I am curios if somebody tried to feed DSC or any other DSD only DAC directly from IAN Canada recklocker, using u.fl connectors, without any usb transports etc.
I have ordered few assembled DSC2 PCB’s, but thats 3 weeks wait, and i dont want to dismember my only DSC
Also Pavel has a new revision of this design, with resistors network array board… i wonder what would be the difference.
It depends on how much power shortage you have. The power consumption is not at all steady, it fluctuates a lot during boot. In particular when the USB ports get enabled and connected devices begin drawing power.
Over my past career, I’ve seen cases where device would boot properly with one PSU model and fail on another model even though both having same specs. The other one just didn’t have fast enough stable regulation for the output.
Mean Well psu i’ve got as a bonus from a guy who sold me a stack of Ian Canada boards. It lights like a Christmas tree So psu is out of the question, its been in an serious exploitation.
You could try with the official RPi5 PSU for comparison. I don’t know anything about Ian Canada. But in normal case the PSU is 5V 5A and by default RPi5 requires it to be a USB PD one (so that the PSU can tell the RPi5 which voltages and currents it is capable of providing).
I have not seen anyone posting a feedback about it. If you compare it with 2.5.2, this one has its own ndk clocks and swappable modules with resistors array
There is a discussion starting from 2017… over the years, there were prototypes of DSC with active output. Op amps and tube buffers… transformers performed best…
What transformers had the DAC you have measured ? Lundahl ? If so, these are worst to use.
Right now most DSC2 are packed with Joyin Chinese transformers, these perform much better.
But, there are even better transformers, that have twice less distortion with higher output then any other transformers available on this fair planet.
There is a number of non public revisions … probably with active output. Few times Pavel had a new development, and before the research can hit release, entire branch is being bought out with copyrights etc and goes into market looking like this:
Who knows… but it’s not confirmed. And even if someone is up to make a such DAC it would be very challenging. The power supply for the output would be awful.
You see, the power consumption of the AFIR matrix goes up together with frequency. And you have to compensate somehow the decline of voltage… passive output performed best in all tests. An active output would require current stabilisation etc, it’s not like you just throw in an Sparkos or Burson and your done.
My original design used 4th order transient optimized LPF plus output driver stage with cable capacitance compensation (to avoid overshoot on transients like square waves). Indeed it produces very clean 7 kHz square wave.
On the diy forum, the person who craft them has its own thread, just search for Output transformers for DACs.
They are outstanding and the price is without reason but… i just need best sound in town.
It is really worth to at least look into these toys. You wont find a better relocker, which is also a transport and a masterclock for DSC. Combined with PI it makes the ultimate endpoint.