Another DAC + Headphone amp

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi guys.

This is my first audio project, but not my first rodeo. I have taken quite a lot of influence from Gary Marsh (HD-1) and ezkcdude (ezdac). The improvements I aim to make with this project are:

- Improved power supply via low noise LT1963 regulators.
- Volume control is done via balanced programmable gain amplifiers. (Some may debate whether this is better or not.)
- PCB will be 2 layers and therefore cheaper and less of a pain to order in small quantities. (Some may prefer 4 layers for better power and ground layout schemes.)

There are some factors which will probably be considered negative:
- Since I am laying the board out in the free version of Eagle, I only have about 4"x3" and 2 layers to play with, so the transformer will not mount on the PCB... the finished DAC probably won't be quite as svelte as the HD-1.
- You will still have to program the ATmega microcontroller. They are my favorite 8 bitters so I'm sticking with them.

There are a few items that may change or be added:
- IR receiver for remote control. I plan on using this board with a power amp as well, so this would be handy.
- Encoder in place of potentiometer.
- Sub 2 x WM8816 for 2 x PGA2320.

Attached is a preliminary schematic, which I post hoping someone will catch simple mistakes that are surely there. I have not rigorously calculated the proper gains for the various portions of the design, but I will be doing that and finalizing the rest of the schematic over the coming days. Your feedback is welcomed .... :eek:
 

Attachments

  • prelim01.zip
    98.9 KB · Views: 395
ghetto said:
So I would have to buy 47 WM8816? :bigeyes:


You design a good PCB and offer it as a GB and that might not be enough!

I would not condone it as a commercial venture, but a GB is a GREAT way to finance YOUR own personal parts and materials.

Design a product, come up with a parts list and offer them at a reasonable cost. People WILL buy a good kit priced right. Collect 1/2 to 3/4 of the money up front, order them and ship out kits. In this case it would "only" be 45-46 kits...
 
Boards are EASY! Components are much tougher..

I'd suggest boards and chip /controller parts.. Resistors, caps and such anyone can get.. Think about just the "big" stuff, and then it would be very easy.


Maybe someone else would do the GB if you designed the kit. If your talking a GB of 4-5 items x 45 kits it is VERY easy.
 
Hey, ghetto. Schematics are relatively easy. Designing the board - much harder IMO. Anyway, thanks for the shout out. Good luck with your project. I only have one opinion right now. I don't understand why you think the LT1963 is so great. The ripple rejection is lower than LM317, and I don't think it's less noisy either. However, low dropout has some advantages.
 
There were lots of errors with the original schematic. :smash:
- Left and right channels were switched at headphone jack.
- Mute on PGA2310 was not connected to a pin on the micro.
- Chip select not connected to a pin on the micro.
- Others that escape me now.

There have been some other changes:
- Reset supervisor to prevent power-on glitches.
- Relay to suppress TPA6120 power-on thumps.
- Bicolor LED (lock / no lock indicator?) Good for debugging.
- 36Khz IR receiver added.
- Alps encoder with switch added. Switch function not chosen yet.
- Ditched LT1963 for +/-12V (DD5 package = huge), kept for +5V. Probably no audible difference anyway.

I think that covers the changes. The PCB is getting closer. Some packages need tweaking, some traces need rerouting... I bet you will find a few things you don't like.
 

Attachments

  • pcb-sneak.gif
    pcb-sneak.gif
    94.2 KB · Views: 862
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.