wall wart + LT3045?

[the_doc735]

Can the SMPS ripple noise be reduced by inserting a LT3045 Ultralow-noise (0.8µVrms) LDO linear regulator in the cable between the device and the barrel plug?

 

cheers!

[Panelhead]

  Yes, but the output needs to be several volts lower than the wall watt output. I like 3 volts difference. 

[mourip]

Although the DC from the regulator output might be clean there is still the issue of the SMPS injecting noise back into the AC for other devices to pick up.

 

It is probably still an improvement over a straight SMPS supply...

[the_doc735]

1 hour ago, Panelhead said:

  Yes, but the output needs to be several volts lower than the wall watt output. I like 3 volts difference. 

the  smps is 9V 0.6A, & the LT3045 I looked at was 9V 1A.

[the_doc735]

1 hour ago, mourip said:

Although the DC from the regulator output might be clean there is still the issue of the SMPS injecting noise back into the AC for other devices to pick up.

 

It is probably still an improvement over a straight SMPS supply...

"SMPS injecting noise back into the AC" - yes very true! Need a cheap solution to isolate "them" (more than one!)? LOL!

[Superdad]

7 hours ago, mourip said:

Although the DC from the regulator output might be clean there is still the issue of the SMPS injecting noise back into the AC for other devices to pick up.

 

Even worse is the massive amount of AC leakage current that the SMPS will spew into the DC lines.  That will NOT be blocked by the LT3045 reg boards.

Life would be simple If all it took to make a great PS was an SMPS followed by a linear regulator.  But physics, like life, is not simple. 

[the_doc735]

2 hours ago, Superdad said:

 

Even worse is the massive amount of AC leakage current that the SMPS will spew into the DC lines.  That will NOT be blocked by the LT3045 reg boards.

Life would be simple If all it took to make a great PS was an SMPS followed by a linear regulator.  But physics, like life, is not simple. 

So, this idea is useless? i.e. no improvement at all? waste of time and money?

[sandyk]

 

22 minutes ago, the_doc735 said:

So, this idea is useless? i.e. no improvement at all? waste of time and money?

 

It will be an improvement, but if you want +9V OUT when using the LT3045, you will need an SMPS with preferably from about +9.5V to 10V output.

 The SMPS will still send crap back into the AC mains to affect other components.

If you EARTH the -VE side of the DC output of the SMPS this will improve the DC output.

 (this may be easier said than done !)

 However my experience with cheap SMPS plugpacks is that the output voltage is often poorly regulated and may be quite a bit higher than the label says.

[Superdad]

2 hours ago, sandyk said:

If you EARTH the -VE side of the DC output of the SMPS this will improve the DC output.

 

But that won’t do anything to get rid of the normal large AC leakage that SMPS units transfer into the audio system via DC connections.  I’m not at my desk, but can post all sorts of graphs (at various bandwidths from various SMPS units) to show the large AC leakage currents.  

The shunting of -ve to AC mains ground only takes care of the high source impedance  leakage.  The majority of the bad stuff will remain.  It is there because the makers of SMPS units are required to use ‘Y’ capacitors from AC side to DC side.

This is a large part of why a conventional transformer>diode>caps>regulator supply will most easily be better than an SMPS. [Our isolated UltraCap DC-DC units—fed by an AC>DC SMPS are the exception to the rule.]

 

@the_doc735: I am afraid that your free lunch will not be served today!  

[sandyk]

54 minutes ago, Superdad said:

But that won’t do anything to get rid of the normal large AC leakage that SMPS units transfer into the audio system via DC connections.

 Alex C

In this case, will this also depend to some extent on whether the devices  to be powered are also earthed by design ?

When the SMPS PSU of my STB was earthed, the SQ from Coax SPDIF into my system where the power amplifier is earthed, back through to the Preamp and DAC ,was quite noticeably improved.

 

Alex  K .

 

P.S.
You will be aware that I know about the capacitor(s) that you mentioned.

 Although Silicon Chip magazine measured the leakage between  0 volts /case to earth across a 100 ohm resistor, their solution was to use a 22nF capacitor to earth instead where the P.A.  was earthed, as the lesser of the evils, instead of a direct earth connection to reduce the possibility of earth loops.

[4est]

8 hours ago, Superdad said:

@the_doc735: I am afraid that your free lunch will not be served today!  

From what I have seen, that'll not stop them from trying. It is getting old...

[Foggie]

Will this sbooster in conjunction with SMPS be a better overall solution?  These can only be used with certain SMPS specs though. 

 

Its a bit unclear if these work with standard SMPS as the barrel connector pics look to work with sbooster products.  Although there is a specific model to work with meanwell/regen combo

 

https://www.sbooster.com/sbooster-ultra-mkii/

[jabbr]

15 hours ago, Superdad said:

 

Even worse is the massive amount of AC leakage current that the SMPS will spew into the DC lines.  That will NOT be blocked by the LT3045 reg boards.

 

Let’s not talk about leakage currents as if they possess some sort of voodoo powers that defy the laws of physics.

 

For example an SMPS with 100 microAmps of leakage current: is that massive?

 

The beloved LT3045 does provide AC impedance, well either the AC ripple is shunted to ground or blocked via impedance (depending on the exact circuit).

So of course the LT3045 will provide additional impedance to AC currents ie PSRR.

 

Right?

15 hours ago, Superdad said:

Life would be simple If all it took to make a great PS was an SMPS followed by a linear regulator.  But physics, like life, is not simple. 

 

That meme is advocated by Analog Device (née LT) and they should know  

Not that it’s the only way to do things. Like everything that’s engineered, final measurements tell the truth about how much leakage current remains. Physics indeed!

[mansr]

15 minutes ago, jabbr said:

The beloved LT3045 does provide AC impedance, well either the AC ripple is shunted to ground or blocked via impedance (depending on the exact circuit).

So of course the LT3045 will provide additional impedance to AC currents ie PSRR.

From the data sheet:

[jabbr]

The concept: 

https://linearaudio.nl/silentswitcher

 

I do remain concerned about cheap wall warts spewing noise backwards into the mains AC and consequently do not use them after one of my “ultra” iso transformers.

[Superdad]

@mansr and @jabbr:

I'm really surprised. It seems like you are really missing the issue of AC leakage, which is common-mode and does indeed sail right through the regulators.  PSRR of regulators is for differential noise.  I know you guys know this stuff. 

 

John has a cold this week (courtesy of all the people on the ocean cruise he just got back from), but he said he was thinking about drawing a whole schematic model of how and why all this AC leakage is generated and transferred into a system.  He has written about it before, but somehow it just is not sticking for many.

 

[jabbr]

1 hour ago, Superdad said:

@mansr and @jabbr:

I'm really surprised. It seems like you are really missing the issue of AC leakage, which is common-mode and does indeed sail right through the regulators.  PSRR of regulators is for differential noise.  I know you guys know this stuff. 

 

Sigh ... its really difficult to have a technical discussion when we use the terms "massive leakage" and "sail right through" without talking about a specific circuit. It all depends on the grounding and specific circuit. Indeed the PSRR can be correlated to impedance and in some but not all cases can affect leakage currents. Specifically when the circuit is floating (i.e. does not share common ground with AC). This situation is often the case when we are supplying low voltage DC to the second stage LDO.

 

Quote

 

John has a cold this week (courtesy of all the people on the ocean cruise he just got back from), but he said he was thinking about drawing a whole schematic model of how and why all this AC leakage is generated and transferred into a system.  He has written about it before, but somehow it just is not sticking for many.

 

 

Yeah I've been asking for circuits for years now because that's the key, really. Here is one which nicely demonstrates the point: http://micro.rohm.com/en/techweb/knowledge/emc/s-emc/01-s-emc/6899. Pretty basic that avoiding shared ground, and employing differential signalling between boxes is a good way to avoid ground loops -- which is what leakage currents are, and if you look at the common mode diagram in the above link, it is demonstrated that the parasitic capacitance is what creates the "leakage" circuit. i.e. without shared ground you are exactly in differential mode.

 

I never suggested that leakage currents are never an issue, nor that SMPS are always good -- ridiculous -- nor did I make an unqualified statement.

 

Here I'll make an unqualified statement: the schematic is always important.

[Superdad]

30 minutes ago, jabbr said:

Yeah I've been asking for circuits for years now because that's the key, really. Here is one which nicely demonstrates the point: http://micro.rohm.com/en/techweb/knowledge/emc/s-emc/01-s-emc/6899.

 

Good link.  And indeed it makes the point.

 

30 minutes ago, jabbr said:

Pretty basic that avoiding shared ground, and employing differential signalling between boxes is a good way to avoid ground loops -- which is what leakage currents are, and if you look at the common mode diagram in the above link, it is demonstrated that the parasitic capacitance is what creates the "leakage" circuit. i.e. without shared ground you are exactly in differential mode.

 

Sure, but that's not generally what folks get when they use an SMPS (and certainly not an ungrounded one).  And my original point--to the OP who asked if a wall wart plus a linear reg would take care of everything--still stands.

 

30 minutes ago, jabbr said:

I never suggested that leakage currents are never an issue, nor that SMPS are always good -- ridiculous -- nor did I make an unqualified statement.

 

I sure did not mean to suggest you did.  My post this afternoon was more a reaction to mansr's suggestion--via his posting of the LT3045's PSRR graph--that a DC regulator would block common-mode AC currents. 

[the_doc735]

2 minutes ago, Superdad said:

my original point--to the OP who asked if a wall wart plus a linear reg would take care of everything--still stands.

 

I am under no illusions, I don't expect miracles! I know the linear reg. would NOT take care of EVERYTHING (especially backwards pollution), but just some sort of general improvement overall?

...take care!

[Superdad]

2 hours ago, the_doc735 said:

I am under no illusions, I don't expect miracles! I know the linear reg. would NOT take care of EVERYTHING (especially backwards pollution)...

 

No worries.  Just wanted to be clear that the real issue with SMPS units is not about what they put back into the wall.  They are required to meet strict standards in that regard so their switching noise backwards into the mains is extremely high frequency, spread-spectrum and low level; your wall AC is already much more polluted with noise  and harmonics from other house usage.  (In fact, here is a video showing that some SMPS units will actually quiet the AC line a bit: https://youtu.be/6S73A_fRtqE)

 

My whole point was that the common-mode AC leakage from an SMPS (caused largely by their required use of caps right across their transformers) is something pernicious that gets into your audio system and is not blocked simply by the use of linear regulators.

 

Our other friends here may wish to argue the finer points, technicalities, more idealized circuits, and other aspects—but don’t get confused and miss the basic and true point about SMPS and leakage. 

[mansr]

9 hours ago, Superdad said:

My post this afternoon was more a reaction to mansr's suggestion--via his posting of the LT3045's PSRR graph--that a DC regulator would block common-mode AC currents. 

Please stop misconstruing my posts to suit your agenda. There was talk about voltage ripple, and that graph seemed relevant.

[the_doc735]

23 hours ago, jabbr said:

The concept: 

https://linearaudio.nl/silentswitcher

 

I do remain concerned about cheap wall warts spewing noise backwards into the mains AC and consequently do not use them after one of my “ultra” iso transformers.

@Superdad

"your wall AC is already much more polluted with noise from other house usage.  (In fact, here is a video showing that some SMPS units will actually quiet the AC line a bit: https://youtu.be/6S73A_fRtqE)"

 

"the common-mode AC leakage from an SMPS is something pernicious that gets into your audio system and is not blocked simply by the use of linear regulators."

 

.....surely though, even if you install a 'mains' seperate power line purely for audio, from the main consumer fuse unit, it will still get contaminated by other connections in the unit (the point where they all converge to the main electric cable into the property)? OR, can it be isolated?

[the_doc735]

doesn't the Innuos ZENith SE have switching technology though? (on its mainboard):

 

it also has a pico and built in toroidal transformer & optical drive & standard SATA cables! Yet this is supposed to be high end gear? Wonder what else is not so special?

[One and a half]

21 minutes ago, the_doc735 said:

 

surely though, even if you install a 'mains' seperate power line purely for audio, from the main consumer fuse unit, it will still get contaminated by other connections in the unit (the point where they all converge to the main electric cable into the property)? OR, can it be isolated?

The advantage of a separate AC line for audio is distance from other appliances. Small noise has the wire resistance to overcome and quite a lot can be killed off. 

 

Of course , if there’s interference from a faulty switch in a fridge or wall switch, that will always be vocal and you hear it. So do something about that, it’s the first sign of trouble.

[the_doc735]

36 minutes ago, the_doc735 said:

@Superdad

"your wall AC is already much more polluted with noise from other house usage.  (In fact, here is a video showing that some SMPS units will actually quiet the AC line a bit: https://youtu.be/6S73A_fRtqE)"

 

"the common-mode AC leakage from an SMPS is something pernicious that gets into your audio system and is not blocked simply by the use of linear regulators."

 

.....surely though, even if you install a 'mains' seperate power line purely for audio, from the main consumer fuse unit, it will still get contaminated by other connections in the unit (the point where they all converge to the main electric cable into the property)? OR, can it be isolated?

I have highlighted a place of convergence with a red oval outline in the next pic: