DAC 2496 (AK4393) DAC KIT With CS8416+AK4393+5532

I have also had several fuses blow and it seems it is related to on/off surge. I'm up to 3 amp slo-blo on the U.S 120 volt format.

How do you like the sound? (assuming it has stayed on long enough)

Thanks, Bob - this is only my second DIY build so I'm pretty cautious with things like fuses; I'll up the rating again.

When it has powered on, it stayed on without problems so I spent half an hour flicking through albums with it this morning, and I'm already very pleased with the clarity of what I was hearing. This is with the opamp from the kit too, as I decided to save the OPA827s until I knew everything else was ok.

Bit early to judge though, especially as I was teaming it with a TeraLink X2 and an O2 amp, both of which are new to me (the amp was my first DIY build, 2 weekends ago).

Voltages seem good, and no components seem to be getting hot, so I think I'll throw the 827s in now and pair the dac with one of my regular amps for a better comparison.

Cheers!
 
Thought I'd throw in some numbers here as I'm still working out an overcomplicated overbuilt shunt power supply that will run in the temperature ranges I'm comfortable with.

Keeping in mind that I have split the analog 3.3v and 5v legs from the 12v for the op amp...

Analog leg powering the 3.3v and 5v on-board regulators is pulling 0.57watts
Digital leg powering the 3.3v regulator is pulling 0.22watts
Op Amp is pulling peaks of 0.56watts per side but averages much less at my listening levels.

There are several issues with the power on this DAC, for one you are loading the positive sine of the transformer by 0.97watts more than the negative sine.
That wattage ends up loading the ground leg and I had some noise there without sufficient gauge ground wire.
Also the positive leg to the opamp is lower voltage the way the board is set up with the analog leg coming off the 7812,
where the negative side does not see much load at all on the 7912 (0.4v difference on my board).

IMHO I would recommend cutting the trace between the LM117-5v and C38, then running a jumper to C27 with a ferrite bead in the line and adding a 1uf tantalum capacitor to ground.
That would balance out the load on the 7812 and 7912 and assuming you are feeding the board a separate AC 9-15v would take load off the center tap ground to the transformer...

Or you could go nuts like me and make a shunt supply with 4 independent legs to regulate them all off board and bypass the 7809, 7812, and 7912 all together LoL :)
 
It does seem like madness that all equipment does'nt have a slow start circuit built in.

This was common practice in the time of the old valve radios and for all the slagging off directed at Chinese made equipment both my Bada amps (h/amp/power amp) have this feature - does anyone know if there is small kit available to buy, it would be a simple thing to insert after the IEC plug.

Reading about this problem on this thread it does appear to be related to the American voltage.
 
I get some pretty crazy pops from the shunt regulator if I power up/off with my amps turned on. I also get random no-output issues due to start up time differences between the digital and analog legs that require a power cycle to clear (something with the AKM coming on line before the Cirius, I2S issue?).

So in revision 3 of my shunt I am adding a DC blocking circuit at the input, stepping up to FRED rectifiers, 50ohm resistors after the FRED's to limit in-rush current at power on, 22Kuf primary caps, Polypropylene DC filtering caps, adding Tantilum caps to the LM317T's (which I've already upgraded to LM338's in version2), ferrite beads, and an Atmega 328P to actuate relays for power and audio outputs to overcome pops and the power up I2C issues.

With 3 revisions, 5 shunt circuits per DAC, and 4 DAC's it's fair to say I've got way more into this project than I antisipated when I started LoL!

P.s. and I just realized I'll be buying 4 more SMD only 4396 boards to upgrade the 4393's I mistakenly got originally too :-/
 
Hi, Can somebody show me how to connect the Input Optical TOSLINK S / PDIF Module to DAC 2496. I am a complete electronic NOB. Thank you!!!

Specifically - where do I connect the GND (black) and + V (green). Thanx.
 

Attachments

  • Schowek01.jpg
    Schowek01.jpg
    104.2 KB · Views: 600
I recently completed the modification of my MINI 2496 board with new parts as per the BOM published by clave freeman on this forum. I'm running the DAC balanced outputs through a balanced passive differential filter network then into a 6N6P SRPP buffering stage. Low pass filtering is essential with the AK4393 chip due to SPDIF type signals present on the audio outputs. After filtering this noise is no longer seen on the scope. The filter used is described in Thorsten Loesch's PDF availble on the web entitled "thermionic valve analogue stages for digital audio"

I was already amazed how good this little DAC board was before i modified it. Following the BOM upgrade the sound stage has opened right up. There is suddenly a great depth to recordings and i'm hearing subtle parts of the music that ive not heard before. Trebles are much cleaner, less grainy, bass notes extend right down to the floor as they should. I wasnt expecting such a radical improvement and i'm not sure what bit of the upgrade made the most audible difference considering it doesn't have any audio stages on board. My guess is that the better smoothing and filtering around the DAC and receiver chips has resulted in improved jitter.

My next upgrade will be valve shunt regulator (tubecad Janus based) for the 6N6P buffer and i might finally get round to fitting it all into a chassis!
 
Ive built two of these mini 2496 dac's, one with a Broskie unbalancer and one following Clave Freemans BOM from a bare board and I think they both sound great but I prefer the valve output although the op-amp version has a lm4562 and I couldn't get the evox rifa smr caps so used x7 ceramic's instead I'm not sure how much difference this would make.
 
As someone who reported some blown mains fuses, I have to report the last episode wasn't the fuse at all and I suspect some of the others weren't either. I checked the suspect fuse and it showed good continuity. What had happened was those tiny power screw terminals were loose and the joshing required to change the fuse was enough to reconnect the power lead. Can't say if this may have been a problem with other builders, but I plan to pull the PCB from the chassis, check the traces and try to replace the terminals with something more stable.