The new DAC features improved clocking (making use of the new “audio grade” “SDA” oscillators from NDK), improved powering scheme (including the addition of a USB type C port for dedicated external power as well as a new layout), and improved filtering capacitors in the output stage.
In order to do a fair comparison, I had to keep all external parameters the same. That meant using two identical Raspberry Pi 3 boards, running the same distribution (Archphile), with same settings, powered from separate but identical power supplies, using the same audio cables, into 2 inputs of my preamp.
Both of the DACs had run for several hours (or even days..) as a “burn-in period”.
Power was delivered through a couple of identical custom-made USB power-only cables and data came in from a file server located several rooms away, through a single ethernet switch.
Control was through Archphile’s ympd web-based MPD client.
I sync’ed the playback of the RPis as best as I could and I set about switching between the two DACs by selecting either one or the other input in my preamp. The volume needed no adjustment – the two DACs output the same signal amplitude (not measured but fairly certain..).
I used a small number of test tracks, representative of both high quality recordings as well as contemporary pop music.
After some considerate going back and forth, I came to a conclusion. The new Boss 1.2 is a definitive improvement on the original Boss. The differences are not night-and-day, but they are there.
The 1.2 manages to have better bass extension, without sacrificing control. It also does a better job of placing the instruments in space, while also managing to sound more realistic – the vocals sound more real, the acoustic guitar sounds more like an actual guitar, and so on.
All of these “symptoms” are classic signs of less jitter being present in the system. So Allo.com’s choice of new audio-grade oscillators and the changes they made to the power supplies have paid off.
Good job keeping the title of “best VFM DAC HAT”. 🙂
Pretty much everyone agrees on this one, still, it took me a while to get on the bandwagon. Probably because I am not using an RPi as my main transport..
In order to test this I needed a proper linear power supply, capable of outputting at least one “real” amp at 5V. It would also have to be as low noise as possible. After some searching I ended up picking the TPS7A4501 for the job. Since its output voltage is adjustable, it would also come in handy for various projects. So I designed a PCB that would do a proper job of “hosting” the regulators along with the necessary rectification and filtering stages and had a bunch of them made.
I also ordered a proper custom-made toroidal transformer, one that would have 5 secondary windings, each of them outputting 6VAC at 1.5A and one sixth winding outputting 3V at 2A. The end game would be to replace all of the RPi’s on-board switching regulators with linear ones.
But for starters I’d just power the RPi with 5V via its USB port. Here are a couple of RPi 3s being powered by my linear supplies. This setup would be used to compare the Allo.com original Boss with its new “v1.2” version. More on that in another post.
Power draw on the linear power supply was measured to be at about 500mA @ 5V.
My “reference” 5V SMPS is this one:
It’s an old but beefy switching mode power supply from an old HP tablet. It’s specc’ed at 5.3V @ 2A.
In order to be as impartial as possible, I took the setup to a friend’s house and had him and a couple of other friends audition the RPi powered either from the SMPS or by my linear power supply.
In both cases, the RPi was running Archphile and was connected to my upgraded-with-ES9028Pro-and-Mercury-Buffalo III DAC via USB.
The difference between the two power supplies was immediately obvious. It was like with the SMPS we had an at-best mediocre source – DAC combination, while with the linear power supply the setup became “proper”, it sounded “in-place” among my friend’s high performing system. The sound stage became better defined, the detail level went up, overall the presentation was more realistic. In other words, it was like the jitter of the system went down, but perhaps the noise levels in the system also decreased.
In other words, IMHO no serious audiophile should be powering his RPi by a run-of-the-mill 5V SMPS, even if he is using it as “just a USB transport”.
The only parts that need matching are T8A with T9A and T8B with T9B. There’s no need to go crazy with the matching – within 5% should be enough.
Power should be ideally +/-16VDC. A bit less is OK (I did my initial testing with +/-12VDC) but more will most likely damage the board. The board is running in class-A so current draw is constant. A power supply with 100mA current capacity should be enough.
Bias current is adjusted by the multi-turn trimmers R26A and R26B. They should be adjusted to their mid value before soldering to the board (~1K). To adjust bias just measure current consumption at one of the rails while turning the pot. Adjust for ~25mA total current draw per rail and per channel. Current draw on the negative rail should be about 1mA higher than on the positive rail. Bias should be re-adjusted if the power supply voltage needs to change.
After bias adjustment and with no input signal you should check for DC at the outputs. If everything went well you should be seeing anywhere between 0 to a few mV of DC voltage.
A few design notes:
This design is inverting. I’ve set up my AK4490 code to also invert the 4490’s outputs so as to end up with a non-inverting overall output. But it doesn’t seem to be making much of an audible difference, since I can’t hear a difference as I’m switching between inverting and non-inverting output.
There is provision for a relay that switches between the default filter for PCM (-3db @ ~90KHz) to a more proper filter for DSD (-3db @ ~50KHz). This feature has not been tested yet..
Regarding the resistors, we’ve chosen to go with mini MELF resistors (type 0204) because of their high quality and in general excellent reputation for audio. We are aware that some of the necessary values ATM are kind of hard to source so we’ve included notes in the BoM with the really critical parts and valid value ranges for the less critical other values. Note that “+/-5% from initial value” does not mean that the parts tolerance can be 5% – it means that instead of for example a 10K resistor you can use a 10K + 5% = 10.5K part. It still has to be matched to its counterpart, of course.
I am including these substitutions, as reference:
Instead of the proposed 604R : —-> https://gr.mouser.com/ProductDetail/Vishay-Beyschlag/MMA02040C6190FB300/?qs=sGAEpiMZZMsU0eETUM64Jwu1lXnAfA1Az%2fUBmHPWxBQ%3d
Instead of the proposed 62R —-> https://gr.mouser.com/ProductDetail/Vishay-Beyschlag/MMA02040C5909FB000/?qs=sGAEpiMZZMsU0eETUM64JzhzDaxYafIopmV6xoU7Pd%2faEeFyYr39dw%3d%3d
Instead of the proposed 49,9R —-> https://gr.mouser.com/ProductDetail/Vishay/CMA02040X3909GB300/?qs=%2fha2pyFaduinSoUQ%252bxM%252bcspKJgxV2WgydeZC1mFkJQ0%3d
Instead of the proposed 499R —-> https://gr.mouser.com/ProductDetail/Vishay-BC-Components/CMA02040X4700GB300/?qs=sGAEpiMZZMsU0eETUM64J9XrXp4g00LTnadULizxluo%3d
Instead of the proposed 33R —-> https://gr.mouser.com/ProductDetail/Vishay-Beyschlag/CMA02040X3309GB300/?qs=sGAEpiMZZMsU0eETUM64J5DAE%2fHuiq02KfqnEeQjWns%3d
Instead of the proposed 10R —-> https://gr.mouser.com/ProductDetail/Vishay-Beyschlag/CMA02040X1009GB300/?qs=sGAEpiMZZMsU0eETUM64J5wRh77yguxgpX8F3yMYvKE%3d
The value of two of the required 1K resistors (the R20A & R20B) is critical, so you should use these —-> https://gr.mouser.com/ProductDetail/Vishay-Beyschlag/CMA02040X1001GB300/?qs=sGAEpiMZZMsU0eETUM64J5DAE%2fHuiq026nqxTapoB%252bc%3d
for the other 6 non-critical 1K resistors you can use these —-> https://gr.mouser.com/ProductDetail/Vishay-Beyschlag/CMA02040X1201GB300/?qs=sGAEpiMZZMsU0eETUM64J5DAE%2fHuiq02Tdrq8lX57XI%3d
The value of the 120R resistors is critical, so are the tolerance of the 49.9R (substitution for 39R) resistors. Their values should be matched.
This concludes Part 3. The next post will detail how everything works together.
Following up on Part 1, it’s time to talk about the “brains of the operation”. Bare with me, this is going to be a rather long read.Hardware selection
The DAC needed to be controlled by a microcontroller so I looked into my options. I wanted something that would:
Be easy to program, so the Arduino IDE was a must.
Be able to drive a TFT.
Have enough storage capacity to store enough code & fonts for the TFT.
Be readily available.
Be relatively low cost, since it would have no influence on the dac’s SQ.
Be easy to build / integrate into a new design, even by a novice.
After some consideration, I decided to use an STM32F103 ready-made module. It would plug in to a “mainboard” of my design, along with the chosen TFT. It would be fast enough, have enough flash & ram, be easy to integrate and develop for and it would cost next to nothing.
Next up was the TFT. I’d seen on Ebay an interesting one that was relatively big, high resolution and inexpensive. It was this one:
It can be found on Ebay by searching for “3.5 tft uno 320 x 480”. Expect to pay 6-8€ inc. shipping.
After some searching I found a suitable library that (after some slight tinkering) would allow my tiny STM32 to drive it properly. (Note: do not download the library from this link. I will provide a customized version of the library together with my code).
I found a ready-made library for configuring the Si570 and modified it to run on my STM32, using one of its hardware I2C ports. I will also include this library in my code. To complete the recipe I also found working rotary encoder and IR receiver libraries.
Next up was the prototyping work. I adapted my TFT HiFiDuino Pro code to run on the STM32 & TFT combo, with support for AK4490 dual mono operation. The end result had this feature list:
Support for either Dual Mono or single chip setups.
Support for the Amanero Combo384 USB to I2S module (must be set up as slave with MCLK/2 and F0,1,2,3 enabled).
Control with one rotary encoder with push-to-select functionality.
IR Remote support.
Support for software volume control, from -99dB to 0dB
Display incoming signal sampling rate and type, determined by “reading” the relevant I/O pins of the USB to I2S board.
Display and control of the AK4490’s digital filter.
Selection of the proper MCLK frequency according to incoming SR and type and programming of the Si570 accordingly.
Control of “DSD Direct” function of the AK4490s.
Control of the DSD Filter’s Frequency (50KHz or 150KHz).
Control of the Sound Mode of the AK4490.
Choice of either inverted or normal analog output for the AK4490s.
Choice of two sets of MCLK frequencies, either 22/24MHz or 45/49MHz.
Remote power on/off functionality (or always on – configurable in the code).
IR remote library, compatible with the SMT32 (modified version of this library, sourced from here, also included in the ZIP)
In the download I am including the modified versions of the libraries (as mentioned above) as well as the necessary font files. Be sure to extract the contents of “Libraries (place in Libraries folder)” to your Arduino IDE’s “libraries” folder.
Added support of rotary encoder and IR remote control.
3.5″ TFT support.
Added support of rotary encoder for volume control.
Bugfixes related to DSD.
Added support for dual mono mode.
Added very basic TFT support.
Code cleanup for first public release.
Added full control of sound parameters through serial port.
First functional version.
Automatic switching between PCM and DSD by monitoring DSDPIN.
After I was certain that everything related to the software was working the way it should, I designed a “motherboard” that would take care of the following:
Accept the STM32F106 board.
Accept the 3.5″ TFT.
Accommodate an 24LC256 EEPROM chip, used to store the DAC’s configurable settings.
Accommodate two sets of I2C signal isolators and I/O expanders.
Include headers for the encoder, IR receiver, power relay, non-isolated and isolated I2C communication, unused uC pins, etc.
This is what I ended up with:
Basic Hardware Requirements:
STM32F106 module (a.k.a. “blue pill”, search Ebay for “stm32f106c8t6 board”)
3.5″ TFT with resolution of 320 x 480 (Search Ebay for “3.5 tft uno 320 x 480”)
24LC256 EEPROM chip
I2C Isolator ICs, I/O expanders, passive components, etc (see BoM)
Compatible IR remote control (Apple Remote or other – in any case you must edit the code and input the proper IR codes for your remote, see below)
STM32 TFT Motherboard Bill of Materials
C1, C2, C3, C4, C5, C6, C7, C10, C11
Ferrite bead, low R, 1206/3216
R1, R2, R3, R4, R12, R13
2.2 – 3.3Κ, 0805
.100 (2.54mm) – 2 Pin
.100 (2.54mm) – 2×4 Pin
.100 (2.54mm) – 3 Pin
.100 (2.54mm) – 6 Pin
.100 (2.54mm) – 3 Pin
.100 (2.54mm) – 3 Pin
.100 (2.54mm) female – 6 Pin
.100 (2.54mm) female – 8 Pin
.100 (2.54mm) – 3 Pin
.100 (2.54mm) – 10 Pin
.100 (2.54mm) – 2×6 Pin
.100 (2.54mm) – 2 Pin
.100 (2.54mm) – 4 Pin
STM32F103C uC board
How do I make it work?
You have to supply the board with 5VDC at ~300mA through header DC_5V.
If you will be controlling a power on/off relay, you can use the POWER_RELAY header:
I2C header (non-isolated):
Isolated I2C ports
The board has provisions for two separately isolated I2C ports, complete with I/O expanders on their isolated sides. The idea is to connect the DAC board to one of the isolated ports (I2C_ISOL1 & MCP_ISOL1) and your USB-to-I2S board to the other isolated port (usually MCP_ISOL2)
For the better part of a year I’ve been busy developing what one would call a “respectable” DAC from scratch. It has been a team effort, with different people becoming more involved with specific aspects of the project, but pretty much everyone involved ended up learning a lot about DACs. Right now, the project is for the most part finished, in that it is fully functional with a USB input and a pair of single ended outputs (or balanced, if you use a passive transformer-based stage). It is controlled by an Arduino-compatible micro-controller, running my aKduino v2 code (more on that in a future post).
I will do a series of posts detailing the design and build process, with each post covering a specific PCB. When the series is complete, the contents of all of the posts will be concatenated into a project page.
So, without further ado, this is the schematic of the main DAC board:
(Right click, Save Image As.. to download it in full resolution)
The design goal was to do a dual mono design so as to maximize SNR and channel separation. A 4-layer PCB design was chosen so as to have a very solid, low impedance ground plane as well as proper power and signal planes. The I2S, audio signals and power after the local LDO regulators are routed on the top layer, the 2 middle layers are ground and power planes, and the bottom layer serves to route I2C signals and some power lines.
All of the local power supplies are implemented using the currently top-of-the-line LT3042 LDOs. The VDDL & VDDR (analog power supplies) are set to 7.2V so as to maximize SNR and dynamic range. There is provision for providing separate pre-reg power supplies for the L and R channels (headers AVDDL and AVDDR) but I don’t consider that to be critical to SQ since there are local LDOs and the power draw is very very small. In my implementation I’m using a common pre-reg for both the AVDDL & AVDDR. The AVDDs and DVDDs are also supplied by LT3042 LDOs set to output 3.3V. The Si570 has its own dedicated 3.3V supply implemented with an LT3042 and features extra filtering.
Overall, the power requirements of the board are:
1) AVDDL: 8-10V DC at 40mA max
2) AVDDR: 8-10V DC at 40mA max
3) DVDD: 4-6V DC at 200mA max
Clocking & input signals
It was decided that the MCLK would be provided by a programmable low jitter oscillator, namely the Si570. This way we could select different MCLK frequencies at will, supporting different sampling rate families and different USB to I2S boards.
Speaking of USB to I2S boards, the DAC board has a very specific requirement: The USB board must be able to receive MCLK externally. In other words, the DAC board and USB board must be clocked from the same oscillator. This is due to the AK4490’s design. Unlike the ESS designs which by default run asynchronously, it needs to be on the same clock domain as its I2S sources.
So, the DAC board needs to output MCLK back to the USB board. There exist a number of USB boards that support that. The most popular ones are the Amanero Combo384 and JLsounds’ I2SoverUSB.
Since I had decided to do reclocking using flip-flops as close to the AK4490s as possible, and the flip-flops are clocked by the MCLK, its frequency needs to be sufficiently high in order to reclock signals corresponding to high sample rates. This translates to 49.152 MHz for sampling rates of 384KHz. If you’re content with going only up to 192KHz, you can use a MCLK of “just” 24.576 MHz. Of course you will also need the corresponding MCLKs of 45.1584 and 22.5792 MHz for the 44.1K families of SRs.
If you decide that you don’t want to do reclocking on the DAC board, you can just not solder on the flip-flops and just connect the proper pads together with some wire so as to bypass them. That way you can run with 22.5792 & 24.576 MHz clocks with SRs up to 384KHz (and probably beyond..).
If you decide that you would also prefer to not use the Si570 and just clock the DAC board directly from your USB to I2S board (or other suitable I2S source) you can also do that. You just don’t solder-on the Si570 and make a couple of changes to the Arduino code (to be implemented..).
But I urge you to try the Si570 & reclocking way first..
Start by first soldering on the power supply components (LDOs, resistors, caps, etc.) and testing that everything works the way that it should.
The LT3042s are pretty tiny and its easy to make a mistake while soldering them. I’ve found that the easiest method is by using a hot air rework station. First I use a regular soldering iron to tin the thermal pad and the pads with a small amount of solder. Then I apply a good amount of high quality solder paste, put the LDO on top of the pads and heat the area of the board until the solder melts. I set my hot air station to a relatively low temp of ~280 degrees C and the process takes less than a minute (per LDO). But you could also solder them with a soldering iron. To solder the heatpad, warm up the pad from the underside of the board and add solder.
After the power section, you should solder on the AK4490s and other low profile components. I do that with a low power (18W) soldering iron with a fine tip.
Next up is the Si570 programmable oscillator. I start by putting a little solder on one of the pads:
Then I add soldering paste and I place the Si570 on top of the pads. I use the fine tipped soldering iron to melt the solder on the tinned pad, effectively soldering the Si570 on the board. I then proceed to solder the rest of the pads by applying heat with the soldering iron to the side of the pads of the Si570 and adding solder.
You should finish up by soldering the electrolytic capacitors and other higher profile parts.
Beware that the spacing around the electrolytics is very tight. You should take that into account when selecting parts. The parts in the BoM are sure to fit in the available space.