Jump to content

Kill the interconnect!

Recommended Posts

Think I've read it all now!




Doing away with your interconnects and hard wiring your setup. Anyone here performed such a lobotomy to their own kit?





Intel iMac + Beresford TC-7510 + Little Dot MK III + beyerdynamics DT 231 = Computer audiophile quality on the cheap! --- Samsung Q1 + M-Audio Transit + Sennheiser PX 100 = Computer audiophile quality on the go!

Link to comment

This is just true.


By the most accidental circumstances I have developed a very absolute means of measuring this. Per connection 3-5dB of noise is added !


Not that I apply it. But I should ... :-)




Lush^3-e      Lush^2      Blaxius^2.5      Ethernet^3     HDMI^2     XLR^2

XXHighEnd (developer)

Phasure NOS1 24/768 Async USB DAC (manufacturer)

Phasure Mach III Audio PC with Linear PSU (manufacturer)

Orelino & Orelo MKII Speakers (designer/supplier)

Link to comment

who now desigsn for Cambridge Audio, mentioned this briefly in one of his design articles. He had to connect his interconnect to a rusty nail before he found any measurable degredation. So assuming the connection is electrically and mechanically sound (and I presume Peter's weren't if he had that much noise on them), I don't see how soldering all the connections would make any useful benefit. At least one loudspeaker designer (Earl Geddes) advocates electrical connection blocks for his crossovers as they are more reliable than badly soldered joints.


Link to comment

By the most accidental circumstances I have developed a very absolute means of measuring this.


shenzi, If I hadn't put forward this one you would have been right. But I did ...


At this moment I can't explain it quite, merely because I didn't investigate it further, but might a Nobel prize for audio exists ... well ... hehehe


First it needs a very special prerequisite : noise on the mains. This noise must be created by a PWM (Pulse Width Modulation) transformer. Besides such devices exist by explicit means (I mean, cabinets just doing it for various reasons, usually used for industry purposes) they may exist in more common devices like a pump. In my case this is a 380V heating pump, and the revs are controlled by PWM. PWM is a technique that cuts the normal sine of the mains into pieces, and by leaving out pieces there's less voltage to the driven device (the pump in my case). One of the characteristics is that more torque is gained from it.

Btw, this is used in model railroading as well, to control the speed of individual locomotives, while they're all on the same voltage (like 14). It is a digitally controlled means, and therefore flexible and useful.


The side effect of it is that it indeed creates noise on the mains itself. This is square wave noise, or unintentional harmonics if you like. It looks like white noise throughout the audible spectrum, but it is not because the noise is consisting of coincidental adding up of those square wave harmonics.

The "white noise" is perceived because it looks random, but it is not ...


The reason I started to investigate was a by me perceived annoying something, like a very high frequency tone, and it started to bother me at working on a new DAC which consisted of a couple of PCBs in the open, hence without cabinet. It was noisy ...


After a few weeks I got the hunch of taking a measurement microphone and just measure. If I heard it and it was really there, it should be able to get captured by a microphone.

And so it did.


I found a huge peak at around 17.5 KHz which even played well above the level of the music at normal playback levels (which are rather high for me). The peak at this frequency was steady as a rock, or in other words it should be sounding like a pure tone and nothing like noise at all. Although I am fairly sure I can't hear frequencies above 12KHz or so, it was clear to me that this was bugging me.


Of course my "investigation" was headed towards getting rid of the peak/tone, which appeared to be impossible. No grounding could help, no switching over to other audio mains groups (I have a couple) helped, BUT, I noticed that eliminating noise helped. Huh ?


During the process, one of the first things occurring to me, was eliminating injected "normal" noise as we know it. would bring down that peak. For example, when in a normalized situation (meaning throughout the same amp level, same microphone level etc.) the peak would be a -55dB, I could bring it down to -85dB by removing "noise". In my situation of that time this was easy, and it came down to removing the noisy DAC from the chain, and then it happened. For myself this was important, because it proved that / why I did not suffer from that tone before, while it should have been there for years (because the pump is there for years).


Since there was no way that the DAC was generating a 30dB of noise, I started to investigate further.


On a side note, people may know that I am a kind of deep into square waves, their (harmonics) behaviour, and the importance to good audio playback. Without this, I presumeably only had this problem, would have changed the pump (although a 3000 euro job) and let it rest. But instead I started investigating more.


The peak and the FFT characteristics showed me this was a square wave;

My knowledge about standing waves (think about sub woofer calibration and room modes) told me that something like a standing wave in electronics was going on. Could that exist ?


Of course I did quite some more things at finding out what actually was going on, and one of them was taking the micropohone out of the equation, and I measured the elctronics themselves. As I could expect, there was no difference, kowning that when loudspeakers emit the tone, it must be in the electronics in front of it. And of course I already checked whether the microphone was capturing the tone from the loudspeakers indeed, which is easy by moving away the microphone.


More down to the clue, at removing the microphone, I wanted to be as close to the earlier situation, so, I just removed the (balanced) microphone. To my surprise the peak got higher.


From here on I worked the other way around, and removed everything from the capturing device (a Fireface800 btw) but left the mike input open;

I could connect an XLR without cable ... peak went higher.

Connect a XLR->RCA converter. Peak more high than before.

Connect an RCA connector additionally -> additionally higher peak.

Connect a short well shielded cable -> higher.

Connect a longer cable -> more high.

Connect a lousy Y connector -> relatively more high.


In the end this just came down to the obviously worse connectors creating more noise. But be careful here, because I wasn't measuring noise as such, but looking at a representative of it : the peak tone.


Back to the squares and their harmonics, and adding up of waves in electronics (??), it became clear to me that the so called "white noise" which also was visible in the spectrum (at some -100dB) would rise just the same but significantly less. Thus, once I saw that happening, I could explicitly see that when some connector added 1 dB of this "white noise" (which really takes attention because you are looking at an all dancing bunch of peaks which are the opposite of steady), the tone rose with 3-4dB. To me this was the clear effect of the "super coincidental" fact of all harmonics of whatever it was, adding up in that one tone.

NOTE : This "super coincidental" really is not so. It's only the why that is lacking currently ...


At this stage, I can already prove that no one has seen this for various reasons. The most important reason is that no one will have met the conincidental facts coming down into one situation. These are :


- Having such a PWM device in the chain;

- Having such a PWM device that exhibits its mains feedback in the audible area (which is not allowed); THis feedback is common btw (which is different from commonly known).

- Not having a preamp (so the main amps are at full gain all the time);

- Having fairly fast main amps (good square wave response up to 200Khz)

- Possibly my high resolution speakers add up.

- All together the pure coincidence of being able to see the tone above the normal noise level.


This means allows to see the influence of connectors and everything. Everything, really everything creating the slightest amount of noise will show that my this means. But as implied in the beginning, it does not allow (yet) to deal with it in some absolute means. Thus, the fact that I see a connector make 1dB of noise (measured at the "white noise level, not at the peak tone !) does not tell me that this is normal noise. It probably is not at all (like e.g. Cambridge Audio measurements could not show it), BUT IT IS SOMETHING !!


It is something we don't know of yet, and you can bet it influences sound.





PS: I could eleborate much more (which I did but scratched it again), but of importance for understanding is the fact that my called "white noise" is already created by the pump. It is this layer that allows each single source of noise to ride upon, that by itself allowing for the add up in the 17.5 KHz tone.


Lush^3-e      Lush^2      Blaxius^2.5      Ethernet^3     HDMI^2     XLR^2

XXHighEnd (developer)

Phasure NOS1 24/768 Async USB DAC (manufacturer)

Phasure Mach III Audio PC with Linear PSU (manufacturer)

Orelino & Orelo MKII Speakers (designer/supplier)

Link to comment

I've had mains noise in my system before, but isolating digital and analog components to separate circuits dropped it to below the threshold of (my) audibility. Interconnects? That chain is broken on the way to a digital amp by an optical cable.


If this was really a serious concern, though, a better answer than hard soldering interconnects might be eliminating the need for interconnects by integrating components. The potentially quietest system would be built around something like an A/V receiver with DAC, pre-amp and power amps all together. Better yet, I suppose, would be something like AVI's ADM 9.1's -- DAC, pre-amp, power amps, speakers, all in one box, maybe all on one board for all I know. Of course that would leave audiophiles with nothing to tweak and no synergy to seek, which is a great passion for many.


I think it's largely this lack of tweakability that has made integrated components (mostly amps) unpopular with audiophiles. In theory, they achieve quite a few audiophile goals.


It isn't just theory, though. I haven't heard the ADM 9.1s, but I've spent quality time with the better Yamaha, Pioneer Elite and Denon AV receivers with integrated DACs, etc. In their pure direct modes, which bi-pass all the surround sound processors, tone controls, etc, they are extremely quiet. When you think about it, this makes sense. They are bi-passing circuits and wire that is already insignificant compared to a couple of sets of interconnects and the circuitry of an entire outboard DAC and preamp. And what you end up with is quite desirable -- the shortest distance between two points, passing through almost nothing able to add noise or degrade signal quality.


The high current, higher-end Yamahas (starting with the RX-V1900, I believe, at around $1200), have some of the lowest noise and distortion levels available at any price. They are also fast, powerful and very detailed. They do have a sound that is open and bright, though, not exactly tube warm. The Pioneers have more of a warm tone, but a bit less clarity to my ears.


It's not a popular opinion in audiophile circles, but personally, I think the better quality AV receivers, running in stereo direct modes, offer a simple, successful path to great sound. Some of them even allow you to bi-wire or bi-amp using the mutliple amps in the box in stereo, thus addressing another challenge that is critical to fidelity: headroom.


Synergy in a box, pre-tweaked by some of the best audio engineers in the world, for less than the price of many audiophile DACs.




I confess. I\'m an audiophool.

Link to comment

Funny you should mention this Tim, because I've been using a Denon 5800 for the past month and the value on these old receivers is astounding. For example, the 5800 cost $3800 new, but I picked it up for under $1000. The quality of the amp section when I biamped it was comparable to my Moon i-1, which is quite a feat for a 9 year old piece of "mid-fi" equipment as a/v receivers are viewed. The DAC section is good, but slightly worse than that of the Apogee Duet I'm auditioning right now. All in all, these old receivers are the best bargain for audiophiles on a budget.




Link to comment

Having been a fan of the Tandbergs in the 70s, I'm coming back to receivers. I tried the little Sonic T-Amp which, although ridiculously underpowered, sounded as good as my Naim amp. When the chance came to pick up a Tripath-powered Teac receiver, I jumped at it. Wouldn't swop it for a pile of boxes and all those nasty interconnects.


btw, my least favourite interconnects have to be those stiff, high end knitted ones. It seemed like every time I moved one, a solder joint went intermittent. That's probably why decent studio interconnect cable is so nice and flexible. It may also have something to do with my soldering skills, of course.


P.S. Having listened to the ADM9.1s for a short while, they are considerably more powerful than the Teac and gave a very solid, balanced presentation which I don't recall hearing elsewhere.


Link to comment

YMMV, but the way I see it, there is well-engineered, good-sounding gear and there is the rest of it and both exist at all price points. "Mid-fi" is a part of the prodigious mythology of the appropriately-named high ($) end, in my view. Another huge chunk of that mythology, to get back on topic, is the notion that somehow a stack of individual boxes, all containing complex circuits and a foot or so of wire, strung together with interconnects and "synergized" by audiophiles is somehow more pure than a single component integrated by a professional audio engineer. If you think about it more than a few seconds and logical flaws are self-evident.


That your 9-year-old "mid-fi" receiver has a DAC that only sounds "slightly worse" than a state-of-the-art professional field recording interface is telling. Of several things.






I confess. I\'m an audiophool.

Link to comment
  • 4 months later...

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

  • Create New...