Should phantom power be supplied by a linear supply?

[user 37518]

Hi ! My experience is that devices with SMPS have a much more unpleasant sound! For example, the jitter of the clock signal deteriorates measurably, interference occurs during the z A-D conversion, etc. I never use an SMPS power supply in studio technology! My instruments: Neutrik A2-D, Tektronix TDS224 oscilloscope, but if I'm looking for noise I use the "good old" Tektronix 465 - five. (analog) By the way, the Neutrik A2-D also has an analog power supply!!! But my DUNE HD media player, which I use for listening to music at home, also sounds better with an analog power supply than with the supplied SMPS! I am an electrical engineer and was a sound engineer when I was younger.

Well, those SSL consoles are full of SMPS and people just worship them.... I am not denying (nor affirming) what you say, but people don't care or don't know...

 

[user 37518]

I certainly don't doubt that different converters may sound different. But all the attempts at making me believe that clock X resulted in better sound that clock Y have failed to convince me, as long as the system was properly locked.

I once reached the conclusion that between good converters you can definitely hear a difference when comparing them A/B, but when used in isolation, it doesn't matter, I know I will get some heat for saying this from people claiming otherwise, among with arguments including a lot of adjectives such as "openness", "forward", "bold", "sterile", "warm" and my personal favorite: "musical".

 

[RogerAF]

this is where I'm at for the +/- 15v supply:

Wow! I think you're going to be blowing that 0.5A fuse every time the rectifier filter caps have to charge, I've seen power amps with less capacitance. I also suggest that C7 and C14 are not necessary and that C6, C16 and C8, C16 (Whoops, look like you have some numbering errors.) Anyway the smaller caps should be right next to the regulators.
Also I agree with Abbey Road re LM3x7 devices.

I have come to favour the LM317/337 over the 78/79xx. They are more robust and less noisy. In addition, they simplify inventory.

Having just looked at a bunch of TI's regulators, there are versions of LM3x7 with noise below 50µV.
If you are going to use 10,000µF caps at the rectifier filter, I suggest you implement some kind of inrush current limiting.
Regarding the SMPS for 48V phantom supply, recommended practice for keeping switching noise to a minimum is to use a linear regulator following the SMPS. So bump the +15V up to 52V with a boost converter, then use something like a LM317HV (60V adj regulator see: https://www.ti.com/lit/gpn/lm317hv). These are noisy, so some kind of circuit like Thor talks about would be good to figure out.

 

[abbey road d enfer]

Having just looked at a bunch of TI's regulators, there are versions of LM3x7 with noise below 50µV.

Couldn't find any... TI's parametric search is a mess.

 

[RogerAF]

Couldn't find any... TI's parametric search is a mess.

https://www.ti.com/product/LM317A/part-details/LM317AT/NOPBhttps://www.ti.com/product/LM337-N/part-details/LM337T/LF01From my shopping cart...
Better hurry on the 337s, only 508 left in inventory...
Note: These are direct from TI. Both DigiKey and Mouser have 0 (at least for 337s).
Cheers

 

[abbey road d enfer]

https://www.ti.com/product/LM317A/part-details/LM317AT/NOPBhttps://www.ti.com/product/LM337-N/part-details/LM337T/LF01

I don't see any difference with the historic 317/337 in terms of noise.
Still rated at 0.003% of output voltage.
IINM, for 48V output, noise is about 1.6mVrms

 

[RogerAF]

I don't see any difference with the historic 317/337 in terms of noise.
Still rated at 0.003% of output voltage.
IINM, for 48V output, noise is about 1.6mVrms

Ahh, but in the Product Description page:
https://www.ti.com/product/LM317A

 

[abbey road d enfer]

Ahh, but in the Product Description page:
https://www.ti.com/product/LM317A

Yes, totally in contradiction with the datasheet. Which one is right?

 

[JohnRoberts]

I have told this story before but back last century when I was using 3 terminal regulators by the truckload, I traced a noise floor issue to noisy 3-terminal regulators. Checking the spec file (data sheet) there was no guaranteed max noise spec. I contacted the rep, it was a major company and their response was that they do not specify output noise. Apparently their way to deliver six-sigma reliability was to open up the definition of what good is.

After they declined to provide a guaranteed noise spec, I declined to purchase any more from them (I black balled them from future purchases).

JR

 

[RogerAF]

Which one is right?

They only have the TO220 types left at TI, I wonder if it might depend on package and layout
But they are cheap

Apparently their way to deliver six-sigma reliability was to open up the definition of what good is.

Well yeah; as compared to what?

 

[ccaudle]

Yes, totally in contradiction with the datasheet.

If you go backwards through the math, they should meet the 38uV feature list number for output noise at the minimum output voltage of 1.2V. Seems like the line in the feature list is incomplete enough to argue it is dishonest. Also a reminder to read all of the datasheet, don't design in devices based on the marketing copy at the top of the web page.

 

[abbey road d enfer]

If you go backwards through the math, they should meet the 38uV feature list number for output noise at the minimum output voltage of 1.2V.

Yes, it is the voltage reference's noise.
As you say, dishonest and misleading.
Actyally, these "new" parts are no better than the old ones, in that respect.

 

[RogerAF]

Before I knew about "noise" I made several linear supplies for one-offs for various studios. I used a LM723 DIP-8 and a series pass as per the data sheet https://www.ti.com/product/LM723JAN-SP one for the + rail and 1 for the - rail. Everybody was happy and I never had a complaint. I will note however, that all of these dealt with line level signals and analog consoles and tape machines, so the noise floor was probably below the tape hiss.
I once had to rebuild 3) SSL 4000 series PSUs. They had a very similar circuit with several series-pass transistors in parallel. Talk about high current capacity!
The 723 is a real old dog, but riddle me this: why is there a radiation-hardened-space version that's currently active?
I think someone is still using these. (Starlink?)
Who knew?
BTW I just looked at the price. Holy S**t! the space grade is almost $200 a pop in lots of 1000. The Mil grade is just under $8. I think I still have some regular plain vanilla versions from the 80s in one of my parts drawers. Hmmm...Ebay?

 

[RogerAF]

Before I knew about "noise" I made several linear supplies for one-offs for various studios. I used a LM723 DIP-8 and a series pass as per the data sheet https://www.ti.com/product/LM723JAN-SP one for the + rail and 1 for the - rail. Everybody was happy and I never had a complaint. I will note however, that all of these dealt with line level signals and analog consoles and tape machines, so the noise floor was probably below the tape hiss.
I once had to rebuild 3) SSL 4000 series PSUs. They had a very similar circuit with several series-pass transistors in parallel. Talk about high current capacity!
The 723 is a real old dog, but riddle me this: why is there a radiation-hardened-space version that's currently active?
I think someone is still using these. (Starlink?)
Who knew?
BTW I just looked at the price. Holy S**t! the space grade is almost $200 a pop in lots of 1000. The Mil grade is just under $8. I think I still have some regular plain vanilla versions from the 80s in one of my parts drawers. Hmmm...Ebay?

Follow up. Jameco has 'em for $1.39. Not sure of manufacturer. Probably not Mil spec

 

[jokeramik]

Best way for clean power is Funk-Tonstudiotechnik : SCMS-Copyprossesoren/Konverter

lowest noise ever.....and precise

Best regards!