DAC AD1862: Almost THT, I2S input, NOS, R-2R

i2s, yes - oops.

It might be oscillation? I'll try the resistor - is small 1R, 10R, 100R? Or 1K?

Have to think about how to use regulators. The system power is 24V so I used dual-output Traco modules for 5+/- and 12+/-, being spot-on there's no room for the voltage drop of an LM317/337, eh?

How about a Zener?
 
Last edited:
It might be ... try 2R2. Look into datasheet for some specification about capacitors (capacitive load) and it can tell you more about that ... try to give only 100uF.
There is no room for LM317/337 (they also don't like output capacitors, only the minimum specified in datasheet)... zener based regulators are nice and simple. Problem is a voltage drop of a zener diode (depending on the temperature). Active regulators need only small output capacitors.
 
Thanks for the pointer, this is valuable learning for me.

The +/-12V converter has a max capacitive load of 520uF per output, +/-5V converter max 1470uF per output. If I understand the schematic, each 12V output sees 4 x (100uF + 100nF) capacitors and each 5V output sees 2 x (47uF + 100nF), plus whatever capacitance exists in the other components. So 5V is definitely not a problem but guess it's getting close on the 12V side?

If I'm truly off the path let me know and I'll go study.
 
Member
Joined 2014
Paid Member
Hi Paddy,

I attached the schematic of the hybrid linear/DCDC converter PSU I had made up for DAC projects. It uses a linear regulator to feed the Murata converters a steady 12vdc input. The +/-15v converter feeds a parallel TPS7A3901 regulator set for the +/-12vdc Miro's dac board requires. The +/-5vdc converter does not have the additional regulator.
 

Attachments

  • Vpower Schematic ver 1_1 (1).pdf
    65.6 KB · Views: 152
I am not an expert in switching power supplies, it's quite complicated for audio.
Install smaller capacitors for 12V outputs, or give separate controllers for each channel ... or install 2R2 resistor instead of jumpers.
Adding analog regulators would be best. DC-DC for +-15V. From that +-12V analog regulators with +-3V room could be enough (use a low input and output capacity).
Vunce helped best :D
 
I apologize Manniraj :(

I’ve used the V5 and V6 Classics, I preferred the V6 over LM6171, OPA1611 and NE5534 opamps. But, the OPA861 I/V boards (Aya II) I fabbed up sound really good also, that’s what I’ve been listening with most of the time.

Thanks Vunce, I will try with my V6 classic one piece and vivid another piece ;)
I will take a look into your preferred option - OPA861 I/V boards (Aya II). Where did you procure this board? Any pointers.

Thanks again.
 
Member
Joined 2006
Paid Member
Hey All,
I need another project like a hole in the head (or another hole....), but this one looks like fun.
Two questions to start:

1. Does anyone have any spare boards they are willing to part with, or is anyone interested in some boards if I order some?

2. Does anyone have any spare AD1862's ordered from Rochester, and if not, is anyone interested in a group buy to defer the cost of the minimum order?

Thanks
J


I’m thinking of ordering from Rochester Electronics (en-US)
: Part AD1862N
.
The minimum quantity is 11 and the price before shipping is $254.65.
If there is interest, I will create a small GB.

The Oshpark PCB is $52.45 for 3 boards.
 
Last edited:
Disabled Account
Joined 2019
Hello,


You should materialize it as a group buy in the commercial section imho : you are two already.


At JPLCB the boards are cheaper if you choose the long postal way but at least think to take the option without lead, less than 20 usd euros for 5 boards.
 
Disabled Account
Joined 2019
Hello,


There is no particular interrest to use an opa861 as a buffer. I will keep the opa861 for the I/V but will try something else as a buffer in your shoes... The C10 is not wanted as well. Which buffer ? the opa 1656 (double) used single channel makes sense with 1206 smd pads for close decoupling. Imho, YMMV :)
 
Hi Paddy,

I attached the schematic of the hybrid linear/DCDC converter PSU I had made up for DAC projects. It uses a linear regulator to feed the Murata converters a steady 12vdc input. The +/-15v converter feeds a parallel TPS7A3901 regulator set for the +/-12vdc Miro's dac board requires. The +/-5vdc converter does not have the additional regulator.

Thanks, Vunce! After studying, I think that my power stuff looks a lot like yours, but with fewer discrete parts :) Really by luck I selected Traco regulated DC-DC converters, so the outputs are actually already regulated.

I checked the DAC output on the bench with an oscilloscope and can't find any sign of noise in the output, which makes me think the noise I hear is indeed ground-related, especially as I had a similar problem with a commercial DAC which was resolved with better grounding inside my DIY preamp.

Next up - try installing the caps and inductors recommended by Traco for EMI control to see if the weird intermod-like sounds heard in specific passages go away.
 
Disabled Account
Joined 2019
Hi,


An advice please:



For the ones having tried several OPA of the BOM : do you adapt the feedback network nise supression with the values of each datasheet or the 1k4 + 4,7 nF suit most of the cases ?


Also : Has anyone tried w & w/o the 22R anti-bouncing resistors of the I2S inputs ?
 
Disabled Account
Joined 2019
thanks Miro,

I assume if the I2S device has also such resistors at the I2S output, one should keep the 22R of the dac board as well ?

I will try 4" + the board traces or shorter if I can side by side with the dac board...I'mnot fan of stacking even to try to make shorter.