Digital noise/interference on the left channel

Hi everyone,

I've recently assembled a breadboard DAC prototype based on the AD1862s with an AD811 I/V stage, when I test this board with my headphone amplifier board it seems to work fine.
I've also tried to add preamp and equalizer stages from Yamaha and Behringer in the middle (between the DAC and HP amp board) and everything is also ok, no sound issues, even if I move the sliders or knobs all way up or down.

Now I'm trying to put a preamp breadboard between the DAC prototype and HP amp (to perform some small L/R balance, volume and tone controls). I've noticed that whenever I play some sound or music in the DAC board I get an hiss, white noise like, on the left channel which is really anoying, even when I turn the preamp input pot all the way down, the noise is still there and it is only present when audio is being played, I don't know what is going on, since the right channel seems fine. It gets worse if I add some treble gain. Any ideas of what should I look for?

All the Best,
Danny
 
The noise is there even without the DAC connected, if there is some digital signal on the setup, when the I2S board is connected and some sound is being played there is noise, the DAC board can be completely off.

EDIT:
I haven't tried yet but it should work fine if no digital circuits are present in the setup.
 
Here are the schematics of the preamp and DAC boards.

EDIT:
I have tried to switch the computer on and off and the LED lights also, it makes no difference, only the digital circuit of the setup seems to cause the noise.
When the balance pot. is adjusted to full left or full right, there is a lot of noise on the cutoff channel, if digital circuits are on. The higher the treble pot. setting the higher the noise.
The noise seems to not be present after the first stage (voltage gain).
 

Attachments

  • Tone Control v1.png
    Tone Control v1.png
    28.6 KB · Views: 67
  • Preamp Sim Files (LTSpice).zip
    4.4 KB · Views: 36
  • AD1862 Circuit Diagram - v2 - small.png
    AD1862 Circuit Diagram - v2 - small.png
    98 KB · Views: 70
Last edited:
So you are certain that the DAC causes the noise? Turning it off kills the noise?
No it is not the DAC that causes the noise it's the digital circuits, the optical to I2S board (DIR9001), the inverter and the two shift registers.
So no, turning it off does not kill the noise, maybe it lowers a bit but I'm not 100% sure.
Is the DAC in the same chassis as the tone circuit? How about some photos?
I don't have a chassis yet :ROFLMAO:
This is a breadboard prototype, which uses 3 breadboards and 2 PCBs, I can send you a picture tomorrow, but it looks very messy. :p
It is powered by 4x9V NiMH 320mAh batteries and a 5V Li-Ion 10000 mAh battery.
 
Moderator
Joined 2011
The chassis is just whatever the pcb is mounted on. Prototypes are always messy.

Can you post schematics and photos of the optical to I2S board, the inverter, and the two shift registers?
Are the batteries charged only when not being used? Is the digital ground independent of the analog ground?

If noise is only in one channel, that is a good sign.
 
Last edited:
Member
Joined 2015
Paid Member
In addition :

Are the the supply pins of the two 4562 properly decoupled ? With 100nF to GND ?
4562 is fast and easily enter in oscillation, you could try adding a 100 Ohms in the negative feedback path of the two output buffers : between 4562 output and neg input.
Not difficult to test on a breadboard.
 
Member
Joined 2015
Paid Member
I have no experience with AD1862 but it seems that pin 1 and 2 should not be connected together to negative analog supply.
Negative analog supply should go be connected to pin2 and a dedicated cap should be connected to pin1.
see AD datasheet :
1692608747379.png


see also DIYAudio member miro1360 implementation :
1692608843230.png

( https://www.diyaudio.com/community/threads/dac-ad1862-almost-tht-i2s-input-nos-r-2r.354078/ )

Not sure it has influence on your noise issue, but that should be checked.
 
  • Like
Reactions: 1 user
Thank you very much for helping me AIM65.


Are the the supply pins of the two 4562 properly decoupled ? With 100nF to GND ?
4562 is fast and easily enter in oscillation, you could try adding a 100 Ohms in the negative feedback path of the two output buffers : between 4562 output and neg input.
Not difficult to test on a breadboard.
I have a 100 nF capacitor in parallel with a 10 uF capacitor connected between +Vs and -Vs, but not to GND.
It doesn't seem like oscillation, because if I stop playing audio in my PC the noise stops.
 
Also, these circuits will not work very well on a breadboard. They need a ground plane.
I definitely agree, but I like to test everything on more flexible prototypes before proceeding with a PCB design. I'm also very bad at soldering things, my hands shake a lot, most of the times I end up damaging the circuits and burning myself, I only have a few successful PCBs. Soldering SMDs such as the shift registers in Miro's DAC board is nearly impossible to me. :cry:
 
Moderator
Joined 2011
I would suggest, instead of soldered circuits on a perfboard, that you try a solderless breadboard.
With careful planning, even without the ground plane, you can get compact circuits that will work
pretty well before moving to a pcb with ground plane. Use the pre-stripped and formed wires that
are made for it. I have had very good results using these with thoughtful planning. They are completely reusable.
Small pcb adapters for SMT to TH are also available. There are many kinds of breadboards, an example:

https://digilent.com/shop/solderless-breadboard-kit-large/
https://digilent.com/shop/breadboard-wire-kit/
 
I can't hear the noise, if I reinforce the GND with like 3 aligator clips per breadboard, however, the noise is always there if I move the balance pot to the extreme positions, what can be happening?

Some photos of the terrible mess 😋
 

Attachments

  • DAC Board + Digital Boards.jpg
    DAC Board + Digital Boards.jpg
    779.4 KB · Views: 44
  • Tone Control.jpg
    Tone Control.jpg
    773 KB · Views: 49
  • HP amp + batteries + GND connections.jpg
    HP amp + batteries + GND connections.jpg
    778.1 KB · Views: 46
Moderator
Joined 2011
Then you are using solderless breadboards. An advantage of these is that the layout can be
rather similar to a pcb layout if carefully planned, preferably done with the short, precut solid wires.

Since one channel does work correctly, and troubleshooting this would be problematic, you can
probably, with some justification, assume the noise will be gone when the actual pcb layout is done.
You could certainly lower the balance pot impedance, and just use a single 10k pot.

With the power off, you can check ground continuity from each circuit ground to the power supply ground.
Especially make sure the balance control wiper reads zero ohms to ground at all times.

Some types of solderless breadboards have breaks in the power supply bus lines (at the sides),
that may not be expected.
 
Last edited:
  • Like
Reactions: 1 user
I can't hear the noise, if I reinforce the GND with like 3 aligator clips per breadboard, however, the noise is always there if I move the balance pot to the extreme positions, what can be happening?

Some photos of the terrible mess 😋
It's a good thing that these AD811s work at all in those conditions and they are still without a heatsinks.
Surely AD811 is a source of noise, i.e. self-oscillation.
 
  • Like
Reactions: 1 user