downloaded.=20 Your Amanero board is now a =E2=80=9CAT91 USB to Serial = Converter=E2=80=9D
Your are ready to update=20
firmware
1- Ensure that you have erased the firmware as shown above, and = ensure that=20 the device is identified as =E2=80=9CAT91 USB=E2=80=9D as a port as = shown below:
Got a hold of DIYINHK=E2=80=99s C-Media based USB-I2S interface = a while back,=20 and it is just now I got around writing something about it.
This is based on the newest USB2 audio chip on the block, The C-Media = chip is=20 the CM6631A. This version in turn is the newer, the =E2=80=9CA=E2=80=9D = version of the CM6631=20 chip which has been implemented in several recent products. The CM6631A = differs=20 from the more common (and older) CM6631 in that it supports 176.4KHz = sample=20 rate. All other features are the same as the CM6631.
According to the product info page [link]:
USB 2.0 Asynchronous operation (every one does this nowadays)
Up to 192KHz / 32bit
Very capable set of input/output (the implementation reviewed here = only=20 implements USB input and I2S output)
I was able to find an older version of the datasheet for the CM6631 = which is=20 pin compatible with the DM6631A [CM6631_Datasheet_v0.8]
It can be seen from the specifications that a more capable system can = be=20 developed. For example, a 4-channel I2S output driving two DACs. = (allowing=20 performing digital crossover functions in the PC)
DIYINHK USB Interface: CM6631A (with the =
=E2=80=9CA=E2=80=9D)
Top
Bottom. Uses the CM6631A
DEVICE DRIVER
The vendor provides a device driver [li= nk]
Other sources: the latest version of the driver is a unified version = for the=20 CM6631 and CM6631A devices, thus available from other manufacturers = using either=20 the 6631 or 6631A part (likely they all offer the same versions. some=20 manufacturer may be faster than others in offering newer versions):
I tried the Emotiva provided driver and it works. Here is a screen = shot:
FIRMWARE
There are two versions of the firmware as discussed here [link]=20 and here [link].=20 However, these versions seem to be related to the SPDIF features of the = chip.=20 For this board, which only supports I2S, I don=E2=80=99t think these = versions of the=20 firmware matters. Note: according the tdtsai [diyaudio], the developer = for the=20 driver s/w, the firmware for the CM6631 is not compatible for the = CM6631A.
SALIENT FEATURES
Two ultra low noise regulators: the TI LP5900 [link]
Uses high frequency 45.1584 MHz and 49.152 MHz oscillators to derive = the=20 audio frequencies (the manufacturer indicates that these are sourced = from=20 Digikey, thus ensuring their quality). This is unique as most devices = use half=20 that frequency or lower.
No markings? According to the datasheet, this seems normal. The = product=20 number is in the box or reel.
The oscillators output are available to be used as master clock for = the DAC.=20 The higher speed oscillator are in the =E2=80=9Csweet spot=E2=80=9D = frequencies for ESS Sabre=20 DACs for synchronous operation.
Since the board does not switch the clock lines, you can only select = one of=20 the clock lines to feed the ESS DAC. In this manner, with some of the = sample=20 frequencies the DAC will operate in synchronous mode and with other = sample=20 frequencies, in the normal asynchronous mode.
For example, if you chose to use the X45 line (45.1584 MHz) as the = clock for=20 the ESS DAC, then when playing 44.1K, 88.2K and 176.4K the DAC will = operate in=20 synchronous mode and when playing 48K, 96K and 192K, the DAC will = operate in=20 asynchronous mode.
You could manually switch the clock lines but this is not only = impractical,=20 but it could also upset the DAC requiring a reset.
Easy bypass of USB power. CAUTION: Maximum input = voltage for=20 the local regulators is 5.5V
According to the manufacturer, the design employs solid ground plane = (a must=20 for high speed digital circuit) and no vias in active circuit (via = inductance=20 always create jitter problem). This is apparent from the photo below and = in=20 comparison with other designs. The overall layout is very clean and = compact.
Datasheet: [link]
Jitter
The datasheet for the Xpresso clocks is actually very extensive, and = the=20 jitter measurement is equally extensive. The phase noise plot is = included. I=20 have shown here the comparison with the Crystek CCHD-950-80 MHz which is = the=20 oscillator I have in my version of the Buffalo II DAC:
Using the 62,5 MHz curve, we get a phase jitter value of 6 psec RMS = (10 Hz =E2=80=93=20 1 MHz). The Crystek CCHD-950-80MHz has a jitter value of about 2 psec = RMS.=20 However, at the lower offset frequencies (the ones of interest to audio=20 performance) the phase noise is not very different from that of the = Crystek=20 clock which is a good thing.
INTERFACING WITH BII DAC
There is something =E2=80=9Ca bit odd=E2=80=9D with this interface: = According to the=20 specifications:
DATA | LRCK | BCLK | MCLK |
16/24bit | 44.1kHz | 5.6488Mhz | 11.2896Mhz |
16/24bit | 48kHz | 6.144Mhz | 12.288Mhz |
16bit | 88.2kHz | 5.6488Mhz | 22.5792Mhz |
16/24bit | 96kHz | 6.144Mhz | 24.576Mhz |
16bit | 176.4kHz | 11.2976Mhz | 22.5792Mhz |
16/24bit | 192kHz | 12.288Mhz | 24.576Mhz |
2X BITCLOCK
BLCK is the same for both 44/48K and 88/96K. This means that for = 44/48K=20 sample rate, the data is running at 128fs and for the rest of the sample = rates,=20 it is running at 64fs.
Even though the Sabre32 DAC specifies BCLK to be 64 fs, it appears = that it=20 also supports 128fs. Why? because when I play 44.1KHz material, = the DAC=20 reports 88.2KHz sample rate and it sounds perfectly fine. (The sample = rate=20 reported by the Sabre32 DAC is based on the frequency of the = bitclock, the=20 DAC reports 88.2KHz for 44.1K material)
The sample rate for the higher sample rates are reported correctly by = the=20 ES9018 DAC.
Is this a =E2=80=9Ccommon=E2=80=9D feature supported by other = DACs?
The AK = 4399=20 supports both a 64fs and a 128fs bitclock:
The Wolfson=20 8804 also supports 128fs Bitclock -in DSP mode (p. 44):
So it seems that a 128fs bitclock is not so strange after all. I am = not sure=20 if any other dacs support 128fs bitclock, but the Sabre32 DAC definitely = does.
OTHER NOTEWORTHY IMPLEMENTATIONS
C-Media has been in the computer audio business since its inception.=20 According to their website, their world=E2=80=99s= =20 first accomplishment are all related to audio. Its claim to fame is = probably=20 the ASUS XONAR series of PC audio cards and interfaces. The latest ASUS=20 XONAR ESSENCE ONE uses the CM6631 part as shown here (There is = a review=20 of the Essence One here: [link]):
The Schiit=20 DACs also use the CM6631 part for their USB option board. As = discussed here=20 [link] it does not support 176K material because of hardware limitations = in the=20 6631 part. The =E2=80=9CA=E2=80=9D part supports 176K sample rate.
The Emotive XDA-2 reportedly also uses a CM6631 chip for the USB = interface=20 [link]
The MHDT USB Bridge also uses the CM6631 chip [link]
Have yet to see any commercial implementation using the CM6631A = part.
NEW VERSION
diyinhk turns his designs very quickly. This version of the board has = been=20 replaced with a new version having isolated I2S [link]
MAC USB 2.0 COMPATIBILITY
There are some reports and fixes to ensure USB2 compatibility with = Macintosh=20 computers: [link]=20 with certain CM6631-based implementations.
This board does not suffer from the reported problem: there are no = 3-pin=20 devices near pin 98 of the chip.
Plugin this device into a Macintosh computer shows that it is indeed = a high=20 speed USB 2.0 device as shown in this snapshot of the CMEDIA device = under the=20 USB description screen
JITTER IMMUNITY TEST
Here is a reported test of the CM6631A (implemented in a different = device)=20 showing jitter measurements with and without processor load and = comparing the=20 asynchronous nature of the USB communication vs a device using USB = adaptive=20 communication. [link]
In his conclusion, this device (as well as other USB-asyncrhonous = devices)=20 show great immunity against processor loading:
Bottom Line:Don=E2=80=99t worry about jitter! = It=E2=80=99s more than=20 likely inaudible in a modern computer system and with decent (not = necessarily=20 expensive) audio gear. I see no evidence that high CPU/GPU load makes = any=20 difference to jitter. Isolating your DAC from electrical noise = polluting the=20 analogue output seems much more important.