Do it! When I did, the DAC started to sound like magic! Oh, I miss it now because I burnt up my AK4393 chip and waiting for 4396s to be delivered. The unbalancer means alls these components around the opamp and the sand gone.
When I remember how Three Wishes from Roger Waters sounded, I almost get tears. The last time I modded the unbalancer, it was with K52-2 AgTa russian output capacitors. The effect was amazing - so much increase in space.
I am now finishing my power supply for the DAC - two channel EZ81 with 47uF electrolytic and then LCLCRC filtering with 4uF non electrolytic caps, 5H chokes and 300Ohm resistor. The final version of the DAC will have a tube voltage regulator.
I'll say it again - do the unbalancer, you won't regret it.
Just received one of these in finished, cased-up form. I'm very impressed with the build quality and sound for the price.
Before first power up I changed the opamp to the Brown Dog dual OPA827 and put clip-on heat sinks on the regulators. Pretty simple for this novice DIY'er. Maybe next up will be to replace the feet with some much less odorous rubber.
I like it! After the kids are finished watching TV, I'll try it out as the primary music DAC in the main rig.
Thanks to Dario for the great development of this thread.
Cheers!
Before first power up I changed the opamp to the Brown Dog dual OPA827 and put clip-on heat sinks on the regulators. Pretty simple for this novice DIY'er. Maybe next up will be to replace the feet with some much less odorous rubber.
I like it! After the kids are finished watching TV, I'll try it out as the primary music DAC in the main rig.
Thanks to Dario for the great development of this thread.
Cheers!
I've had a chance to listen to my build which was accomplished using ClaveFreman's BOM. I've been comparing it to my Oppo BDP-95 and at first I thought that the BDP-95 trounced it and then i changed the crummy internal hook-up that i was using with pure silver 26 AWG jacketed in teflon. That made a huge difference and I can say that this board with the OPA827 on each output is far closer to the Oppo than it was before.
Musically, this dac is incredible for the money (just over USD $100) with Clave's BOM. This is the kind of DAC that makes you sit back and enjoy the music and l've listen to track after track with it and it's simply enjoyable. Is it as good as the 95? No, but the 95 uses an I-OUT part that is another generation ahead, it mostly beats the 4396 on noise. It really is impossible to compare parts unless you want to set up identical setups, measure, listen and make your own conclusions. In reality, it's silly because the core of each of these DACs is so different. Plus, the player feeds the DAC with I2s and I feed SPDIF to the AKM board. Again, kind of unfair.
The one thing that I must say is that surface mounts and components don't do bass, treble, midrange, increase or decrease soundstage width or depth. What they do is either a better or worse job than than their peers, but they don't "do music" better than other's, much to the chagrin of many. Having said that, the aggregate of Clave's BOM build is an excellent sounding DAC that competes very well with XYZ over-priced commercial DACs that I have owned or auditioned.
My next project will to build Broskie's Unblancer for this DAC. The Unbalancer takes the full diff output from the chip and outputs SE signal, but the summation occurs in a cathode follower instead of an Opamp.
I can't coment on the boards stock sound because I ordered the kit and never tried the stock components, but the stock LPF values aren't correct, so it's really a worthless comparison anyway!
The BOM was purchased through Mouser, Parts Connexion, and I bought the OPA827s off of an Ebay vendor in Washington state. The dual to single dip header adapter for the Opamps came from Cimmaron Technologies.
What I can say is if you need an affordable DAC that sounds great, build this one
It might take a few weeks yet, but I am preparing a thread on DIY Sulzer/Jung supplies having this DAC as my first application.
I still have to think of a way to lower the output voltage down to 3.3VDC to feed the digital parts directly, bypassing the original SMD regulators.
For the output stage I'm planning two modules versions to start with: one for dual IC and just one +/- regulated supply; another one with two single ICs and two separate +/- supplies.
Sulzer/Jung supplies are considered the most sophisticated feedback-based regulators. At the time these regulators were introduced in '95 in The Audio Amateur, they were applied on a Philips DAC960 with excellent results. A version of these regulators is also sold by Burson and others.
AFAIK they hold the best specs around in noise, line rejection and output impedance, better than any 3-pin regulator available.
There's also a high voltage version that has not been tried yet which I'm willing to do to use on the low current stages of a power amp of mine. But this could also be applied to power a high-voltage valve stage.
"Valve stuff" is not easy to find in Brasil. By "stuff" I mean mostly valves themselves and supply inductors, so such a regulator might be an option for the supply at least.
I still have to think of a way to lower the output voltage down to 3.3VDC to feed the digital parts directly, bypassing the original SMD regulators.
For the output stage I'm planning two modules versions to start with: one for dual IC and just one +/- regulated supply; another one with two single ICs and two separate +/- supplies.
Sulzer/Jung supplies are considered the most sophisticated feedback-based regulators. At the time these regulators were introduced in '95 in The Audio Amateur, they were applied on a Philips DAC960 with excellent results. A version of these regulators is also sold by Burson and others.
AFAIK they hold the best specs around in noise, line rejection and output impedance, better than any 3-pin regulator available.
There's also a high voltage version that has not been tried yet which I'm willing to do to use on the low current stages of a power amp of mine. But this could also be applied to power a high-voltage valve stage.
"Valve stuff" is not easy to find in Brasil. By "stuff" I mean mostly valves themselves and supply inductors, so such a regulator might be an option for the supply at least.
Last edited:
Some of you guys really take the buscuit.
Sorry, but meaning what exactly?
What is it you do not like or approve?
That list is already included in upgrade BOM
I would stick to the fantastic R-Core eBay sellers are offering with the MINI2496.
As others reported the most rewarding swap is for opamp rails and analog 5V (U8, U9 and U10)
But you still need U9 and U6 for the analog 3.3V (CS8416)
I've yet to try shunt regulators but I'm pretty sure they will.
You're welcome
But we must give credit to Erin, the thread opener that discovered this little gem.
Unbalancer
I have built a DAC using an unbalancer.
This is a great circuit as it uses the full voltage swing of the DAC.
The unbalancer is Broskie Cathode follower (BCF) with a gain stage before it.
The only reason for using an unbalancer is for the gain otherwise the BCF will do the same job.
I wanted the gain as the DAC is feeding a valve single ended 6B4G amp with 6SN7 driver.
I experimented with different power supply configurations rangeing from the on board supply, changeing the diode to hexfreds, adding choke smoothing to the hexfred option but finally settled on a Shuguang 274B valve rectifer with a 3.9uF PIO cap (russian) a 20H choke then a 1500uF 360v electolytic bypassed with a 4uF film cap.
This gave the smoothest, most grain free sound.
For a BCF I suspect that the power supply would not be so critical as cathode followers have no gain.
Using a BCF with standard supply should be fine.
Add a simple LPF using a cap and resister between the DAC legs and the circuit. One for each leg is needed.
I hope this is helpfull
Mark
x2, i'm planning to build unbalancer as well
do you already have the board? they're out of stock at this moment in glassware store..
should i just build them point to point.. i already bought some of the parts so if i must go with BCF, these parts will go to waste
I have built a DAC using an unbalancer.
This is a great circuit as it uses the full voltage swing of the DAC.
The unbalancer is Broskie Cathode follower (BCF) with a gain stage before it.
The only reason for using an unbalancer is for the gain otherwise the BCF will do the same job.
I wanted the gain as the DAC is feeding a valve single ended 6B4G amp with 6SN7 driver.
I experimented with different power supply configurations rangeing from the on board supply, changeing the diode to hexfreds, adding choke smoothing to the hexfred option but finally settled on a Shuguang 274B valve rectifer with a 3.9uF PIO cap (russian) a 20H choke then a 1500uF 360v electolytic bypassed with a 4uF film cap.
This gave the smoothest, most grain free sound.
For a BCF I suspect that the power supply would not be so critical as cathode followers have no gain.
Using a BCF with standard supply should be fine.
Add a simple LPF using a cap and resister between the DAC legs and the circuit. One for each leg is needed.
I hope this is helpfull
Mark
Hi Dario,
I need a USB input can I buy this.
Finished CS4398 + CS8416 USB DAC With Transformer+case | eBay
Can I upgrade it with your Bom.
Rey
I need a USB input can I buy this.
Finished CS4398 + CS8416 USB DAC With Transformer+case | eBay
Can I upgrade it with your Bom.
Rey
thanks! I would stay with +-15v & 5v salas regs then.
Regarding r-core, I hve already ordered several 7VA pcb toroids. I think it would be the best option considering price/performance.
Hi Rey,
while very similar that DAC is different so you can inspire from my BOM but it doesn't apply as is.
You're welcome
Hi, Is it better to buy opa827AID or opa827AIDGKT.
cheers
It must be OPA827AID, the GKT version is MSOP (different package format) and is not suitable.