I think the correct way to pose the question about noise floor modulation audibility is can anyone perceive when it has been reduced - not so much "can anybody hear it"? This seems to be the nature of perceptual significant issues & we are working at low levels of this
I've wondered about coming at this from the perception side & working back i.e. can we create NFM at levels & with characteristics/spectra/pattern that is easily audible & then play around with these characteristics to tease out what aspects are perceptually significant & what aren't. In other words exaggerating the condition to examine its perception.
Any ideas what the best approach for doing this might be with music signals?
Putting aside the regs for a moment. With my 3 yrs of mucking around with ps on my amps & dac, I have discovered that the ps side affects SQ far more then many have thought. It’s not about just using the biggest cap to reduce ripple will do, infact using to big a cap will degrade sound, this proberbly is due to the charging pulse. Now what ever I build I will always focus on tuning/ tweaking the ps side first before I move on to other parts of the circuit. Pls do not assume that local decoupling will do it’s job of cleaning up what’s coming from upstream. You guys can experiment to verify what I say.
Evolution took billions of years of trial and error and chanceYeah, tell that to evolution
Last edited:
Indeed, look what can be achieved by stumbling around in the darkEvolution took billions of years of trial and error
So true.
I have found that it is often certain combinations of certain parts also, but more typically just better/appropriate. No doubt the layout contributes also. Line conditioning is part of the power supply too.
Yes Phase that’s true also it’s a combo of parts that takes whatever we’re building to another level & that’s the reason that I do use 1 brand or type of caps on my projects.
Line conditioners does help too. Me my only experience is with Jon Risch Ac conditioner but it will only work optimumly if you follow Jon’s recommendation of loading the conditioner with a 100 watt light bulb or resistor which I found that many have forgone. Now I only use the first stage of this conditioner on my build, it does help a little in cleaning up AC.
Cheers
Line conditioners does help too. Me my only experience is with Jon Risch Ac conditioner but it will only work optimumly if you follow Jon’s recommendation of loading the conditioner with a 100 watt light bulb or resistor which I found that many have forgone. Now I only use the first stage of this conditioner on my build, it does help a little in cleaning up AC.
Cheers
I'm not sure I'm in agreement with sumotan.
I agree the PS usually has a huge affect on sound, but if you ensure that the signal currents don't enter the PS or that they cancel out in the PS then the audible effect of the PS becomes much less if anything.
This is a big reason why balanced designs sound better.
I agree the PS usually has a huge affect on sound, but if you ensure that the signal currents don't enter the PS or that they cancel out in the PS then the audible effect of the PS becomes much less if anything.
This is a big reason why balanced designs sound better.
How to ensure this Hellokitty ? The only way that I know is proper grounding. I don’t listen to balance system. Im not smart enough to design, I build from kits or boards that are proven designs & start to tune/tweak from there. Along the way I experiment alot & through all this I leanrn alot & know what & what not to use to get good sound.
You can ensure the signals don't reach the power supply by using a current source as a load which will act effectively as an almost infinite impedance. Obviously this method depends on design.
A balanced-bridge amplifier is a good example if you have current sources for loads.
The signal currents will be equal and opposite when they enter the power supply and cancel out. There will always be some small level of mismatch though.
Balanced designs also cancel out the input signal currents if there is any so they don't have to travel back to the source avoiding potential ground issues altogether.
You don't need to design in order to do this.
Just purchase two amplifiers of the same make and connect their outputs across the load and then hook up the + and - phase of the balanced signal to the inputs of the two amps.
Just be careful because this will make the load appear half its actual value to the amp.
However, dac analog stages are internally balanced, which makes me wonder why there would be a huge need for an ultra low impedance supply there...there must be a detail I'm missing.
A balanced-bridge amplifier is a good example if you have current sources for loads.
The signal currents will be equal and opposite when they enter the power supply and cancel out. There will always be some small level of mismatch though.
Balanced designs also cancel out the input signal currents if there is any so they don't have to travel back to the source avoiding potential ground issues altogether.
You don't need to design in order to do this.
Just purchase two amplifiers of the same make and connect their outputs across the load and then hook up the + and - phase of the balanced signal to the inputs of the two amps.
Just be careful because this will make the load appear half its actual value to the amp.
However, dac analog stages are internally balanced, which makes me wonder why there would be a huge need for an ultra low impedance supply there...there must be a detail I'm missing.
Last edited:
Balanced is one way to ensure signal currents are resolved in the active stage and don't influence the PS. There's another though - the SE configuration where the active element (say a source follower) acts as a shunt across the load. This sounds to my ears as good as the balanced solution and uses half the components.
Balanced for interconnects though is definitely the way to go.
@hellokitty - which DAC chips draw signal invariant supply current? Have you found any?
Balanced for interconnects though is definitely the way to go.
@hellokitty - which DAC chips draw signal invariant supply current? Have you found any?
No but since they always have balanced outputs the signal currents will always be 180 degrees out of phase so the signal currents should cancel out at the supply. There must be something else in the analog portion drawing current?
You mean a SE bridge? That works too as long as the source follower on the opposing end is loaded with a current source.
Unless you meant something else.
Last edited:
I use this kind of SE buffer on the output of some of my DAC designs. In this case it works well as a headphone buffer - the schematic : lingDAC - cost effective RBCD multibit DAC design