Arduino controlled Dual Mono AK4490 DAC (Part 1)

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)

This is the 4-layer PCB:

This is the parts placement diagram:

And this is the BoM (v1.9) in xls format: Dual AK4490 DAC (main board BoM) (190 downloads)

Design considerations

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.

Power

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..

Construction hints

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.

That’s it for Part 1. Stay tuned for Part 2: The Controller.

ES9028Pro power wows

Following-up on my previous post, I soldered a Crystek 100MHz clock and had a new set of Trident replacement boards made for my upgraded Buffalo III.

The new boards are direct replacements for the Tridents, supplying 2 x 3.3V for AVCC_L & AVCC_R, 3.3V for the oscillator, 3.3V for the AVDD and 1.2V for DVDD.

Upon powering up the board with the new clock and power supplies, I noticed a problem. Whenever I tried to play material with sampling rate over 44.1KHz, the sound became garbled.

In my test rig I had also used the same power supplies with no problem, so I guessed that the problem had to be the new oscillator. But the new oscillator appeared to be working just fine.

To further complicate things, when I turned off oversampling and the DPLL filter the sound improved (but still sounded bad because I was doing no oversampling).

After some probing and prodding, I realized that my 1.2V supply was sagging below 0.9V when the sampling rates went above 44.1KHz. That should not be happening, since I had set the current limit on my LT3042 regulators to about 125mA with the data sheet specifying the 9028’s 1.2V supply current consumption to 82mA. But it appeared that my 9028 was pulling in a lot more current.

So I disabled current limiting and tried it again. This time everything was working just fine, with no sagging, but my LT was getting hot. Real hot. Like 80 degrees C hot when playing 352.8KHz material. So I did a couple of things. First I desoldered the 1.2V power supply and inserted my DMM in series in order to measure the actual current draw. I got these results:

44.1KHz -> 140mA
DSD64 -> 147mA
DSD128 -> 157mA
352.8KHz -> 200mA

So up to 200mA! No wonder my LT3042 was having a hard time.. I was working it very close to its thermal limits.

I decided to mount it vertically and to add a heat sink to the back of the PCB.

With this arrangement I was getting about 62 degrees C with 44.1KHz material and up to about 75 degrees with 352.8KHz. This is tolerable for 44.1KHz playback but not cool (pun intended 😛 ) for high sampling rates. Since 200mA is pretty much the LT3042’s upper current limit, I’ll have to design a new power supply, either with two paralleled LTs or some other LDO. The bigger problem is the heat dissipation. I will have to make the PCB as big as possible in order to have a lot of room for heat sinking.

But why does this chip’s actual power consumption differ so much from the published specs? As always, the devil is in the details. The stated 82mA is for a sampling rate of 48KHz and a clock frequency of 40MHz. I’m using a 100MHz oscillator so power consumption ought to go up, but I never expected it to increase so much. I wonder what will happen if I input a 768KHz signal.