12/25/2025 2 PM - Main components

This is a dac+amp, so it needs chips, this is a T1 project so I can afford to have such expensive chips:
Dac chip: ESS ES9039SPRO

• THIS IS a VERY expensive and VERY nice chip, why did I pick this? Because it is a "32-bit audio DAC with ultra-high dynamic range & ultra-low distortion" what that means is I will never have to buy a dac ever again. okay okay I'll be serious:
• 1 Everything is separate unlike cheaper chips like the es 9069 it's not a one for all chip with usb built in and everything, all it is, is ONLY a dac without anything, increasing my pcb making time thus increasing my enjoyment.

Digital bridge: XU316-1024-QF60A-C32
Like the dac chip this is REALLY high end, it runs at 3200MIPS so 8000Mhz which means better sounding audio and audio fidelity

Yes it took 1 hour to find these components and YES i have to design most of the symbols myself

12/25/2025 5 PM - SO what have I done the past 2HOURS

So with some /s inspiration from higher level desktop dacs I have come to this final component list.
NOTABLE THINGS
korg nutube 6p1 dual triode: I really wanted a vaccum tube in the dac design for that awesome crystal quality audio. Instead of insanely high 240v traditional vaccuum tube, I picked the nutube which can run at voltages as low as 5v which is nice.

Knob and screen, rp 2040
usually dacs are just digital in audio out, but I wanted to add my own couple twists to it, because I can interface with the ES9039PRO via an rp2040 I thought, huh let's add some more stuff. So obviously you can change the volume, but theres also gonna be bitrate etc on the screen.
name
ES9039SPRO
XU316-1024-QF60A-C32
ABM8G-24.000MHZ-4Y-T3
SG-8101CE 100.000000MHz TCHPA
W25Q32JVSSIQ
CX90B-16P
TLV75709PDBVR
TLV75712PDBVR
TLV75718PDBVR
TLV75733PDBVR
OPA1612AIDR
USBLC6-2SC6
DLW21HN900HQ2L
BLM18PG121SN1D
LT3045EDD#PBF
ADUM3165BRSZ
DLW5BTM102TQ2L
BLM31PG601SN1L
B0505S-1WR3
6SVP470M
rp2040
PEC11R-4215F-S0024
Waveshare 1.5inch OLED Module
PCA9306DCUR
korg nutube 6p1 dual triode
LT3467ES6#TRPBF
G6K-2F-Y DC5
2N2222AUB
BUF634AIDR
GT-3321667P-01
SJ-3523-SMT-TR
ABM8G-12.000MHZ-B4Y-T
NCP114ASN120T1G
TS-1088-AR02016

12/27/2025 4:35 PM - I finished the schematic

what??

You see the title, that I finished the schematic, well I did, but over the course of 3 days with HOURS everyday working on this, in the morning, in the afternoon in the middle of the night, deadline is in 4 days so I'm locking in.

format

The following devlogs will have been of the past, I'm sorry I didn't do it as I went because I was too locked in. It won't be in chronological, or logical order but I'm sorry 3:

12/27/2025 4:39 PM - Wired up the USB

So I have wired up the usb, you may be wondering WHY SO MANY COMPONENTS for one usb?
It's all to get rid of noise/protect my expensive dac.
I've got esd protection filters chokes etc to make sure that the audio is CRYSTAL clear for my very nice ears

12/27/2025 4:42 PM - Wired up the voltage regulators

Wired up all the voltage regulators I almost made some mistakes lmao I wired up 5v to the 0.9v regulator, would have made a lot of fire/boom/aaah/pain lmao

12/27/2025 4:47 PM - Decoupling capacitors Excl rp2040

This one is pretty self explanatory without decoupling caps everything willl go boom boom boom and I will have fried many a many components incl the 70 buck dac

12/27/2025 4:51 PM - DAC

This one was HARD well it is quite obvious, so this chip is 70 dollars, which makes it only be on high end production dacs not diy ones like this, what that also means is that there are no similar diy projects so guess what!? I have to spend FOREVER looking at datasheets and figuring stuff out. Also includes crystals cause they take no time.

12/27/2025 4:55 PM - MCU

SO this is the big boy mcu, the XU316-1024-QF60B-C24 present in many big boy desktop production dacs. Like the Dac this also had no diy projects and also took forever of reading through datasheets.

12/27/2025 9 PM - SE9039SPRO Wired up

THis is the Dac chip that costs SO MUCH MONEY, why does it cost so much, easy audio quality the more xepunsive it is the better the quality, and in this case it's the dac chip, you have to wire up 3v3 1v2 and 0.9 vots which is quite confusing

12/27/2025 10:08 PM - wired up AMP Plus Xorg Nutube

So, if you look at the schematic you will see complicated stuff, a lot of capacitors booster and power sources, you may ask yourself WHY, so the components at the top is to convert the 5v ALL the way up to 24v to drive the nutube, as it's a vaccuum tube it has certain characteristics which is why theres so many components, for the amp they need 12v which is why theres a booster, one for l and one for r to minimise noise.TWO HOURS because wiring up the nutube was such a pain I also did audio processing pf

12/27/2025 10:11 PM - Wired the RP2040

SEE here is the rp2040 wired up, I won't go into detail as it's just a normal rp2040 with a knob and a screen plus connected to the dac amp and mute loop

12/27/2025 10:16 PM - Started with routing the pcb

So as you can see I only have the components in a rough place not in the final and I have some stuff wired up WHY? you may ask, the why is because I like to place the decoupling caps AFTER I have wired most things up, it may not be correct but it's what works for me, theres SO MUCH SPACE in between components because I need a MASSIVE grounding layer for the least amount of noise and interference possible

12/31/2025 - Finished routing the pcb

So, what I did was I finished routing as you can see in the title. I did record this on lapse but uhh it stopped working :pf: so no recording, but it was fun and there were some roadblocks and hurdles on the way to completion, I'm going to list the major ones:

• Decoupling capacitors, I routed a lot of the pcb, then I realised OH YEAH DECOUPLING CAPS so I had to place them, but then I realised I did not include enough in the schematic to make sure the pcb didn't go BOOM and I lose bajillions of dollars. So I kept frantically going back and forth and adding some, removing some and realised I had routed some things to the wrong ground plane/power source.
• 4.4mm Balanced ports, as you know a 4.4mm balanced connections requires 4 signals, L+,L-,R+,R- if you scroll down you will see that I just tied L- to ground and R- to ground, technically this could work but actually it may damage some of the headphones you plug into it. So naturally I went an thought, okay lets double the signals, but that entails the following:
  • Another korg nutube vaccuum tube
  • Double the amps As audio gear is very expensive I was reluctant to do this, as I probably would not use the 4.4mm balanced port as I use 3.5mm IEMs. THe outcome of this is that I scrapped the 4.4mm port and saved about 100 dollars on amp equipment (yes it's that expensive).
• Layers. IN the end I went with a 6 layer pcb :sho: you may wonder why so many layers? - Ground, as this is high end audio equipment you try to minimise as much interference and noise as possible, I already did this by isolating the ground planes of the RP2040, dac/amp and the usb in. But this was not enough, which is why I added more layers for more ground capacitance, thus less interference. - Routing, as I have a relatively small desk I need this to FIT on my desk reasonably, and with just F/B copper, can't easily do this, I also contemplated 4 layers, which would have worked, but the more ground the better. ### todo Silkscreen, mounting holes (need a metal case acting as a faraday cage to stop interference from the outside