![]() ![]() |
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
104.60.35.252
In Reply to: RE: so you don't know how to isolate the quiet side from noisy side? posted by Leo loves music on April 24, 2025 at 10:52:43
Can you spend time on my other query please?
And do enlighten. How would you have fixed it without fiber?
Cut to razor sounding violins
Follow Ups:
1. Let's examine equipment grounding for components like amplifiers, preamplifiers, and DACs. When a power cord's ground is directly connected to the equipment signal ground and RCA interconnects are used, a portion of the audio signal can inadvertently travel through the power cord's ground conductor. Due to the power cord's typically lower impedance compared to the interconnect cables, a significant portion of the signal current may take this path. For instance, with a $1000 silver interconnect, potentially 50% of the signal could be shunted through the power cord, effectively meaning you're hearing a significant contribution from the power cord. A similar issue arises with SPDIF signals, where noise on the power cord's ground can contaminate the digital data. To mitigate this, inserting diodes or a 10-ohm resistor between the equipment ground and the power cord's ground can help. This approach aims to maintain safety while impeding signal and noise propagation through the power cord.
2. For SPDIF data signal isolation, fiber optic connections inherently provide it. However, with coaxial connections, a pulse isolation transformer is necessary, as previously discussed in my thread regarding the Tascam CD-200 and Cambridge CXC modifications. The transformer's secondary coil must be AC/DC isolated from the output equipment (streamer or CD transport).
![]()
![]()
I def dont understand.1. I have a switch that fed a poe device that feeds a spdif output device to the dac. So ethernet to spdif. I changed the connection at the switch to be fiber optic. So its fiber ethernet to a new fiber/ poe device. Then copper ethernet to the spdif converter. So 2 changes. Fiber ethernet out of the switch instead of copper and a better psu for the PoE than the one in the switch. The new device has a psu that has 2 prongs. The dac has 3 prongs. Where do i put the resistor? (Pict attached. Switch on bottom, yellow jumpers on right are fiber to the new fiber/ Poe injector. Psu is ifi 15v )
2. Not using fiber for the spdif connection. Using coax. The ethernet/ spdif device outputs aes, coax and optical. Optical and Aes are not currently used.
Cut to razor sounding violins
Edits: 04/25/25
The picture you tried to post didn't load but my take on it is this:If you don't have an audible problem I wouldn't mess with it.
In the scenario I presented the Inmate was having HUM issues with his streamer in the system. Switched to cheap CAT5 patch cable to the streamer with no metal RJ45 connectors. Ground loop broken. No HUM. Problem solved.
Edits: 04/25/25
Thanks for letting me know about the pict.
I certainly have a different take. Always trying to improve. Always.
And while I def get the "if it aint broke don't fix it" mentality I don't apply it to audio. If I can improve things I will try. And often we have problems we aren't aware if. Like CDs. Before ripping and computers and then streaming, no one knew how much cds really sucked.
Its like $130 to add the fiber in my case and even less if you don't have PoE. It seems to have cleaned up some things.
When you see the pict you will see my audiophile cables and maybe if you look close the isolation. Even the networking components get basic isolation.
![]()
Cut to razor sounding violins
than overt hum.
I use unshielded CAT6 cable for the long run upstairs to avoid that potential problem.
"There is more to transmitted noise than overt hum. "Of course, but I was describing the overt HUM issue CAUSED BY designer audiophile cables.
Relative to audio all data transmission might be described as transmitted noise. Extracting the data is what counts while the rest is ignored.
Edits: 04/25/25
this is one case where shielded cable may be detrimental.
Sounds like the baby and the bath water, like you are saying ALL audiophile ethernet cables cause hum. And if noise gets to the dac it certainly seems audible. Sure it wont mess up your printer but if your system is resolving you might hear it depending on your hearing.
But sure its diminishing returns at some point.
That said, you can buy 100 cork testube stoppers for $8. Thats enough footers for a whole system. Start there and see.
![]()
Cut to razor sounding violins
"...you are saying ALL audiophile ethernet cables cause hum."
Not at all but any Ethernet cable with its shield terminated at each end to metal RJ45 connectors CAN cause ground loop hum in some systems. That was proven right here in the Asylum. As it turns out many such audiophile cables CAN cause ground loop hum too.
" And if noise gets to the dac it certainly seems audible."
Seems is the operative word. A controlled test might include an Audio Precision analyzer at the output of the DAC with:
#1) a length of cheap CAT5e cable going to the streamer
#2) same length high-end shielded *CAT7 or CAT8 cable to the streamer
Measure 'noise' at the output of the DAC with Audio Precision analyzer and compare the two cables. Has anyone done such a test? I'd like to see it out of curiosity. And even IF the high-end cable measures better it won't matter because the SNR is already well below what's required to perfectly extract the data.
*CAT7 was never an officially recognized standard and in fact CA6A is the better cable.
Look at all the bogus technical marketing speak. None of it even matters for the intended application:
![]()
![]()
Isn't it almost impossible to have grounding issues if you have a fiber network? I mean sure jist being in the same house is often enough regardless
I might be able to test. I have 2 computers one gen purpose and one audio only. Its currently disconnected but when it was, one day I heard my general pc's mouse noise. Took a while to trace but I eventually realized that the audio pc was not EXACTLY headless. I had reconnected the hdmi cable to the audio pc and never disconnected. They were tied together at the TV (monitor). Disconnected and mouse noise gone. Can see if its diff with fiber. Though the devices share the same outlet so the ethernet connection may have nothing to do with it.
Yea. ANOTHER audio thing to figure out. Can't get DSD to work with my Tascam DA-3000 through Spdif-3 bnc connections. My Antelope Zeo stopped working with IOS 18.4.1 and the Neutron player seems to only output 32bit which means half my gear wont work. :(.
![]()
Cut to razor sounding violins
I know what I don't know. :)
That seems to be the concept and why Sonore created their fiber based version of my endpoint. Stray currents and noise do travel along conventional Ethernet which is why I use the Tripp Lite filter.
" Stray currents and noise do travel along conventional Ethernet"
That's why the Ethernet standard requires transformers at each end for isolation. Transformers are very good at generating and receiving differential signals while providing high common mode noise rejection. This makes them an excellent combination with twisted-pair wiring.
Just a thought.... If optical network coupling provided noticeably better audio quality from a streamer you would think that more manufacturers would incorporate optical. IMHO, it's mostly marketing and a way for some companies to differentiate their product.
Just my 2-cents worth as I can't argue with what other people hear or think they hear.
![]()
There are market factors. Fiber is more expensive and though it is cheaper than it used to be, its not exactly plug and play like copper. As i understand it not all modules really work together though they should. For example I was strongly advised to buy the modules from my switch manufacturer and not another cheaper brand.
You need a special switch and cables and need to be careful with the runs and bending. Much easier to sell plug and play use existing equipment, etc. its kind of like I2s in that some manufacturers offer it but most don't.
And if your premise about copper ethernet is true (that transformer and utp wiring handles noise, why do they make ethernet isolators for hospitals? What am I missing?
![]()
Cut to razor sounding violins
Optical GBICs and SFP's along with optical cables are designed to run at certain wavelengths so it does matter which compatible components you choose for your data center application. Optical cables are usually color coded. Think of an optical cable as a waveguide for light much like metal waveguides are used for microwave RF propagation.
Most everything you said is true in a data center BUT- optical Ethernet to a network streamer for audio applications offers no practical benefit - IMHO.
![]()
I looked at the isolator link you posted. Its the exact one I am using between the walljack and my switch. Do they add? I have a 2nd one somewhere and could put between the computer and the switch, though the fiber should make that unnecessary I would think.
![]()
Cut to razor sounding violins
Its the exact one I am using between the walljack and my switch. Do they add? I have a 2nd one somewhere and could put between the computer and the switch
Potentially yes because you've introduced a new source of noise in the second case.
In the case of Ethernet ports on networking gear they all have small isolation transformers onboard so the idea of ground isolating Ethernet is kind of a waste. But audiophiles have found ways to mess that up with high-end shielded Ethernet cables with metal RJ45 connectors at each end. These have the cable's outer shield terminated at the metal connector body at both ends of the cable. Perfect scenario for creating ground loops. Not too much of a problem in data centers, big problem in audio.The solution for audio - don't use high-end designer Ethernet cables.
There was an example here in Audio Asylum. Inmate switched to cheapie CAT5 patch cord. No more ground loop hum
P.S. I think it was inmate Ivan303 from Santa Fe NM with the Hum problem. We haven't heard from him in several months.
Edits: 04/25/25 04/26/25
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: