Perhaps, but I'm sceptical that I'd be able to tell the difference if I upgraded it.
If the Mac Mini is used with an optical connection to the dac, noise from its psu makes little difference.
Johan-Kr
Johan-Kr
You just trade noise for jitter. With a proper built USB interface, noise isn't as much of an issue. Core audio becomes the issue at this point.
Take the same Hypex modules, send 2 signals into a real speaker load and show us the graphs please.
Try 1.1khz and 1khz
Try 30 Hz and 15Khz
Try 2khz and 2200khz
at 4 watts just to see
Try 1.1khz and 1khz
Try 30 Hz and 15Khz
Try 2khz and 2200khz
at 4 watts just to see
I'm using a USB connection to the DAC's internal hiFace TWO interface. I can't hear any difference in the noise floor with the Mac Mini powered on or off. It would be interesting to measure if any noise from the Mac Mini ever arrives at the preamp or speaker terminals.
I's using Audirvana to bypass Core Audio.
It doesn't create noise that you can hear out of the analog outs. But it gets into the USB interface and causes jitter in the clocks, and interference in the processors and FPGA's in the USB interface. It's very apparent when you switch to a premium streamer powered by an ultra low noise supply. I have a Mac mini as well and I can't listen to it any more since listening to my streamer running Linux. But the better the USB interface is the less it's impacted.
The best interfaces have galvanic isolation, as well as synchronous reclocking after the asynchronous reclocking. A great example of a DAC that does this is the Phison PD2.
Wow, that must be a pretty badly designed DAC! Do you have any jitter measurements to share?
I would be really happy to spend a couple of hundred euro and convert the Mac Mini to DC power if I thought it had a good chance of improving the sound going to the DAC over asynchronous USB3.
But looking at the available solutions, I'm not very impressed. None of the vendors even go so far as showing a plot of the noise in the standard SMPS.
I don't have an electrical or audio engineering background, but I do have computer science background. When websites that sell products claim that that turning off unnecessary language packs is going to improve sound quality on a modern computer, I know I'm reading BS.
I'm also pretty confident in the ability of the Mac Mini to transmit a bit perfect data stream over USB (it manages to send far more information bit perfect to a USB hard drive).
So the question is, if I can't hear any analogue effects from the Mac Mini SMPS, and have no reason to believe the data stream itself is being disrupted, how can I know if my DAC is being internally disrupted by power supply noise travelling down the USB?
But looking at the available solutions, I'm not very impressed. None of the vendors even go so far as showing a plot of the noise in the standard SMPS.
I don't have an electrical or audio engineering background, but I do have computer science background. When websites that sell products claim that that turning off unnecessary language packs is going to improve sound quality on a modern computer, I know I'm reading BS.
I'm also pretty confident in the ability of the Mac Mini to transmit a bit perfect data stream over USB (it manages to send far more information bit perfect to a USB hard drive).
So the question is, if I can't hear any analogue effects from the Mac Mini SMPS, and have no reason to believe the data stream itself is being disrupted, how can I know if my DAC is being internally disrupted by power supply noise travelling down the USB?
It should show up as additional jitter and noise. One would think that it would be relatively easy for the vendors of these products to show the benefits (if they actually worked), but for some strange reason they seem to want to avoid doing that...
I thought you understood your own tests. For starters, look at the scale and weighting.
Jajelos, why not use dBV or dBu for your scale? No weighting. Also, present results notching out the fundamental (ti33er, the 1kHz is the stimulus signal aka the fundamental) plus a wide band measurement. State all parameters, scale and stimulus.
??? How are these two statements consistent with each other?
If only Scott Wurcer was watching this thread
Fig 15-23 of the LM4562 data sheet might provide insight (at least with respect to that device)
Please post the full circuit so that we know what the results relate to.
You can have some fun deploying Douglas Self's techniques of buffering the input from the gain stage to allow much lower feedback resistor values and hence much lower noise.
PS the other plot I would be interested in seeing is the noise spectral density, unweighted, input shorted.
Last edited:
dB scales are relative to 0dB i.e. unity. You need to define 0dB. dBu, dBV, dBm or whatever.
No most DAC's have poorly design USB interfaces. If they were designed properly, the transport quality wouldn't matter so much. USB audio even done as good as it can get is still mediocre compared to a good Ethernet interface.
I'll be taking measurements with the new aPX-555 DSIO module. I have played around with about a dozen USB interfaces so far. The worst is the M2Tech OEM board. The best is the interface in the Phison PD2.
Here's 3 boards I'll be testing along with a JLsounds board with reclocker.
Best bet it to ditch the Mac mini altogether. Both the OS and hardware wasn't purpose designed for audio. In 2 weeks I'll have a very good solution on my website. I've been building servers for audio for 18 years now.