SMPS and grounding

[Cornan]

3 hours ago, lvc10000 said:

Well, my feedback is not as I expected so I've continued testing a bit more before posting. In the last 4 weeks I've tried JSSGing Satas (System SSD & music storage HDD), the Sonore DC cable of my LPS-1 and the Sonore mini USB stub cable going to my MicroRendu. And I tried the PC with both linear & ATX power supplies.

For the 3 types of cables I tried, the findings in my system go in a similar direction. I was listening for a noticeable variation in distortion for example however in my case I keep coming back to a reduction of the top end information. And the more cables added, the results seems to add-up further in the same direction.

It is seductive with one cable in the loop in the sense that it brings out better focus to the medium and voices for example and fleshes out a bit further/more. However as counter example, a cymbal goes/tones down a notch, or for example a piano's wooden resonances information are reduced/hidden. Tough call. 

 

A tip is to try it on network devices ONLY. Routers, network switches (and its GND screw if any). No other devices. That’s where the JSGT magic happened in my perticular setup! ?

[barrows]

3 hours ago, lvc10000 said:

And the more cables added, the results seems to add-up further in the same direction.

It is seductive with one cable in the loop in the sense that it brings out better focus to the medium and voices for example and fleshes out a bit further/more. However as counter example, a cymbal goes/tones down a notch, or for example a piano's wooden resonances information are reduced/hidden. Tough call. 

 

And there's the rub.  Using subjective analysis alone in these kind of circumstances can be difficult.  What is perceived as a reduction in high frequency energy, often is a reduction in distortion, and hence, more accurate.  I have run into this type of "problem" many, many times in my experience doing customer service and trying to help customers/dealers achieve the best sound.  One might change something in their system which we know is an improvement, from a technical perspective (measured lower distortion/noise), but then they find the system perhaps lacking in "air".  The reality is that the "air" they heard before was really an artifact, and not accurate.

If we rely only, and completely on subjective listening tests, without also applying a technical understanding, and measurements where possible, we can often go down "wrong" roads, and get detoured away from accurate sound in playback.

[Cornan]

16 minutes ago, barrows said:

but then they find the system perhaps lacking in "air".  The reality is that the "air" they heard before was really an artifact, and not accurate.

 

I tend to agree with most of your posts, but not at all with this perticular one I’m afraid! 

The sence of air is not an artifact. IME it is actually a confirmation that you are heading in the right direction but need to optimize something in the chain to get closer to the presence, 3D & 4D experience. Likely the choise of cable material, capacitors or voltage regulators IMO.

[barrows]

13 minutes ago, Cornan said:

 

I tend to agree with most of your posts, but not at all with this perticular one I’m afraid! 

The sence of air is not an artifact. IME it is actually a confirmation that you are heading in the right direction but need to optimize something in the chain to get closer to the presence, 3D & 4D experience. Likely the choise of cable material, capacitors or voltage regulators IMO.

Perhaps I did not explain my point adequately.  I did not mean to suggest that the idea of "air" is ALWAYS an artifact.  Only that when distortion is reduced, sometimes, some of the "air" apparent in a system will be reduced, and this could be perceived as a "problem" when it is actually not: if the reduction in distortion results in a loss of desirable "air", the problem is elsewhere in the system, and not in the change which resulted in the loss of some "air".  So, by reducing the distortion, one has discovered an actual (new, because it was previously hidden) problem in the system which needs to be addressed.

 

The point I am trying to make is that often changes which are actually in fact improvements, can be perceived as negative, when really the improvement is just revealing  a new problem to be addressed.

[charlesphoto]

Another thing to keep in mind is that more transparency can also reveal a lot of new sounds and recording flaws, for better or worse. It can be a bit jarring at first. This happened with many on the recent Naim  firmware update of their streamers. Of course it can be smoothed over with upsampling or MQA, but both of those can rob a lot of dynamics, imo, ymmv. A lot can depend on the type of music listened to as well. My system does a lot better with jazz, edm, acoustic etc than it does with rock (ironic considering I’m a well known rock photographer). Keeping an open mind about it all and the changes is the most important part. Keep in mind there’s never any “right” way. One can throw $ at it, or diy it, but it all comes out relative to your wallet, ears, brain , and room. Some of us are of a more restless nature, and have discovered simple ways of getting increased sound quality. But it’s all relative. 

[Cornan]

13 minutes ago, barrows said:

Perhaps I did not explain my point adequately.  I did not mean to suggest that the idea of "air" is ALWAYS an artifact.  Only that when distortion is reduced, sometimes, some of the "air" apparent in a system will be reduced, and this could be perceived as a "problem" when it is actually not: if the reduction in distortion results in a loss of desirable "air", the problem is elsewhere in the system, and not in the change which resulted in the loss of some "air".  So, by reducing the distortion, one has discovered an actual (new, because it was previously hidden) problem in the system which needs to be addressed.

 

The point I am trying to make is that often changes which are actually in fact improvements, can be perceived as negative, when really the improvement is just revealing  a new problem to be addressed.

 

Spot on! ?

It is actually quite common that an improvement revels another issue down the road in this hobby. Something that you’ll

always needs to be aware of. The true challenge is to understand to rule out really bad changes from

plausible great changes. That line is sometimes very narrow! ?

[charlesphoto]

8 minutes ago, barrows said:

Perhaps I did not explain my point adequately.  I did not mean to suggest that the idea of "air" is ALWAYS an artifact.  Only that when distortion is reduced, sometimes, some of the "air" apparent in a system will be reduced, and this could be perceived as a "problem" when it is actually not: if the reduction in distortion results in a loss of desirable "air", the problem is elsewhere in the system, and not in the change which resulted in the loss of some "air".  So, by reducing the distortion, one has discovered an actual (new, because it was previously hidden) problem in the system which needs to be addressed.

 

The point I am trying to make is that often changes which are actually in fact improvements, can be perceived as negative, when really the improvement is just revealing  a new problem to be addressed.

 

I’ve battled chronic myofascial pain my whole life, and it’s very much getting sicker to get better at first as one reveals those built up layers. As far as hifi goes, I know what you mean. As great as my little amp is I know i could do a lot better, as every lowering of the source noise floor allows me to listen louder with less fatigue. Thankfully it’s about as good of a 30w amp one can find and i just loathe to let it go or spend the $ - for possibly another potential problem! 

[Cornan]

37 minutes ago, charlesphoto said:

 

I’ve battled chronic myofascial pain my whole life, and it’s very much getting sicker to get better at first as one reveals those built up layers. As far as hifi goes, I know what you mean. As great as my little amp is I know i could do a lot better, as every lowering of the source noise floor allows me to listen louder with less fatigue. Thankfully it’s about as good of a 30w amp one can find and i just loathe to let it go or spend the $ - for possibly another potential problem! 

 

For sure any DAC or amp is just showing a fraction of its potensial as is. IME you can get just about any device to sound great. Especially a DAC. GI and great voltage regulation will get you far. Amp as well, but not as much. A new fuse will get that amp happy! Amps is best used as close to the source as possible, just as voltage regulators. That’s one of the reasons why I use active speakers with biamped amps. Less transformer artifacts. More current. More music glow. Less noise.

Balanced wiring and safety ground not connected to the transformer to reduce interfearence from AC mains.

Just a few of so many tips! ?

[Lukasluis]

On 4/30/2018 at 1:06 PM, wwc said:

I use a Cisco Catalyst 2960 - C Switch-- un-shunted.   I also have the Netgear FS105, though I haven't compared it yet.

Have you tried the 2960 shunted?

Just wondering if you already compared the Cisco Catalyst switch with the Netgear FS105. Care to share your feedback please? Thanks

[Lukasluis]

On 5/22/2018 at 12:51 AM, charlesphoto said:

Luis V.: 

 

One point: many of us have now discovered a huge bump in sound quality using Supra CAT8 (or other CAT7) as a dc cable. The extra shielding seems to be doing something pretty incredible. I think my LPS-1 was seriously throttled, even with very good Canare starquad, and sounds extraordinary with the Supra CAT8 cables I’ve made. So, yes, dc cables need shielding. Speaker cables no. Switch it all depends on the switch. I use Cisco 2960’s that have an internal SMPS and beat the pants off of any shunted consumer unit I’ve tried.  Stay tuned: streaming audio is a moving target to say the least. 

Charles, 

 

Care to share how did you wire the twisted pairs of the CAT8 cable for DC cable please? I assume all shieldings will be disconnected on both ends and add a loop wire will be connected. How about the B+ and -, which wires will be used? Thanks. 

[asdf1000]

4 hours ago, elan120 said:

Hi Alex,

 

This is the ZyXEL GS-108B V3 switch, and it is being powered by LPS-1.  I have not done other mods yet as I plan on comparing this modded switch with your suggested Cisco WS-C2960-8TC-L first

 

Hi @JohnSwenson

 

Until the Uptone EtherRegen comes of course - does the Cisco WS-C2960-8TC-L block leakage currents too?

 

I know you've tested the Cisco SG 100D-08 (with PSU DC output grounded) but what about the Cisco WS-C2960-8TC-L?

 

Cheers!

[charlesphoto]

On 7/6/2018 at 3:24 AM, Lukasluis said:

Charles, 

 

Care to share how did you wire the twisted pairs of the CAT8 cable for DC cable please? I assume all shieldings will be disconnected on both ends and add a loop wire will be connected. How about the B+ and -, which wires will be used? Thanks. 

Detailed in the iso Regen thread. Just think cool colors pairs to - and warm colors to +. Personally I'm not longer using the CAT8. 

[Lukasluis]

30 minutes ago, charlesphoto said:

Detailed in the iso Regen thread. Just think cool colors pairs to - and warm colors to +. Personally I'm not longer using the CAT8. 

Thanks. I'm just curious, what DC cable are you using now? 

[charlesphoto]

Solid Neotech 12awg with Kemet caps (and an Lt3045). Also pictured and explained in that thread. Best, CP

[charlesphoto]

12awg hook up wire. 

[BigGuy]

4 hours ago, charlesphoto said:

12awg hook up wire. 

Link please...

[Lukasluis]

On 7/8/2018 at 2:32 AM, charlesphoto said:

Solid Neotech 12awg with Kemet caps (and an Lt3045). Also pictured and explained in that thread. Best, CP

Thanks

[asdf1000]

On 9/20/2017 at 12:27 PM, JohnSwenson said:

. The SMPS are already designed to be double insulated so the AC cannot show up on the DC output, grounding it does not pose any electrical threat.

 

Hi John S,

 

I fully appreciate the fact there is no electrical safety threat with grounding the DC output. And appreciate the great improvements in low levels of leakage noise that you've shown in measurements.

 

But does grounding the DC output now defeat the galvanic isolation from mains RF/noise, of SMPS?

 

Does this DC output grounding connection now provide a path for mains RF to downstream components? Since you now have direct connection (even if it's just ground) from mains to whatever the SMPS is powering?

 

Is that one potential downside of this grounding method (although the net effect is an improvement overall)?

 

Or have I got this all upside down (as usual) ?

 

Cheers!

 

[JohnSwenson]

If the grounded SMPS is powering an LPS-1.2 (or -1), it will provide isolation for what is left over on the output of the SMPS.

 

If you are powering network equipment the builtin transformer will also do a good job of blocking what is left.

 

The issue is if you are powering something else (such as a renderer, preamp, poweramp etc) where the ground of the SMPS MAY provide a path for a leakage loop that you don't want. It may not provide such a loop, it is going to be very situation specific.

 

John S.

[asdf1000]

37 minutes ago, JohnSwenson said:

The issue is if you are powering something else (such as a renderer, preamp, poweramp etc) where the ground of the SMPS MAY provide a path for a leakage loop that you don't want. It may not provide such a loop, it is going to be very situation specific.

 

Cheers John. Noted, that with the LPS-1 and ethernet switch, there is leakage blocking/filtering after the SMPS DC output.

 

My question was moreso about other stuff (in case people are using this method for all their SMPS's like DAC, amps etc).

 

Correct me if I'm wrong but SMPS's have galvanic isolation (between AC input and DC output) by design so I wanted to ask, if you now add a direct connection between AC and DC output (even if it's just earth in this case) are you now defeating this galvanic isolation. 

 

I guess from your answer, it's a yes - this isolation is now defeated but as you said, whether it results in a leakage loop is very system dependant so too difficult to answer?

 

While your original grounding trick recommendations were limited to the LPS-1 and recommended switches, you can probably guess many went around their system grounding every SMPS they could hehe.

 

So it's best to just use your grounding trick with LPS-1's and with the recommended ethernet switches, and not blindly grounding everything in the system / house. Unless you have an audible humming 'ground loop' where grounding one of the "other" things may help with the humming I guess?

[JohnSwenson]

7 minutes ago, Em2016 said:

 

Cheers John. Noted, that with the LPS-1 and ethernet switch, there is leakage blocking/filtering after the SMPS DC output.

 

My question was moreso about other stuff (in case people are using this method for all their SMPS's like DAC, amps etc).

 

Correct me if I'm wrong but SMPS's have galvanic isolation (between AC input and DC output) by design so I wanted to ask, if you now add a direct connection between AC and DC output (even if it's just earth in this case) are you now defeating this galvanic isolation. 

 

I guess from your answer, it's a yes - this isolation is now defeated but as you said, whether it results in a leakage loop is very system dependant so too difficult to answer?

 

So it's best to just use your grounding trick with LPS-1's and with the recommended ethernet switches, and not blindly grounding everything in the system / house? Unless you have an audible humming 'ground loop' where grounding one of the "other" things may help with the humming I guess?

 

 

 

Partially. Connecting AC safety ground connects the SMPS negative output will connect the negative output to the AC neutral IF as in the US the neutral is connected to the safety ground.

 

This CAN lead to a traditional ground loop and or a leakage loop.

 

Remeber the two loop types:

 

Traditional ground loop: caused by current flowing through the AC wires magnetically coupling to the safety ground. Longer wires generally have greater potential developed this way. Very small difference if everything plugged into a single power strip. Large difference if say preamp and poweramp plugged into sockets at opposite end of large room.

 

Leakage loop: leakage that goes AC, through power supply, through various wires either to safety ground or another power supply. A strategically placed connection to safety ground can prevent a leakage loop from going through a audio devices power supply. The impedance of the safety ground connection is much lower than going through another power supply.

 

The result is there is NO simple "always do it this way" approach.

 

You need to take into account how the the AC wiring is done and how you want the system setup.

 

A system where EVERYTHING is connected to the saftey ground in a single power strip can possibly be the best system. All leakage currents wind up being shunted through the single safetey ground. But add just one NON-grounded supply (say a computer) and all bets are off. As is common in many systems today where you have some boxes grounded and some not you really have to sweat the details and figure out where leakage currents are going and work to prevent them from flowing through the most sensitive points (DACs, preamps).

 

John S.

[asdf1000]

6 minutes ago, JohnSwenson said:

Partially. Connecting AC safety ground connects the SMPS negative output will connect the negative output to the AC neutral IF as in the US the neutral is connected to the safety ground.

 

This CAN lead to a traditional ground loop and or a leakage loop.

 

Remeber the two loop types:

 

Traditional ground loop: caused by current flowing through the AC wires magnetically coupling to the safety ground. Longer wires generally have greater potential developed this way. Very small difference if everything plugged into a single power strip. Large difference if say preamp and poweramp plugged into sockets at opposite end of large room.

 

Leakage loop: leakage that goes AC, through power supply, through various wires either to safety ground or another power supply. A strategically placed connection to safety ground can prevent a leakage loop from going through a audio devices power supply. The impedance of the safety ground connection is much lower than going through another power supply.

 

The result is there is NO simple "always do it this way" approach.

 

You need to take into account how the the AC wiring is done and how you want the system setup.

 

A system where EVERYTHING is connected to the saftey ground in a single power strip can possibly be the best system. All leakage currents wind up being shunted through the single safetey ground. But add just one NON-grounded supply (say a computer) and all bets are off. As is common in many systems today where you have some boxes grounded and some not you really have to sweat the details and figure out where leakage currents are going and work to prevent them from flowing through the most sensitive points (DACs, preamps).

 

John S.

 

Thanks John!

 

When I mentioned mains borne RF in my first question - is this directly related to leakage currents and leakage loops you talk about?

 

Are they both essentially the same thing / directly related? 

 

Or is mains borne RF a completely different /separate problem

 

 

[JohnSwenson]

Mains born RF noise IS a separate issue. The leakage I'm talking about is primarily line frequency, 50/60Hz and harmonics.

 

How much of an issue it is and what to do about it are very system dependent. Again no simple generalizations possible.

 

One problem is that several common solutions for RF actually increase leakage loops. At least for me and my my systems it is better NOT to use these attempts at RF attenuation.

 

I haven't spent a lot of effort on RF issues, for me they have been much lower importance than leakage.

 

John S.

[Ralf11]

I'll add that mains borne RF is a plausible problem but is rarely (or never) quantified.  You can look at the Mains thread for more - it may be titled Mains Machinations.

 

Isolation by distance works well in this universe

[asdf1000]

5 hours ago, Ralf11 said:

I'll add that mains borne RF is a plausible problem but is rarely (or never) quantified.  You can look at the Mains thread for more - it may be titled Mains Machinations.

 

That thread is a little different.

 

SMPS's employ mains RF filtering (I'm told by DAC designers).

 

My original question was only about mains RF - if you add a wire connection between mains power and the SMPS DC output, do you now defeat a lot of the mains RF filtering built into SMPS's.

 

Probably less of an issue with the LPS-1's and John recommended ethernet switches of course. Possibly an issue for those grounding other SMPS's (DACs, preamps etc). Just to re-iterate for others, John never advised to go grounding those other devices but I still think people may have.

 

So I'm just trying to understand what are potential downsides of that. I think it's clear from John S now - it may affect SQ, it may not. Very system dependant.

 

Better to just stick to grounding the SMPS for LPS-1's and John S' recommended switches and leave other SMPS's alone. Unless you have an audible ground loop hum where the grounding trick may help somewhere.