That filter will get you a cut off of around 1024Hz..
The first dac I made this way sounded nice, but the common mode noise only vanished sufficiently after building it balanced and using a really high speed flip-flop, as well as using a crystek clock and good coax cable. It amazed me how sensitive it was to all that. It began with a nice, airy and fluid yet somewhat fuzzy sound. Now with many tweaks and upsampling to dsd512 (hqplayer) it's really good. For a much lower noise floor filtering should be done better as well as the known serial to parallel conversion. Happily there are still some tricks in the bag, but this got me around -100dB for s/n figures, my goal though was transparency and it really is. Better than many dacs I know of. In short: interesting project YashN!
The first dac I made this way sounded nice, but the common mode noise only vanished sufficiently after building it balanced and using a really high speed flip-flop, as well as using a crystek clock and good coax cable. It amazed me how sensitive it was to all that. It began with a nice, airy and fluid yet somewhat fuzzy sound. Now with many tweaks and upsampling to dsd512 (hqplayer) it's really good. For a much lower noise floor filtering should be done better as well as the known serial to parallel conversion. Happily there are still some tricks in the bag, but this got me around -100dB for s/n figures, my goal though was transparency and it really is. Better than many dacs I know of. In short: interesting project YashN!
The Italian project specified a 1nF cap to ground, so maybe that will be better.
Correct, and that's where the boys in the "The best DAC is No DAC" thread are heading.
Are you using HQ Player in client-server mode, i.e. with an NAA?
Yep, yep. Thanks.
The Italian project specified a 1nF cap to ground, so maybe that will be better.
I'm thinking the same thing;-)
Correct, and that's where the boys in the "The best DAC is No DAC" thread are heading.
I am kind of surprised this seems to be your first shot at "no dac", because I was just over at CA reading the T+A thread. Some nice info there from you as well. Are you by any chance good in programming or anything else as well?
Are you using HQ Player in client-server mode, i.e. with an NAA?
I did, but the last few months I am using a specially made usb-link from pci to amanero, with a linear supply (both). I am busy with some hardware tweaks for motherboards though, which have to be evaluated in the next 6 months or so. I didn't like the Linux NAA's because of Amanero(you know: dsd128 max and all that), then again: I'm not really liking W10 as a headless device either, but if I had to pick: I'm much better at Windows
Do you use an NAA?
Yep, yep. Thanks.
Btw: I am a big fan of "Keep It Simple Stupid" as well. There seem to be way too many variables to control in the first place,so better make it as simple as possible indeed.
Cheers!
Simple is good, and that's why I get better sound with my SET amp than with my SS amp although the latter is 100W (more than 10x supposedly more powerful than my SET Tube amp
).However, the ideal way to use HQ Player is to use it with an NAA connected to the server computer by Ethernet, and that NAA then connects to your DAC with USB, providing extra isolation from the computer noise. Give it a try (something like the Beaglebone Black as NAA is cool).
My bad: please read the previous post again.
I like this solution of a custom implementation of usb way better than the NAA-solution. I've tried raspberries, Udoo's, beaglebone and j1800 Intel. The only thing beating this current implementation is the botic driver from Miero.
mterbekke, it sounds like you've accrued some interesting experiences. A few questions if I may.
Better than what? How?
Are you referring to the use of the flipflop when you mention parallel?
Any clues?
Cheers
Ray
Better than what? How?
Are you referring to the use of the flipflop when you mention parallel?
Any clues?
Cheers
Ray
Raspi has a habit of using a combined Ethernet/USB interface, crippling the bandwidth.
The NAA configuration provides an additional isolation layer that the USB cannot provide.
People who have experimented with all these configurations actually prefer the NAA with Optic Fibre and Linear Power Supplies.
I'm not yet equipped for a small SBC NAA as I only have an old Raspi, but I used my Macbook Pro as NAA as a test, it sounded great.
So for now, I am really looking forward to how this direct DSD can sound with as simple a circuit as possible.
Thanks Ray, sounds very good to me.
I'll also be using your specified values for the rest, 25uF + 47K.
I'm not really convinced this isolation search is the best solution imaginable. From a hardware perspective there are all sorts of problems (soft and hardware related) which partially get solved by isolation. When you look closer at the hardware, it's easy to spot all sorts of solutions. One of them is using a lower power (less emc and garbage on the power supplies, clock lines etc) source, as well as being able to get rid of some noisy switched mode power supply. I've had a few cases of apple notebook's having a way better sound than many of the shuttle/acer/HP solutions. They probably had a way better power supply feeding the usb chip inside. We now know this is because they had a way better power supply feeding the usb chip& it's surroundings. Kind of like a Regen/Wyrd built in. None of it is galvanic isolation related, but has all to do with clean power supplies, good clocks and good connections. Not to say it isn't needed, but it depends if you look at it from a customer, or from an electronics engineer;-)
Yeah, I was looking for colors to highlight the text, but typing on a phone made me look once and think "nah, don't think it works this way". Sorry for that!
It believe it was Gordon Rankin once who mentioned the higher number of BER errors in the USB interfaces on PC compared to Apple.
Agreed, things are quite complex when you want to extract the most of out of the interface.
Just to see how complex they are for some people (even moderators of this forum - ha!):
- people don't know that what goes on a USB cable as signal really are voltage levels with ramp up and ramp down times, that this is really analogue, and hence is subject to signal integrity issues.
- people think that since a file can be transferred without error with USB, then audio streaming through USB also occurs error-free (as if that's the thing that matters)
- some Microsoft guy on What's Best Forum thinking the theory of operation of the Regen and Sonore Microrendu is because "there's a re-send when signal integrity issues occur with async USB" (at least he thinks John Swenson offered this as explanation when they met, which cannot be true!).
If you now start talking about the other issues we know of, these people will go insane
Hey thanks, mterbekke, actually I just used my small Canon digital camera and was experimenting with some of the settings in a very dark room for the darker pics, not sure I did those well, I still need to learn this.
I retouched them on Mac OS X with the default viewer.
I believe it's AC:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.