USB Cables... again

[manisandher]

Just a quick post to share my latest USB experiences...

A couple of weeks back, I decided that I really ought to try my Berning OTL SET amp with my 114dB/w horn speakers. The amp is a strange design - DC-coupled, no output transformer, switch-mode power supply, etc. It's pretty quiet by SET standards...
- Signal to noise: 92 dB (20 kHz bandwidth)
- Line-frequency hum components: 60 Hz: -94dB; 120 Hz: -100 dB; 180 Hz: -104 dB

... BUT, with my ear fairly close to the MF/HF horn I could hear some 'crackling' noise coming through the speakers as soon as the PC was switched on. And lo and behold, I could 'hear' my mouse whenever I moved it. I could also 'hear' the HDD doing its work. No question about it, this was noise being propagated from the PC to the NOS1 via the USB cable.

A couple of years ago I bought a USB cable from eBay which had separate data and power lines, with the ground passing through the data line (not the 5V power line). I fished this cable out and connected it to the PC via only the data line. Et voila, no crackling noise whatsoever...

... AND, a totally different sound too. A much cleaner, fuller sound, with much deeper bass. On first hearing, it sounds too laid back and almost dull. But with prolonged listening it sounds more realistic. The highs, although more subdued, are sweeter. And the real test is that you can listen to music for hours without any agitation.

[I tried the cable (just data) in my office system too - totally different mobo, CPU, PC configuration, etc, and the change in sound was very similar.]

Not really sure what to make of this. My mobo is the same as the one that Nick has just replaced. It could be that the mobo is cr*p and having just the USB data line connected helps. As to the change in sound... Well, either it's better because of decreased USB noise entering the NOS1. Or it's worse - the USB cable suppressing transients and the like. Will try a few other things and report back.

But one thing's clear to me - even the NOS1 is affected by USB cables and how they deal with noise from the PC.

Mani.

[juanpmar]

Just a quick post to share my latest USB experiences...

A couple of years ago I bought a USB cable from eBay which had separate data and power lines, with the ground passing through the data line (not the 5V power line). I fished this cable out and connected it to the PC via only the data line. Et voila, no crackling noise whatsoever...

... AND, a totally different sound too. A much cleaner, fuller sound, with much deeper bass. On first hearing, it sounds too laid back and almost dull. But with prolonged listening it sounds more realistic. The highs, although more subdued, are sweeter. And the real test is that you can listen to music for hours without any agitation.

Mani.

Hi Mani, this issue of the influence of the usb cable looks interesting. If the NOS1 don´t use the power part of the usb cable I don´t know how the cable can influence the sound (old discussion here in the forum). Anyway could you provide some details of the type or brand of the cable you use?

Best regards,
Juan

[Flecko]

+1. Thanks Mani, sounds interesting.

[manisandher]

Anyway could you provide some details of the type or brand of the cable you use?

It's a 1m Ridge Street Enopias USB cable. I paid USD150 for it from eBay. Apparently, it was >USD1000 new, which is simply crazy. There is very little information on the Ridge Street website about the cable - it looks like they may have stopped making digital cables - it doesn't look anything special. But it has certainly totally killed the crackling noise. Maybe the added change (improvement?) in SQ has nothing to do with the cable's properties (materials, geometry, etc), but rather the separate the data and power lines. Certainly, the NOS1 seems perfectly happy just connected to the data line. But having the ground connected is essential.

Mani.

[Jud]

Would be interested to know what you think of the Mapleshade Clearlink USB Plus, which I use, and Alain did (not certain if he still does).  15 or 30 day satisfaction guarantee (not sure which).

Best I've heard, including one twice as expensive.  What made me think of it is the mention of separating the power leg.  In the Mapleshade, the power leg and the signal and ground legs (latter two twisted together) are loosely held within a sort of open weave mesh.

[AlainGr]

Mapleshade Clearlink USB Plus are still my USB preferred cables I have them between my music PC and the NOS1. I like the philosophy behind it and of course the SQ it provides.

Now I have to run to begins this day will lots of jitter in my precipitation

Alain

[PeterSt]

I added this to that other topic from where you referred to this (and then removed it there). Here it looks a litlle unsollicited, but I think it can be important anyway :


Of course noise still matters; if you add it.
So far I have not seen a situation that more USB noise added "brightness" of any kind. The other way around (soften the sound), mucho. This is no guarantee of course ... only that it would be the first time to see it the other way around. Also, you won't be the first one who is fed up after a week or so with that more blanketed sound (I know, my words). So just keep in touch with it all.
And mind the bass. More bass ? then more noise for sure.


Try the influence of a separate "ground" cable in parallel to your RCA interlinks. Similar things may happen. Whether things change depends on many things. But just saying ...

Regards,
Peter

[manisandher]

Of course noise still matters; if you add it.
So far I have not seen a situation that more USB noise added "brightness" of any kind. The other way around (soften the sound), mucho. This is no guarantee of course ... only that it would be the first time to see it the other way around. Also, you won't be the first one who is fed up after a week or so with that more blanketed sound (I know, my words). So just keep in touch with it all.
And mind the bass. More bass ? then more noise for sure

Hmm... I think it might be the first time then. I getting absolutely the sweetest sound I've ever achieved, I think. Maybe not 'more' bass, but deeper and paradoxically more tuneful.

But will keep on eye on things.

Mani.

[PeterSt]

Mani,

Maybe not 'more' bass, but deeper and paradoxically more tuneful.

After I posted yesterday I realized I always say "more" and this time I should have explicitly said "deeper". But that was only after I heard my own system again ...

As I told in the other topic, I had a small hum coming through the horns which I could solve yesterday. And, gone was the SUPER deep bass. I have created too very lengthy posts about it with all my reasoning how that happened (through the Orelino speaker). Both times I was (too) reluctant to post ... No matter I could not hear colouration I felt it could not be right. And it wasn't. It was impeded by "noise" ...
Still that bass (sub-low) was crazy.

And so I'll bet you it will be wrong. Expecially the combination with "sweet". You really are not the first (if it already wasn't you yourself ??). But we'll see. And if you like it better, well, you just do and why not !

Best regards,
Peter

[manisandher]

And so I'll bet you it will be wrong.

Well, I've just been playing around with a few things this morning and... I'll take your bet!

There is no question in my mind that the sound is better with the power line disconnected. With it connected, the amount of USB noise is simply not acceptable - it's easily audible when the music is playing.

Apart from the USB noise, the sound itself has that 'digital glare' which I believe contributes to 'false detail'. And unlike you, Peter, I'm very familiar with this effect. I used to own an Esoteric D70 DAC (my first introduction to the PCM1704 BTW). It was a very educational unit. You could connect it to the transport in one of three ways:
1) PLL. Initially this sounds very nice, but you soon realise it's 'digital glare' and 'false detail' all over.
2) RAM buffer. The worse. Dead. Totally dead. Yes, just like your 'blanket' description.
3) Word clock (with additional BNC cable). Easily the best. Loads of detail, yet clean and pure.

And the difference in SQ that I'm hearing between using a regular USB printer cable and the split power/data cable is very similar to the difference between 1) and 3) above. I'm pretty certain anyone listening to the two cables would prefer the latter.

BUT there are a few caveats that I should share:

1) I'm using a 5m USB cable extension before the 1m USB cable to the NOS1. My music PC currently sits in my basement. It's a totally silent machine (apart from the HDD), so I'll bring it up to the main room some time and see if shorter USB runs makes a difference. I did experiment with USB cable lengths a while ago and couldn't hear a difference between 5m, 3m and 1m runs. But with my current USB noise issue, it might be worth experimenting further.

2) My mobo may be totally cr*p. Nick found this with his. It could be that the mobo is creating more USB noise than normal. I believe Brian also has the same mobo, but I've not heard him mention anything untoward about the sound he's getting from it.

3) The USB noise has only now become an issue because until very recently I've been using balanced connections between the NOS1 and power amp. With balanced connections, there is absolutely no USB noise, irrespective of USB cable type or length. Indeed, with the Sauermann monos, the only thing you can hear with your ear inside the 114dB horn is some low-level white noise. With the Berning SET, there is noise of certain frequencies, but it's not mains hum - the amp uses a switch-mode power supply. I'm generally against these, but understand why David Berning uses them (for a whole bunch of reasons that he describes on his website). In any event, he's a smart cookie, so I'll assume he knows how to make a proper power supply.

Peter, not sure how to conclude our bet. If you have any ideas, let me know...

Mani.

[PeterSt]

Haha, you changed the rules !
... sort of.

So, first you change "something" which even makes USB noise audible through the speakers, and next ... well, next don't expect jitter to be crazily high as a first thing.
(you have read that John Swenson articles (especially the first part IIRC) so you know what happens with that.

So, true; when you start out with USB noise (somehow) things are the worst. The properties of that are similar (more deep bass, filtered highs) with the difference of the bass being not tight at all. What you have "added" to the situation is that the noise is audible through the speakers. Well, "noise" of whatever source or kind, because USB noise is just one 8Khz (and 16KHz) tone and this will NOT be audible (120-130dB down or so).

If you think you solved this specific issue by means of that power line disconnected (well, that is what you say now), then I will believe that unconditionally. I do *not* believe yet that you don't have USB noise, which is a first thing to present itself when something is wrong in general. But possibly.

In the far end you're talking about a completely different issue. For my part not USB related at all.
And I think (or am fairly sure) what you applied with your USB cable, is not about the USB cable at all. Just another ground connection or cut of that.
But not easy to judge, because when you now way "without power" ... well, wasn't that a possibility in the first place (the NOS1 doesn't need that) with my idea that in the end it works out the same for its gnd connection (as the data line's gnd).

All bets are off now. As long as you can't win (or to ensure that). Hahaha.
Peter

[manisandher]

OK, I'm pretty certain Peter will tell me that I'm doing totally the wrong thing, but I'll go ahead anyway...

I've drawn up my current grounding/PE scheme. I use a 'star arrangement' with a 3m copper chemical rod in the garden. The guy who installed it believes it's around 5 Ohms to earth, based on the clay soil in which it's inserted. In any event, it should be much better and electrically much cleaner than the house PE. (I won't go into details here, but I use 2 balanced transformers - one for the basement and one for the listening room - to isolate this earth from the house PE.) There are actually 12x 3m earth wires connected to the solid copper bus - 6 going into my listening room and 6 going into the basement (I've only shown 5). Now the interesting stuff...

Every component has its PE connected to one of the 12 earth wires. Being of equal length, each of these should sit at exactly the same potential wrt to earth.

Now, in a regular setup, the NOS1 will derive its ground from the USB cable (I'm not talking about the NOS1's PE here - this is derived from 4 and is just connected to the chassis). But the ground is itself derived from the PC's ground plane (which I'm assuming is ultimately derived from the  PC's PE connection, 1).

Now, I'm using a >5m length of USB cable. Everyone says this is a bad idea. I'm not so sure. Pat over at AR-T uses very long lengths to make his spdif cables, to reduce the affect of reflections - at MHz freqs this matters. And maybe there's a case for long USB cables too? But of course the major downside is the susceptibility to RF interference with such a long length. So what I've done is to connect the USB cable's screen at the PC end to earth (2) and at the NOS1 end to earth (3).

But this has now done exactly what Peter wanted to avoid - use the PE for ground in the NOS1. What puzzles me though is that this is unavoidable isn't it? I mean, where does that ground get it's reference from anyway? The PC's PE, no?

Anyway, what I really wanted to talk about was the affect that having 3 connected or not has on the sound. It totally changes when it's connected. It's cleaner with what sounds like a quieter background. It's darker with less 'sparkle'. But along with this, there's much less sibilance and edge to the sound. It's much easier to listen to. 'Blanketed'? I wouldn't say so, but compared to 3 unconnected, it certainly sounds less bright and lively.

Any thoughts as to why connection 3 in my scheme is having such a profound affect on the sound?

Mani.

[CoenP]

You not only referenced to the earth rod but also ceated a new path for currents. Current runs in loops like from pc to dac via 2 and 3 (and back through the usb cable) or via 2 and 1. Etc etc.

Like stiffness in mechanics, the current (like displacement) will yield toward the least resistive (most stiff) path.

And then we have induction: the wires acting as a current generators within ac fields.

Currents all over the place. Some for the better, others for the worse. This is trial, error and learn material.

Maybe a bright guy is able to solve this puzzle... Peter? Anyone?

Regards, Coen

[manisandher]

Reading the John Swenson articles, it's clear that there have to be return currents or things just wouldn't work. But they create ground plane noise. What I'm trying to do is to feed this noise away from the components and into the earth by using a star arrangement to reduce as much as possible any potential difference between components.

Mani.

[CoenP]

Reading the John Swenson articles, it's clear that there have to be return currents or things just wouldn't work. But they create ground plane noise. What I'm trying to do is to feed this noise away from the components and into the earth by using a star arrangement to reduce as much as possible any potential difference between components.

Mani.

John S also talks about the groundplanes being polluted by circuit noise. This means that you have to choose a quietest point on your grounding where you connect the earth reference. If you make multiple refs on the ground you create the possibility to run currents through the ref (aka loops) especially when this is a low impedance path for the noise involved.
So if the pc is refecenced to ground at a certain point and you add another ground reference at the usb terminal you have two referencepoints. My point is that the effects of this are now rather unpreditcable.
I would rather experiment what the difference is between referencing at the pc ground at the plug versus referencing at the usb side.

Since the NOS1 circuitry is internally "floating" I would expect that it is not adviced to ground the poweramp circuitry. You will create a path for currents to pass through the NOS1 (that is that a low impedance path is via the NOS1).

I realise this need more thinking and experimenting.

Regards, Coen