SMPS and grounding

[austinpop]

6 hours ago, JohnSwenson said:

I've been experimenting with some ways to make the SMPS a little easier or applicable to a broader range of equipment. First off is a way to use it with equipment that does not have a 5.5x2.1 barrel jack.

 

I found another I had on hand, in this case the 7.5V Mean Well, this was exactly the voltage my switch took, IT plugs into the adapter, then a 5.5x2.1 to something else is used to plug into the device. In this case it was one of the adapters that came with the iPower. This works great. Using this technique I was able to ground the power to a bunch of different devices that had different plug sizes.

 

Both Amazon and Ebay have kits of these adapters available which include many different plug types.

 

Next up is an interesting kit from iFi called the Groundhog. It has a gounding system and several adapters to ground equipment in various ways. The one most appropriate is this:

 

This kit is unusual, it contains an IEC SOCKET, not a plug, you plug in a standard IEC power cord for your country. Most of you will already have several of these in your cable drawer. It doesn't matter what the country is, the IEC end is the same so they will plug right into the Groundhog. They call the adapter I show plugged in the "DC Spade", it is actually a little clip designed to clip onto the barrel of a 5.5 plug, thus grounding it. Clip it onto the barrel, then just plug it in.

 

This replaces the adapter shown in the first post with an off the shelf system you can buy, you don't have to build anything yourself. It is not dirt cheap, ($49), but for those that don't want to deal with building the adapter it is a nice alternative. It does exactly the same thing - it grounds the negative DC output of a power supply.

 

I hope this makes it a bit easier for people to try this approach.

 

John S.

 

Hi John,

 

Many thanks for for this simple DIY solution.

 

Would you consider this necessary with non-floating LPSUs like the HDPlex? I am powering my upstream switch, router, and modem with one of these, and wonder if I should be using these techniques to ground?

 

Also, would you get similar benefits by using a little IT like the B&K 1604A? Obviously this isn't the ultra low capacitance Topaz, but just wondering what your measurements show?

 

[austinpop]

Question for @Superdad und @JohnSwenson

 

A while back I used this approach to achieve better SQ with the Meanwells powering my 3 LPS-1s:

 

 

The P5 is just AC mains, and the Dectet is just a power strip for the purposes of this discussion. Would shunting the non +ve outputs of the Meanwell to AC ground still help, or does the use of an IT achieve the same effect? Should I be plugging the yellow plug ground receptacle (as shown in the first post of this thread) into the Dectet? Or into the P5 - i.e. bypassing the IT?

 

I know it's easy to try this, but it does involve cutting off the tips of the Meanwell outputs - and we all know how painful that can be - which may impact the resale value at some point.

[austinpop]

Just now, d_elm said:

No need to cut DC wires, use connectors as @JohnSwenson showed.  Connect three ground wires together and attach to a single AC plug ground lead.

 

Ah - gotcha. My brain hadn't kicked in yet. Monday morning. 

[austinpop]

A couple more questions:

 

Anyone have an Amazon link for this sucker?

 

And is that bare wire between the male and female DC terminal blocks?

[austinpop]

Gaah - turns out the screw terminals I ordered from Amazon are crap! About 90% don't want to tighten properly.

 

Avoid this POS, please: https://www.amazon.com/gp/product/B01ER6QWAY

 

I will try another brand...

[austinpop]

1 minute ago, lmitche said:

Please let us know what you choose.

 

These male ones I ordered a while ago to make DIY starquad cables were good: https://www.amazon.com/gp/product/B00PMVLBNW

 

But I cannot find females of the same brand. I ordered these, thinking that uxcell has usually proved reliable so far: https://www.amazon.com/gp/product/B005CMP434

 

Will keep y'all posted.

[austinpop]

3 minutes ago, thyname said:

 

Damn! I have to ground all three of my LPS-1s?

 

I know it's a pain, but he Meanwell.

[austinpop]

Would it be better going forward if we refer to it as “high source impedance” and “low source impedance” noise?

[austinpop]

32 minutes ago, JohnSwenson said:

If you cannot use one of the specified switches as the last point in front of your audio endpoint, then your best bet is to add one, they are not very expensive, just make sure you ground the power supply going into it. This arrangement will make sure that whatever leakage is going on in your network will not make it through into your audio system.

 

I have done exactly this. The FS105NA is now acting as an Ethernet isolator, essentially, immediately upstream of my chain.

 

I know John is saying that with this arrangement, you are essentially immune to noise from whatever is upstream of the FS105 switch. I would love for that to be true, as it would be nice to create a barrier (Build A Wall!) immediately upstream of your audio gear, and not worry about grounding every single network device in your house, or putting them all on LPSes!

 

I just have not tested this hypothesis. Yet. Coming from John's measurements, I have high hopes that it will pan out.

 

Now I'd love to see how his experiments with clocking and signal integrity on Ethernet go!

[austinpop]

52 minutes ago, sandyk said:

 Alex C

 Best to put this member on your Ignore list along with the usual disruptive types who like to crap in Audiophile threads in general, even where the poster is as  highly qualified and experienced as John is !

 

 

Yes, ignorance (of disruptive trolls) is bliss!

[austinpop]

5 hours ago, jamesg11 said:

Just picked up a Netgear GS105v5.

 

On a bridged macmini to ultrarendu setup -

— I should connect the macmini ethernet out to the Netgear switch & then into the rendu, & connect the modem-router into the macmini thunderbolt-ethernet adaptor?

 

ie. switch is best between macmini & rendu, rather than between router & macmini?

 

Correct. 

 

Also, make sure to ground the output (DC -) of the PSU feeding the switch, even if it is an LPS. Normally the ground shunting is used to shunt the "high source impedance" noise generated by SMPSes. 

 

In the specific case of the switch, you're shunting the high impedance noise from all the SMPSes upstream of the switch, that made it onto the ethernet cable into the switch, so even if the PSU feeding the switch is not itself an SMPS, and hence a generator of high source impedance noise, you need the shunt.

 

I'm sure @JohnSwenson or @Superdad will correct me if I mangled that explanation.

[austinpop]

Mine also includes ravioli and cappellini.

 

So there!  

[austinpop]

1 hour ago, vortecjr said:

Something like this could make an appropriate ground for this technique.

 

 

 

BoM and links, please!

[austinpop]

@Superdad and @JohnSwenson

 

All these new grounding findings with SMPSes and switches has probably confused the heck out of your customers. After all, until now, I thought the emphasis with the LPS-1 and ISO-Regen products was ground isolation. Now there are new guidelines where ground shunting is recommended.

 

May I suggest you write up an overarching "Grounding strategy" that encompasses these previous and latest findings? Sort of an "i before an e except after c" cookbook.

 

As I understand it, your past advice was to break leakage loops to ground by isolating AC ground from digital gear using:

1. LPS-1 ultracapacitor supply to isolate DC output from AC mains
2. ISO-Regen to isolate leakage loops to AC ground through USB cables

In fact, I seem to recall you recommending that in an ideal system, exactly one component was connected to AC ground, downstream of component(s) isolated by LPS-1s and ISO-Rs.

 

Now it seems the advice would be all of the above, PLUS:

1. Shunt (ground) the outputs of all SMPSes, including the Meanwells energizing an LPS-1
2. Use a Netgear F(G)S/105/108 switch
3. Shunt (ground) the output of any PSU (LPS or SMPS) supplying  the Netgear.

I'm sure I've already got something wrong above!

 

See what I mean?

[austinpop]

Apropos to switches...

 

There has always been an alternate school of thought - promulgated by the likes of Shunyata and Synergistic Research, to name just a couple - that advocates grounding everything.

 

Without getting into a whole rathole about this, one of the outcomes of the above are the availability of grounding cables that can connect to unused ports of a variety of flavors:

1. spare RCA ports
2. spare XLR ports
3. spare Ethernet ports
4. spare USB ports
5. spare 3.5mm headphone outputs, etc.

I want to focus on #3 - an Ethernet port grounding cable. @JohnSwenson's findings with switches leads to a recommendation to shunt the DC input of the switch, to drain the "high source impedance" noise traveling on ethernet cables from SMPSes connected upstream.

 

My question: instead of the double-barrel grounding gizmo John described in the first post, is it equally effective to shunt using an Ethernet port grounding cable, connected to an unused port on the switch?

 

Or is it not the same thing?

[austinpop]

9 minutes ago, JohnSwenson said:

With the switch (one of those Netgear ones) what makes the "shunt all leakage from other boxes" work is that the ground plane in the PCB inside is grounded. So in reality do whatever grounds the ground plane in the switch. The negative pin on the power input jack is always directly connected to the ground plane, so it is a perfect way to achieve this.

 

Other approaches may or may not work. In many switches the metal shield on the Ethernet jack is not directly connected to the ground plane, thus grounding it does not help. Some switches have a ground screw on the chassis, again this only works if it is connected to the ground plane in the PCB. You can test alternative configurations by using a meter in continuity mode and test between the alternative connection option and the negative of the power jack.  Or just ground the negative of the power jack and be done with it.

 

On the "ground everything" approach it is not a good idea, If really implemented that EVERYTHING is gorunded it shunts high impedance leakage, but does NOT block low impedance leakage.

 

Now that we understand what is actually happening a more targeted approach can be applied:

Get rid of high impedance leakage at the source (ground SMPS) and use the isolation techniques where appropriate to block the low impedance leakage. If done right this gets rid of all leakage from getting into sensitive parts (DACs in particular).

 

But it is more complicated than the "ground everything" approach, but if done right works better.

 

Just to let everybody know I am not stopping here. This is how it stands NOW, I'm continuing investigations into how to simplify this and come up with "hopefully" simple ways to deal with all this. Since this is not done I can't say what this will be or when. You all have a choice, use what is here now, or do nothing and wait for hopefully simpler ways to deal with this. I'm trying to come up with solutions people can implement now which cover significant parts of the total puzzle which don't cost a lot of money. A year from now I might say rip it all out and do "this" simple thing and it takes care of everything, but I don't know what that is yet.

 

John S.

 

Thanks, John! Informative as always.

 

And I, for one, am perfectly happy to implement your ideas, knowing full well they may be superseded with something better.

 

One quick question for you or anyone else. How can I tell if an analog component is already "grounded?" I'm thinking specifically of my DAC and my amp. Should I be looking at continuity between the connected AC cable's ground prong and the outer barrel of an RCA output? Or the ground prong of an XLR output? I'm confused.

[austinpop]

9 minutes ago, JohnSwenson said:

If your DAC or power amp are powered by SMPS, then it is best if that SMPS output is grounded somehow. If they are powered by an LPS, you can probably not worry about it.

 

Most likely the power amp has an internal supply so it is a little hard to tell. If the power cord only has two pins, then it is a pretty sure bet the internal SMPS is not grounded. If the cord is three pin, the SMPS may or may not be grounded. In many cases the shield of the input RCA jack is connected to the negative of the PS. Do a continuity check between the ground pin on the power cord with the shield of the RCA jack. If it is connected you are good to go. If not, you may want to try grounding it. The ifi groundhog has a jack designed to specifically do that (ground an RCA jack).

 

Some DACs have internal PS some do not. If the supply is external and two prong, it's obviously not grounded. If three prong it may or may not be grounded. Apply the same test as above. A problem with a DAC is that many do not have extra RCA jacks if you decide to ground it.

 

Checking is not too hard,  but you have to decide whether you are going to do anything about it if you have an un-grounded SMPS.

 

Purely un-substanciated observations seem to point that leakage going through a DAC is more detrimental to sound than leakage going through a power amp, so if you want to do anything focus more on the DAC than the power amp.

 

John S.

 

Thanks, John!

 

I ordered a Groundhog earlier today precisely for that reason.

 

i seem to be hearing the occasional pop and tick since I implemented your shunts, which makes me think I may be completely floating now!  

[austinpop]

5 hours ago, JohnSwenson said:

If your DAC or power amp are powered by SMPS, then it is best if that SMPS output is grounded somehow. If they are powered by an LPS, you can probably not worry about it.

 

Most likely the power amp has an internal supply so it is a little hard to tell. If the power cord only has two pins, then it is a pretty sure bet the internal SMPS is not grounded. If the cord is three pin, the SMPS may or may not be grounded. In many cases the shield of the input RCA jack is connected to the negative of the PS. Do a continuity check between the ground pin on the power cord with the shield of the RCA jack. If it is connected you are good to go. If not, you may want to try grounding it. The ifi groundhog has a jack designed to specifically do that (ground an RCA jack).

 

Some DACs have internal PS some do not. If the supply is external and two prong, it's obviously not grounded. If three prong it may or may not be grounded. Apply the same test as above. A problem with a DAC is that many do not have extra RCA jacks if you decide to ground it.

 

Checking is not too hard,  but you have to decide whether you are going to do anything about it if you have an un-grounded SMPS.

 

Purely un-substanciated observations seem to point that leakage going through a DAC is more detrimental to sound than leakage going through a power amp, so if you want to do anything focus more on the DAC than the power amp.

 

John S.

 

Just a couple of data points. I did continuity checks between the ground pin of the power cord and the shield of an RCA output jack for my DAC (Ayre Codex) and headphone amp (Cavalli Liquid Gold).

 

In both cases, there was no continuity.

 

I’ll see what grounding does once I get my Groudhog.

[austinpop]

@JohnSwenson and others who happen to have an iFi Groundhog... please see my post here:

 

https://www.computeraudiophile.com/forums/topic/31728-ifi-audio-groundhog-the-official-thread-launched-5th-of-may-2017/?do=findComment&comment=739588

 

My Groundhog cable has a resistance of ~15 ohms. Is this normal or advisable?

 

Could you measure yours?

[austinpop]

3 hours ago, austinpop said:

@JohnSwenson and others who happen to have an iFi Groundhog... please see my post here:

 

https://www.computeraudiophile.com/forums/topic/31728-ifi-audio-groundhog-the-official-thread-launched-5th-of-may-2017/?do=findComment&comment=739588

 

My Groundhog cable has a resistance of ~15 ohms. Is this normal or advisable?

 

Could you measure yours?

 

Well, so far my results with the Groundhog are quite promising. I already have several of the JS grounding hydra gizmos for DC barrels, but I bought the Groundhog to see if the same grounding improvement could be realized on my DAC, amp etc, using a spare RCA jack. So far, I am definitely hearing a small further reduction in digital glare when plugged into both the amp and the DAC. The DAC seems to have the most pronounced improvement, as John predicted.

 

However, that 15 ohm series resistance still bothers me. I'm hoping someone can validate that on their Groundhog, and explain why it's preferred to a straight wire (no resistance) like in John's gizmos.

[austinpop]

2 minutes ago, sandyk said:

 

Perhaps the resistance value is low enough to completely remove the high impedance sourced residual noise as measured from a typical SMPS, but high enough to reduce the possibility of inadvertent earth loops causing very low level hum where you may need to use 2 different mains outlets ?

 

That's certainly plausible, and did occur to me too.

 

But is it preferred?

[austinpop]

Well, we'll find out.

 

I posted over in the iFi thread, and also opened a support ticket on this. We'll see what they come back with.

 

Meanwhile, it would help if anyone else here who has a Groundhog would do a quick resistance measurement between the ground prong of the IEC and the outer shield of the RCA. Thanks!

[austinpop]

Well heck - why do I need a groundhog for grounding RCA plugs. This'll do it: https://www.amazon.com/gp/product/B01HEN66K8/

 

Another hydra coming up!

[austinpop]

8 minutes ago, gstew said:

By my count, I have 8 galvanically-isolated (at lower frequencies) zones....

 

Greg, 

 

Have you tried grounding each of these zones? That's the path I'm on.

[austinpop]

13 hours ago, wakka992 said:

can't you just solder a wire to the ground of a RCA plug? assuming you've plug laying around...

same for 3.5mm jack I assume

 

I would, if I could solder, and had all the tools to do so!

 

2 minutes ago, sandyk said:

 

 I did that for my Teac STB which is connected to my DAC via Coax SPDIF, for a small but worthwhile audible improvement.

Just plug it into an unused RCA socket ./

 

I'm going to do the same, once these puppies come in:

 

 

Reminds me of the flying toasters screensaver from the early 90s!