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

[Sorenm]

I have put 'the DAC' together but have no sound...

This is a PCM63 board fed by the JLsound USB board, and with Sparkos opamps. When I started trouble shooting, I found 0.8vdc/offset on one output and 1.9vdc on the other, which seems wrong. When I swap the PCM63 chips left-to-right, the DC voltages/offset follow, maybe indicating something wrong with DAC chips. Does anyone have any insights?

Cheers,
Soren

 

[zoom777]

Hi Soren.
If you disconnect the JLSOUNDS board from the DAC inputs, do you have this DC offsets at the outputs?

How have you configured the JLSOUNDS?

 

[Sorenm]

I did try with the JLsound powered off and the DC offset was stil there. The JLsound is configured for PCM63, which is the same as AD1862...now I want to recheck the hookup.

I verified all the different voltages in the sockets before plugging in the PCM63 and opamps, as they are precious. First, I tried swapping the opamps left to right which made no difference, but swapping the PCM63s does flip the offset between the channels.

 

[NIXIE62]

Hard situation. If everything is oriented correctly, then it is almost certain that those PCM chips are faulty. Where did you get them?

 

[zoom777]

They are from me. But I have installed them and played for a few days.... I am wondering why too. What can cause dac chips to have DC output?

 

[miro1360]

What can great help with diagnostic are detailed photos of your build with all connections and configuration visible

 

[Sorenm]

Thanks for the input Guys.

A good nights sleep made me spot my error, and what may have fried the chips. I use the external oscillator board for the JLsounds USB board and in my quest to keep the wiring neat, wired it erroneously from below. In my search for errors I moved to wiring to the top of of the oscillator board, mirroing my working 1862 DAC, and I just realized that in the first iteration DL/DR/LRCK/BCLK (pins 9/11/13/15) were wired to ground on the wrong side of the 20 pin connector

I feel bad having removed two pcm63 chips from this world

Cheers,
Soren

 

[zoom777]

@Sorenm ,
It pains me to learn of this. But are you very certain that the chips are destroyed?

 

[Sorenm]

Me too. I can't know for sure they are dead and will stare some more at the board and triple check everything as I would also want to reassure myself before plugging in replacements, should I find them. The asymmetric DC offset that follows the chips, when I swap them, suggest they produce it ... and that I fried them.

 

[RickRay]

I would fix the wiring and verify that the chips are bad. The wiring could be causing the offset, but I don't have high hope for that.

 

[Sorenm]

There is hope. I got music from one channel. It seems the sockets aren't snug enough around the legs of the chips, but I was able to get one channel to drop the dc offset and make music. Now, does anyone have suggestions for improving contact between chip and socket?

 

[Brijac]

It took a little bit of force to get my pcm63 chips to get in a socket snugly. Just go slow little bit by little bit. Unless your socket is bad, in which case, desolder and replace. Glad you haven't killed these gems

 

[grunf]

Me too. I can't know for sure they are dead and will stare some more at the board and triple check everything as I would also want to reassure myself before plugging in replacements, should I find them. The asymmetric DC offset that follows the chips, when I swap them, suggest they produce it ... and that I fried them.

it is not very easy to fry them via digital inputs if the voltages have not exceeded Vd

 

[grunf]

There is hope. I got music from one channel. It seems the sockets aren't snug enough around the legs of the chips, but I was able to get one channel to drop the dc offset and make music. Now, does anyone have suggestions for improving contact between chip and socket?

that's good news

do you have any picture of that sockets ?

 

[Sorenm]

They come in/out almost too easily. The socket is the following: https://www.mouser.com/ProductDetail/571-2-1571586-9

Photo below.

 

[grunf]

These are good sockets and should have a firm grip. Did the IC go all the way in? Then he shouldn't go out easily anymore.

 

[Sorenm]

They go all the way in and with a little wiggle, I can get the DC offset to disappear, but only one side/chip will stay like that without pressure from a finger. Maybe the legs atrophied with age... Joke aside, any idea for improving the contract?

 

[Brijac]

This is a very very snug fit. If it isn't in as much as this, it probably has more way to go.

If it is, do replace the socket then. Go for mill-max.

 

[Sorenm]

Thanks @Brijac. Do you have a preferred part# from mill-max?

Cheers,
Soren

 

[Brijac]

I used this one, it serves me good (on the picture).
115-47-628-41-003000 manufacturer no
575-1154762841003000 mouser no

I also use dip8 one for opamps, it endured more than 100 swaps