DAC power supplies

[iansr]

I am very interested in Andrea Mori’s (ex DIYA) DAC Lite, FIFO buffer and state of the art clocks. Some of the wildly enthusiastic builders of this DAC have used powers supplies from a co called MPAudio who are into using multiple LT3045s. See here:
https://www.mpaudio.net/post/how-to-maximize-psrr

I am curious to know what the panel think of their approach to minimising noise. Are they on to something or is there a better way for the types voltage. / current requirements that we are talking about here?

[Nick]

Well. I sell power supplies that are two regulators in series, so I see nothing wrong with the idea.

But depending on what you are supplying, not sure that the ripple rejection they are talking about in that link is the thing I would worry about. the fact that you are using a low noise power supply, on a board, and then connecting it to the load with wires, would make me worry more about the inductance of the wires haveing more effect on the noise at the load than those figures for the supply. But hey, better numbers must be better . I would also worry about the ceramic output capacitors on the LT output (assuming the spec sheet is followed) introducing more nose due to being microphonic that any other source of noise.

Just initial thoughts.

[iansr]

Thanks Nick, I was hoping you would comment given your interest in PSs.

[Neal]

Well. I sell power supplies that are two regulators in series, so I see nothing wrong with the idea.

But depending on what you are supplying, not sure that the ripple rejection they are talking about in that link is the thing I would worry about. the fact that you are using a low noise power supply, on a board, and then connecting it to the load with wires, would make me worry more about the inductance of the wires haveing more effect on the noise at the load than those figures for the supply. But hey, better numbers must be better . I would also worry about the ceramic output capacitors on the LT output (assuming the spec sheet is followed) introducing more nose due to being microphonic that any other source of noise.

Just initial thoughts.

Where’s the like / upvote / plus one on this forum!

[iansr]

Nick, probably a numpty question but is the inductance issue that you mentioned not then common to all PSs ?

[Nick]

Yep, its common to all power supplies that are sold as a plug and play DIY solution. Its not common if the power supply and regulator is integrated into the the product. In the DAC I made, I used individual LT1761 regulators for every supply pin on every device as close to that devices pin as possible with gnd and vcc supply plains supplying those regulators supplied by board level LM317 regulators.

[Nick]

Further to that, I would also say that good enough is good enough. My point was when you point to a document that quotes "0.000001mV ripple" that reality may get in the way of there being any point in chasing that number, or comparing two things where the numbers are down in that place.

[iansr]

Yep, it’s common to all power supplies that are sold as a plug and play DIY solution. It’s not common if the power supply and regulator is integrated into the the product.

I assume then Nick that is why the aftermarket PSs you make have a second regulation section towards the end if the umbilical to the device that requires powering?

[Nick]

Yep, closer to the actual load.