USB to PC chassis ground ? (noise) - The biggest change ever !

[PeterSt]

From here :

Anthony, you are funny ...

I just wanted to start a new topic with a suggestion of applying the Silverstone card to the PC without it's bracket touching the PC chassis. This, while I didn't even investigate whether the bracket is connected to the general ground of the card. Btw, it even could be that this is not the case, while the other cards do.

Anyway, now you come up with this riser cable, which exactly does what I wanted to propose (outside of other cutting wire opportunities).
Or remove the bracket ...

I feel this can make a great difference, and I guess you will agree with me right away ...

Regards,
Peter

Yes I agree, but I doubt that the bracket is connected to the ground of the PCIe-USB card, but you never know.

I was thinking about starting a new topic about these upcoming experiments with the USB card to see what (if anything) we need to do to improve noise (in theory - I assume that noise is what makes the difference) from the USB port of the pc.  Maybe you could start the thread off Peter with some reference links to the other threads that have been going in that general direction and then I will chime in with what I am hoping to trial.

Anthony

So, let's try to let this topic be *only* about what happens to USB Yes or No connected to the PC's chassis.

USB travels a very strange path which is related to PE (Protective Earth). Something like : When you use 5V from USB that right away influences PE and a 5V offset is there. It is (way) more complicated, and/but I have always seen this as a major downside. However, I "utilize" this for turning groundloops the way I want with the NOS1. You can see this when you pull the USB cable, and most often you will perceive whatever anomalies (hum etc.) from the speakers, totally depending on the remainder of your gear. And if not that you will see that the DC Offset of the NOS1 is not right anymore.

While working on the USB PCIe cards, I now suddenly see an opportunity ...
An opportunity to maybe avoid this strange over-PE connection;

If I recall correct the chassis of the PC is involved here. I for sure forgot how exactly, but something like : The PC's chassis is connected to PE, and why is USB too ?
Back at the time (when I developed the NOS1-USB) it looked mighty dangerous to "utilize" this behaviour because I could expect that not for all PC's this would be the case (btw, what about laptops on batteries ?). Still, no single NOS1 fails on this setup (see DC Offset protocol), so it will be the same for all PC's. But, I am not even 100% sure that it *really* is about this, whatever I am talking about which I hardly know myself. Now :

Supposed that the specs of USB are such that by some path it has to be connected to the PC's chassis. Mind you, I don't even know whether this will be the "shield" part (if in order at all) or just signal ground or something which is even plainly wrong (which is what I prefer to think). Thus, any MoBo connector is (has to be, oughts to be) setup like this;

PCI(e) cards do the same, or at least can do the same, by accident or not; We already found that at least for some investigated cards the bracket with which it's screwed to the chassis is connected to the ground planes of such a card. So, ground of the card is connected to the PC chassis, and *that* is connected to PE. And so is USB ground connected to PE.

Already starts to sound nasty eh ? Yeah, maybe through our speakers too.

Now, with such a card we at least have the opportunity to cut the "USB ground" to the chassis and thus to PE. Just remove the bracket ...

Well, sort of, because a PCIe card really isn't anything to comfortably use without it fixated (way small footprint of the PCIe bus connection) so it seems a better option to put some heatshrink tube around the whole bracket (maybe now it won't fit anymore ?) plus use a plastic (nylon) bolt to fixate it. Also not really easy because you need to have such a bolt. Heat shrink tube of the proper size the same of course, but your wife will have a hair dryer (to heat the tube) or use some torch.
Still, when it is about trying only, removing the bracket will do, *if* the bracked can come off in a fashion that it can be put on again ...
All 'n all it's possibly not even easy to try this.

What will happen next I can't predict. I mean, for those with the NOS1 DAC it just as well might happen that the groundloops now go where I don't want them to go. Maybe it hums, or low oscillations will be audible (like when the USB cable is not connected - but this won't happen to everyone).
Possibly you now have a 10 fold more difficulty to get the DC Offset right, but KNOW that it still can be OK. Just keep on trying (referring to the NOS1 Install Guide here).

You may wonder why I am not trying this myself ? Well, in the few hours in a day that I can work on this, I usually run into other things first. Finalize the software of a PC which has to go out - talk through findings of the day about something (always working on new things/designs) - just a tad too much.
And yes, while typing in forums is an easy task, but I do that somewhere else and harmlessly.

Anyway, for now this is theory only and possibly someone sees something wrong with it. Also, the Silverstone card which actually triggered this subject possibly doesn't even have its bracket connected to the ground of the card. And no, I did not check that either (should have done that yesterday). But *if* it doesn't have connected the bracket to ground, we already can see a reason for the card sounding so good ? Thus one way or the other, I think this little subject could be of importance.

Regards,
Peter

[juanpmar]

Hi Peter,

I made the isolation you said, I believe it is well done. No problem to set up the NOS1 DC offset but it can be done easier if the USB cable is disconnected.

I hear less hum, if any, putting my ear touching the speakers but my speakers are not very sensitive. I did it just to see what happens as you suggested and for me the sound is fine. Better? I really can´t say for sure but I´d say yes. Don´t trust me, I usually find better sound after I make some modification, I´m very prone to placebo  . Anyway the sound is really good, let me say that with less Sssss (I mean my own Sssss not sure either if it is the same Ssss as yours).

To isolate it I used heatshrink tube but only in the upper part of the bracket, it is fixed with blue-tack. I´m sure that the card or the bracket is not touching the PC case, the white piece that can be seen in the green plastic is blue-tack (white) to separate the bracket from the case. As there is not screw the card is like floating, there is a gap between the bracket and the case but the connection with the PCi slot is correct.

Regards,
Juan

[PeterSt]

Hey !!

That is great Juan !

No problem to set up the NOS1 DC offset but it can be done easier if the USB cable is disconnected

But are you now saying that you sure saw a difference with getting the DC Offset right compared to when the card was connected to the chassis ??
If so, it tells something !
So ?

Peter

[juanpmar]

Hey !!

That is great Juan !

No problem to set up the NOS1 DC offset but it can be done easier if the USB cable is disconnected

But are you now saying that you sure saw a difference with getting the DC Offset right compared to when the card was connected to the chassis ??
If so, it tells something !
So ?

Peter

Not, I don´t think there is a difference, it took some times to get it right like always. The only different thing I saw is that repeatedly -1 appeared in one channel and always in the same channel, five or six times. I haven´t seen it before so persistently. I don´t know if that means something.

When I commented about the DC offset I only wanted to say that I´ve found that to set correctly the DC offset is easier when the USB cable is disconnected from the PC. Today I disconnected everything and I had to set the DC offset again and this is why I made the comment, also because you said something about that.

On the other hand I don´t think that when you disconnect the USB cable that connects the NOS1 and the Pc you loose the DC offset, am I wrong?

Juan

[PeterSt]

On the other hand I don´t think that when you disconnect the USB cable that connects the NOS1 and the Pc you loose the DC offset, am I wrong?

Not per se, but not per se correct either. Let's say "it depends".

When I commented about the DC offset I only wanted to say that I´ve found that to set correctly the DC offset is easier when the USB cable is disconnected from the PC.

Now it depends on what you were used to regarding this, but all I know quite certainly is that with the USB cable DISconnected it is far more difficult. This should be "standard" (so, also for you).
Notice that I am seeking for different behaviour to this regard, and certainly not better behaviour or anything (I mean, I don't care about that at this moment).

So ...

I hear less hum, if any, putting my ear touching the speakers but my speakers are not very sensitive.

What about this one then ? Because of our both English this can go two directions :
1. I hear less hum if I heard any hum at all ever, but certainly not more;
2. I hear less hum now, actually nothing anymore hum than before;
What is your choice ?

Peter

[juanpmar]

What about this one then ? Because of our both English this can go two directions :
1. I hear less hum if I heard any hum at all ever, but certainly not more;
2. I hear less hum now, actually nothing anymore hum than before;
What is your choice ?

Non of them but the fault was mine. What I wanted to say is:

Before isolating the card I heard a little hum. Now, after I isolated the card I can´t hear any hum. If I put my ear touching the speaker I only hear a slight ssss in the tweeter but not a hum in the woofer, nothing.

Juan

[PeterSt]

2. I hear less hum now, actually nothing anymore hum than before;

No wonder that you can't choose with such a line of text ! Heck, I don't even know what I really wanted to say there.

So if my conclusion is right that you used the card before but then (without the isolation) you clearly could hear a hum from the speaker, then that is set (but otherwise correct me again). This assumed :

What about when you used the MoBo connection ? was the hum the same ?
I realize that I don't even know whether you used the MoBo connection (I may have been the last of Mohicans doing that ), so if not, just compare with whatever you used prior to the Silverstone.

Mr Noise

[juanpmar]

2. I hear less hum now, actually nothing anymore hum than before;

No wonder that you can't choose with such a line of text ! Heck, I don't even know what I really wanted to say there.

So if my conclusion is right that you used the card before but then (without the isolation) you clearly could hear a hum from the speaker, then that is set (but otherwise correct me again). This assumed :

What about when you used the MoBo connection ? was the hum the same ?
I realize that I don't even know whether you used the MoBo connection (I may have been the last of Mohicans doing that ), so if not, just compare with whatever you used prior to the Silverstone.

Mr Noise

Hi Mr. Noise,

I don´t use the MoBo connection since I was using a Tecknet USB3.0 card and that´s some quite time ago. With the TeckNet card there was some hum but also my speakers were more sensitive then the ones I have now.

What I´m sure now with the speakers I use today is that with the Silverstone I heard some hum before I isolated the card and no hum at all after I isolated the card.

So:

- Before isolation= hum

- After isolation= no hum

 

Juan

[PeterSt]

Ha !
See, this is a typical reason I shouldn't test this myself. I have no hum anyway so how could I hear a difference (but it could hum *with* the isolation). Still it will be possible that it can be seen by measurement, but that too is not be expected really in my situation because I know I don't even see a thing right now. But since I am the only person on the planet which really doesn't have any noise at all (which is only because I can measure it and I'm a bit keen on it all), it could help YOU (all).

Remember, perceiving some hum (or hiss) is not so bad because most will have tweaked it so much down (at full gain of course) that it won't be disturbing. In the mean time though, it will be devistating to good SQ.

And then to think that I have been brewing on this for actually years. It is only that I am using a PCIe card myself now that I could see it (all stupid theory of course). But you wanna know how I really came to it ?

For the first week or so I had been using the Silverstone card without it fixated with the bolt; I was tweaking with other things, and coincidentally it is quite tough to screw it in and out. Below you see a picture of it (sorry it is out of focus) where you see the hole in that bar exactly in the middle of the picture, where under that is the free space where the Silverstone gets in. The hole is for the screwdriver to go through and now try to grab or let loose that bolt with one thumb and finger where each is at the other side of that bar (I know, some pliers do wonders, but that usually fails too when aiming for the correct angle of the bolt opposed to its hole). And I already experienced two times that the BIOS was reset because of mangling there (one time the bolt falling down).
Hehehe, long story ... So finally to the point :

Yesterday, *without* looking at the PC itself, I suddenly was placeboed by a SQ change of the past few days (but I changed more things) with the knowledge that last week I finally fixated the Silverstone. So I envisioned that without fixation maybe the bracket didn't touch the chassis, and this is how I got the idea.
But really, if you reread the first post in this topic, you will be able to see how I first expressed about the readily present knowledge I had that PE is involved, while on the spot of writing that post I just reasoned out how that would happen. I didn't tell this explicitly, but it's really the typing of that post which made me suddenly see. And as usual (say 99% of cases with all these things I run into), all by sheer coincidence; If I didn't have had that card loose in it, if I didn't have experienced a SQ degradation (placebo !), if I didnt fixate the card at around the same time of this degradation *and* if I didn't have the almost impossible to obtain knowledge of the *how* USB is related to PE (which in itself comes from the attempt to power the display over USB and *that* not working well because of how the NOS1 attacks DC Offset which is a story of months in itself ...)
Then we possibly never knew.

All assumed that we didn't make other mistakes throughout this topic and that for example the bracket of the card is not connected to the groundplanes at all. But STILL it could be EMI in that case ...

Coincidental me.


PS: Great thanks for your help here Juan. I can tell you that you should feel proud to be so helpful *again*, while actually lacking the gear (speakers) to enjoy audio from the normal angle. Have a few on me !!

PPS: Someone is going to make me an isolated Silverstone tomorrow ...

[juanpmar]

Peter, I´m never sure if you are laying a trap for us or not  , anyway it's fun and if we move forward it is better still.

This could be a placebo or not but I need (maybe you tomorrow Peter, when someone isolate your card) that other people test the card isolated to see if they hear the same as I do. The highs are smoother and in general all seems to sound better. Really good.

Besides the isolation of the card could be also other circumstances that could contribute to the good sound:

- My Silverstone is getting enough hours on it to start to sound better. I think it has now around 50 hours of running time.

- Today I´ve changed again the CPU speed to 1200GHz and the RAM speed to 600GHz (before I had 3300GHz/1600GHz).

 Hey Peter, thanks for your kind words!

Juan

[acg]

Below you see a picture of it (sorry it is out of focus) where you see the hole in that bar exactly in the middle of the picture, where under that is the free space where the Silverstone gets in. The hole is for the screwdriver to go through and now try to grab or let loose that bolt with one thumb and finger where each is at the other side of that bar (I know, some pliers do wonders, but that usually fails too when aiming for the correct angle of the bolt opposed to its hole). And I already experienced two times that the BIOS was reset because of mangling there (one time the bolt falling down).

This made me laugh Peter.  Personally I remove the screw from the cross-member (through which you put your screw-driver) and cast it aside before I screw in the USB card.  Three screws instead of one I know, it's a pain!!

Cheers,

Anthony

[listening]

Hi Peter,

I unplugged the silverstone card from the mobo and measured bracket vs. molex ground and pci ground - it's connected. I removed the bracket from the silverstone and hopefully will hear no placebos 

Georg

[juanpmar]

Hi Peter,

I unplugged the silverstone card from the mobo and measured bracket vs. molex ground and pci ground - it's connected. I removed the bracket from the silverstone and hopefully will hear no placebos 

Georg

Hi Georg

Great! now we know something more, I hope you hear also some kind of improvement.

Juan

[acg]

Hi guys,

I have done several in/outs with the Silverstone card and have isolated it by simply removing the bracket from the card when I want it isolated and then replacing the bracket and screwing it into the chassis when it is not to be isolated.  It takes a while between re-boots and fiddling with screws (Peter, I even dropped the chassis screw onto the mobo at one stage and held my breath when it was time to restart).

Regardless of whether the bracket is in place or not I have a hum through my speakers that is only evident when I put my ear within about 10cm of them...it is standard transformer noise...and is evident in both the woofer and the tweeter.

DC offsets seemed to be the same in both test scenarios.

What did I hear?  Not much difference.  Sometimes I think that with no bracket that there is a little more clarity to the sound but that does not seem to be a stable observation.

Then, instead of powering the Silverstone from the SMPS I power the Molex from my 5v lab LPS.  Nice improvements...blacker black, more flesh (ringing, decay) and separation on the cymbals, the double bass is easier to follow etc..  This is when the Silverstone is bracketed and non-bracketed.

Cheers,
Anthony

[PeterSt]

Shiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiit


So, my Silverstone got a coat yesterday. Not her official one because the size of the tubing we had laying around was too small to do it neatly, but today we'll get a bigger condom and peek some pictures.

Now a few coincidental things happened;
After talking a few things through I sat down to do some analyser tests for other things and only hours later I switched on the audio PC. The PC appeared to have lost its BIOS again
(what is it in that USB3 card-corner ?) and so up to the photos with the special settings. Well, I didn't make those photos and of course the screen it was about (this is a bit hidden) was not on photo. D*mn. A bit of shouting why not ALL screens were taken, and oh well. Point is also that I wanted to see whether those settings now could work from scratch or that reboots were needed etc. (this is all about special PC behaviour and how to get that done). But alas, since it was a week ago only that I had to put in the BIOS settings maybe I remembered.
15 Minutes later the PC was running, and hey, all as intended ! So ha, I now finally know what settings to set. BIOS photos will be up soon.

Same as this all doesn't make sense to you (what the heck is the guy talking about), because of this I totally forgot about the card's isolation. Ciska went out and I brought up XXHighEnd and pressed Play, continuing the album where I left off the day before. Loading takes a few secs, and in that time I walk to the "offiice" corner to deal with emails and such. Notice that this is along one of the walls, say the left wall where the left speaker is some 7 meters from where I'm sitting (speaker is to the right of me now).

I think I just raised a hand to let it land on the keyboard somewhere, but what I recall is that I lost all of my muscles and sort of collapsed. Hard to describe. But envision something like
Whaaaaaat ?

I will try to show you how I behaved. Remember that 9 year old "aria" (Dutch) girl Juan put up ? Here : http://www.youtube.com/embed/cWWqp-IVfb8?rel=0
Scroll to 1:40 in that video and watch the guy in the right (also see picture below).
You really should briefly look at that YouTube because otherwise you will never understand the impact of this change. Seriously.

I stood up from the chair behind my desk (watch the guy doing exactly so) and walked around the half-height wall there so I could be in the middle of the room, leaving behind my jaw;
Already from that spot behind the desk I heard the sound everywhere. So envision me looking around me what was happening, at first not even realizing this was the music which just had started. This could not be ...

During my little walk around that wall (which is a walk towards the left speaker) I noticed a MOST STRANGE phenomenon : there was no music. Yeah, how to describe such a thing. Better could be "music came from nowhere" (which somewhat resembles "it is everywhere").

If I could detect anything from music source it was from the upper far corners above the speakers (3 meters high and my "corners" are round). What came to me was not a wall of sound, but something which could be called "inside out" and a headphone feeling. Maybe I can call it "from within my brain only". It felt like I was hypnothised or drugged or something which was so odd that it couldn't be true.
The next half hour -being alone in the room- I had only be expressing outloud "shiiit" this and that and "what ?" and I don't know anymore. Envision someone virtually banging his head to a wall and with a bang per word "HOW-IS-THIS-POS-SI-BLE" (6 bangs and enough to wake up I'd say).

What explicitly slipped my mind was that I received a third ear. That third ear made me localise better. I also thought "ok, this must be the real start of my 3D localisation project I started 'Phasure' for (never mind for now if you don't know about it).

Let's remember, I just continued the album from the day before. It was as if the tracks I played yesterday were supposed to be 3D scape sound exposures and test signals. Music floating from left to right and back and towards and up and down and, man, I can't describe it. And *again* I noticed that actually there was no music. Such a thing can only happen in a dream.
When the album was finished I put it back to track 1 to relisten to what I should have heard the day before. But no, that now too were "test signals".
In the mean time all of the tracks went into my Demo Gallery, but I think after the 5th I stopped that because that's not much useful.

With the notice that this was a new album to me in the first place, after that I put up a known one and the story was the same. Nothing is producing music but still it is in my head. Call it the ultimate layed back. But inside out.
No more lyrics.


So, somewhere within that first album Ciska came back in. Of course I was watching her to see her response.
Nothing. Hmm ...

Ok, so we'll go the usual route and sneak in questions;
P: Don't you notice something maybe ?
C: Oh, don't know the music.
P: Nah, just general.
C: Don't know the music; too difficult. And always your sh*t ambient.
P: No no no, just try to make up anything regardless. Just try.

Then within 7-8 seconds this came out :

C: Far away. Speed. No, more speed. Spacious.

Ah ok. So force someone, and within 7 seconds you knew after all eh.
Next came the question whether this was from the working BIOS settings and my response was "forgot your condom excercise already ?". Aha.

Maybe women need some focus, but while she too now knew what it was about (or could be about), automatically this came out :

C: So clear (like in crystal);
C: Loose;
C: Soo clear (now like in can hear/understand all the words);
C: Oh, those drops ...
C: Super much detail;
C: More intense.

I was surpised about the last one and envisioned that only women can make this up. However, now I'm writing it, I feel the resemblance with my "from the brain". And stupid thing is, she did not come up with anything like that, but I now thus see she maybe did after all ("intense").

The drops is almost the best description; possibly all of us have been in some 5.1 arrangement and watched something like a waterfall movie. Now drops could sound all around you. That's what happened here too, and when she said it I didn't even hear drops (the music was not meant to let hear drops) but the way you are in a cave and drops are dripping all around you is what this is about. All around you.

While she mentioned the speed which I didn't notice as an element so much because this actually did not allow me to listen for elements as such, I am not sure but throughout further listening it well can be the key "logical" phenomenon doing this. I say "logical" because probably it is nothing physical. Only about infinitely more separation or something.
BUT
While I always feel the woofers when something extraordinary seems to spring from the speakers and because a track was playing of how I know the woofers feel, I now did that too. And now things get dangerous ...

Maybe people have read about my W8.1 experience, how it muffles the sound and how my woofers excurt so much that it becomes unresponsible (don't play louder or they'll break);
Now it was the other way around. Hardly any movement at all.
(Bert, might you read this : Your 15"ers over there).
Listening to it (before feeling) gave me the impression that bass was now so lean that (and again !) it actually wasn't there. Well it was, but I could not focus on it. This, btw, is really really hard while -with this track now too- the sound scape (mind the scape which might be poor English but which is in my mind now all over) floating from, well, everywhere. I just can't understand it.
For those who experienced it (e.g. Paul), the sugar cubes in quadruple. No, more. And really, one of the first things which came to me is that it seemed that the SPL was equal everywhere in the room (sugar cubes did the same - I measured).
It can't be expressed as "sound does not come from the speakers" because it already didn't. Instead I now have 20 speakers throughout the room. And a third ear.

So, all what happened is that USB shouldn't be connected to PE anymore. And let's remember, noise and such I already did not have (-120dB after gain hence measured at the output of the amps (at full gain)). It is totally freaking unbelievable what this thus does. It feels like 10 fold any previous "huge change".

The biggest change ever.

Peter


PS: Changed the title of the topic.