The Best DAC is no DAC

Another suggestion I have is for the connection from relay to the audio output. Instead of having wires connecting them, why not have them as traces?

Relay Pin 2 & 9 route to JLSound Pin 11 (Analogue Ground)
Relay Pin 3 & 8 route to L and R Output Pads (Output terminals)
Relay Pin 4 & 7 route to the LPF output (ie connection between Rf and Cf)

I'm not sure if that was considered but it'll be less messy to do muting on the PCB. Also, this would be for SE. For FF board, you'll have to figure out what is the ground reference for the differential output.

I'm not sure if the 555 timer is required if using logical AND of Pins 7 and 8. I could try to capture the logic states to see.

I originally (on the earlier boards) had the mute function connected to the outputs. On the latest boards I discontinued that to give more flexibility as some people who use buffers like to mute the output after them. Also, the PCB tracks tended to be rather parallel, for obvious reasons, and the steer from the collective wisdom was to avoid that compromise.

For FlipFlop muting just short the pos and neg lines.

Will be interested to hear the results of your deliberations about triggering muting based on pin 7 & 8.

Ray
 
Member
Joined 2009
Paid Member
You have a powerful PC, mine old Vaio only have speed to resample PCM for DSD64, with Amanero for DSD128 I have to use play RAM driver and have to wait a long time to listen.
I have also a WaveIO but I have problems with the latest drivers so I don't know if is capable of native DSD. I looked the HQPlayer but I guess I will have the same problem with the Vaio, also my OS is 32bits. With Amanero I can't use Kernel real time drivers because is only for 64bits.
 
You have a powerful PC, mine old Vaio only have speed to resample PCM for DSD64, with Amanero for DSD128 I have to use play RAM driver and have to wait a long time to listen.
I have also a WaveIO but I have problems with the latest drivers so I don't know if is capable of native DSD. I looked the HQPlayer but I guess I will have the same problem with the Vaio, also my OS is 32bits. With Amanero I can't use Kernel real time drivers because is only for 64bits.

Time for a new PC I suggest. IMO you need to use at least DSD256 to get the best from the no-dac approach.

I posted you some info about the PC I use a couple of weeks back; I'm using a reconditioned HP Z210 workstation that only cost me about £200, then just added a SSD.

Ray
 
Thank you, I guess correctly so I start step a step and I look for a 2nd hand powerful PC min. i5 for DSD256

Felipe

If it were me I would leave the i5 and look for a used workstation with a Xeon processor; Xeon are more optimised for number crunching processing than the i5/i7 ranges (because they don't have the on-chip video etc.).

I would aim for a processor that benchmarks at around 8000+. Here's a site with processor benchmarks;

https://www.cpubenchmark.net/high_end_cpus.html

I have a Xeon E3 1240 3.3GHz processor which benchmarks at 7974 and comfortably upsamples CD rips to DSD256.

Check out the used commercial kit resellers in Spain

Ray
 
ok, I finally found some time to get the soldering iron out and put together a balanced board and it's been generally smooth sailing putting it together.
but.... I had it built and tested it out just on the JL USB/I2S board and like Fabio I was seeing 0.2mv dc offset on 1 channel and 5v on the other. So I noted that he reckoned a faulty flipflop and set about changing mine for another new one. wow that was fun.... :smash::smash::smash:

But with a new flipflop fitted and everything checked 3 times I still get exactly the same behaviour. I've measured and checked again, but then tried something different and powered on the pc which connects to the JL board, waited for that to fully boot, then add the power to the JL board as I would do normally. As soon as the LED lights up on the DSD board, the DC offset drops to around 2mv on both channels and everything appears to behave itssef nicely. I've reproduced this a couple of times.
This leads me to believe that my original flipflop wasn't faulty and didn't need changing and that I'm a little worried about the potential for squirting DC into my buffer/amps if this isn't booted up in the right sequence :s

As long as the PC has made the USB connection and the JL usb board has activated properly and lit the LED on the DSD board it doesn't seem to matter if you remove the power from the computer. The LED stays lit on the DSD board and the DC offsets appear to stay stable, so that's kind of reassuring.
But this isn't ideal...
I've not heard any music from it yet, but I'm going to put caps inline on the outputs for now until we're sure it's all happy.

hmmmm.....

J
 
Ok, wow. This sounds wonderful! :)
I used the 120r rhopoint GG102D from my dddac for the low pass filter with some 33nf polypropylene caps
491bae1042f2ab3e42a3a24e1a92a60d.jpg


6368feb0f5f952677973189d4dc3cba5.jpg


ed471dcd5cd5f12b04a107da4f220109.jpg


6a9d5cf3bb8e90604a5fe8b8a8072c1f.jpg


I'm just getting set up with my new listening room so I don't have a great deal to compare with at the moment, but wow. The level of detail and clarity is amazing! Properly spooky :)

d27c1eb9164c16544c3a9e974ec0c907.jpg


c5882284909acc9e45607793fd38430d.jpg


Now I just need a nice way of remote controlling hqplayer from a phone or tablet and I'm all set :)

Thanks Ray :)
 
...but then tried something different and powered on the pc which connects to the JL board, waited for that to fully boot, then add the power to the JL board as I would do normally. As soon as the LED lights up on the DSD board, the DC offset drops to around 2mv on both channels and everything appears to behave itssef nicely. I've reproduced this a couple of times.

Useful information James, something for people to look out for.

I've never encountered this situation because I know that if the JLSounds USB board is powered up before the PC (in my case my NAA) it won't be recognised so I always power up as you describe.

Ray
 
Now I just need a nice way of remote controlling hqplayer from a phone or tablet and I'm all set :)

The obvious solution is Roon but at a much lower cost (and less functionally rich) I've found Muso to work well at controlling HQPlayer and you can control Muso via a web session on your tablet type device. The web interface is quite basic but seems to work consistently.

http://klarita.net/muso.html

Ray
 
dwjames: So turning on the JLsounds board after the CPU is on does the trick? What were you doing that gave DC on the outputs? Just the JLsounds board without any CPU connected?
Yes, this is with just the JLsounds board powered up with nothing connected on USB. If the CPU is turned on and active first, then apart from a brief blip of DC when it's first turned on, there is no issue and DC offset is negligable on both channels.
I would guess that the default power-on state of the JLsounds board has one of the outputs with power on. Maybe this behaviour can easily be changed in a firmware update or if it's needed for something else.
I wonder if playing PCM material into the JLsounds board also does the same? I didn't try measuring that.
 
The obvious solution is Roon but at a much lower cost (and less functionally rich) I've found Muso to work well at controlling HQPlayer and you can control Muso via a web session on your tablet type device. The web interface is quite basic but seems to work consistently.

http://klarita.net/muso.html

Ray
Thanks, I had a brief look but hadn't seen that it will produce a web interface
that's viewable on other devices, I'll give that a go :)
 
The thread has been quiet for a couple of days, I guess you're all busy with your soldering irons?

Anyway, I expect the additional boards to arrive fairly soon so if anyone from the initial GB hasn't received their boards if you drop me a PM we can try and sort something out.

After that is sorted I'll make any remaining boards available to latecomers.

Ray