I say it has a lot to do with it in terms of understanding what makes you feel compelled to attack in a troll-like manner. Anything anyone says that might go against your own designs (even if no one knows exactly what your designs are), then it is something you can't tolerate to hear, even if the comment you feel was offensive was not directed specifically at you or your work.
Calling your unsubstantiated claim unsubstantiated is not attacking. And I did not find it offensive, just ridiculous.
To make your claim less unsubstantiated you might start with explaining why the change in phase noise would not show with phase noise analyzer but would still be audible (to you). You could also link to studies that show how low phase noise is actually audible.
That's actually a good question, and one I have been trying to understand. I think there is an explanation though. If you have patience to seriously try to understand the explanation before rejecting it outright, then I would be willing to take a try at saying what it is.
It seems that DIY, like everything else, is subject to change and a need to evolve.
True. In the 1920's, some people made their own resistors and capacitors, which no one does nowadays. Maybe I'm old-fashioned, but having boards assembled is too non-DIY for me.
I'll get to it before too long. If I knew you would agree to my terms then I would have done it already.
Well, you could take a mixed approach and have the fab make the 8-Layer boards and populate the critical stuff on their machines, including C0G 0402 format bypass capacitors in big arrays and you populate more what Jerry Seinfeld in the commentary on the "Soup Nazi" Episode called "chunky style".
Why deny yourself the benefits of modern technology?
Thor
Yes, I think so.
First I designed and made a valve DAC. The goal was to see how far I could get with building a DAC that has all its analogue and mixed-signal circuitry made in valve technology, to learn something about valve switching circuits and oscillators, to learn how to use an FPGA and to learn Verilog (I still regret not having chosen VHDL), to finally make a complete high-order sigma-delta modulator by myself instead of just the bits and pieces I deal with at work, to spend some time hyperconcentrating on all this to keep my mind of things that actually matter, and to end up with a usable audio DAC. All these goals have been met.
Then I designed a CMOS logic gate DAC that's configuration-file compatible with the valve DAC, so I could test alternative configuration files on something with a non-lethal supply voltage. That goal has also been met.
And then I designed the DAC of this thread to test whether I could reduce the noise floor by separating the RTZ logic from the actual DAC, as recommended by JohnW. For fair comparison, I wanted to keep the FIRDAC length and board technology the same, although I did change the filter a bit to keep it from dominating the noise. I also wanted to make a Hilbiber-style bandgap and see how low I could get its noise. The DAC indeed has a lower noise floor than my CMOS logic gate DAC and the bandgap is exceptionally low noise, so I think these goals are also met.
Of course I had no specification list with numbers and test conditions. It's a hobby circuit after all.
Attached is a quick try at the general idea I am considering. Haven't proofread the document yet or it would take longer before it could be posted, so please bear with me on that.
I think there is a difference in number of averages between the plots, but that's beside the point.
Spectrum analysers can usually also be used as modulation analysers. If you have access to one that is good enough to measure phase noise plots like this, maybe you can use it as an FM demodulator and measure momentary frequency versus time.
Spectrum analysers can usually also be used as modulation analysers. If you have access to one that is good enough to measure phase noise plots like this, maybe you can use it as an FM demodulator and measure momentary frequency versus time.
What is the connection between the phase noise plots (or their analysis) and your final hypothesis? Also do you have any real evidence that the differences in those phase noise plots result in audible difference at dac output? And since you are showing visible difference in phase noise plots it does not exactly match your original claim ("Ferrite beads destroy ultra-low phase noise whether or not it looks like it on a phase noise analyzer").
If the phase noise plot does not change with ferrite beads, yet the sound does change, then what is changing is akin to the differences in different types of white noise. You can't tell what type of white noise you have, nor tell the type of phase noise you have from standard plots of either one.
White noise would quite likely be apparent both in phase noise plots as well as spectrum noise floor so even that explanation does not match your original claim. There are other more credible reasons for this apparent sound change such as expectation bias as you have made similar claims for quite some time. To make your claim more credible you should first show evidence that what you claim as difference in sound is real. Commonly used technique for this is AB(x) testing.
If you want to make your claim more credible here is my advice:
- This is not about random beads in random spots. Study e.g. Ott to find out more about the recommended usages of ferrite beads and what type of ferrite bead to use. Ott gives plenty of recommendations for ferrite beads.
- Make sure your layout is not causing e.g. issues with EMC or signal integrity. Again Ott gives recommendations. You could also probe the board to see that there are no EMC issues.
- Since your claim is related to using ferrite beads with clock and clock buffer PSUs the only feasible way for the ferrite bead to have an impact on sound is through the clock output (assuming EMC is dealt with). So measuring the clock output e.g. with phase noise analyzer should show something. Ferrite beads are typically used as noise filters so your suggestion that ferrite beads used in PSU cause white noise to appear on clock output sounds a bit far-fetched. And even if that was the case the noise should be measurable.
- In addition you should prove that your claim about audibility stands scrutiny. As said the common technique for this is AB(x) testing.
At least the design it's DIY, can be a mix so smallest SMD parts soldered by machine & not too small soldered by DIYer's.
How about we stop all these attacks on each other right now. Any more will see the offenders given time out from the forum.