Some more audiophoolery: the Lan Silencer

[racemaniac]

https://ifi-audio.com/products/lan-isilencer/

For when you hate all the noise your router is causing in your streaming device due to all the noise being passed through the network cable .

[Coordonnée_chromatique]

https://ifi-audio.com/products/lan-isilencer/

For when you hate all the noise your router is causing in your streaming device due to all the noise being passed through the network cable .

It seems to be the same technology ?

ISOUSB211-Q1 High/Full/Low Speed Isolated USB Repeater
https://www.ti.com/lit/ds/symlink/isousb211-q1.pdf?ts=1716370578378&ref_url=https%253A%252F%252Fwww.ti.com%252Fproduct%252FISOUSB211-Q1

2 Applications
• USB Hub, Host, Peripheral and Cable Isolation
• Medical
• Factory automation
• Motor drives
• Grid infrastructure
• Power delivery
• USB Audio

3 Description
"The ISOUSB211-Q1 has
inbuilt programmable equalization to cancel signal
loss caused by board traces, which helps in
meeting USB2.0 high-speed TX and RX eye-diagram
templates"

[Haenk]

"galvanically-isolated inputs to reduce digital jitter and overall system noise.

The LAN iSilencer boasts premium galvanic isolation technology, where electrical circuits are separated to eliminate stray currents. Audio signals can pass between galvanically isolated circuits to block differences in ground potential or currents induced by AC power."

Oh well. How will a transformer in a network device reduce digital jitter? And what audio signals are they talking about? There are only some Ethernet packages - which happen to contain encapsuled, encoded data, which might be or not be part of a digital audio stream/file.

I can only see a benefit in avoiding ground loops, if the shield is not connected. But they are not even talking about this?!?

[Daixiwen]

Ethernet is already galvanically isolated so all what that thing does is to add an extra transformer in series

[xrunner]

Only $89? They must be new to the audiophool market with such a low price. Raise it to at least $500.

[dave j]

I'm surprised it's a UK company. That name would be too easily pronounced "Iffy Audio".

Although that does seem rather apt for a purveyor of audiophoolery. 

[PlainName]

https://ifi-audio.com/products/lan-isilencer/

For when you hate all the noise your router is causing in your streaming device due to all the noise being passed through the network cable .

Damn! Wish I'd thought of that

[aeberbach]

So at the risk of everyone pointing at me and laughing, I bought one of their products a few weeks ago:

https://ifi-audio.com/products/idefender-plus/

It works. I use a USB audio interface connected to a Mac via USB, with balanced inputs from a PC next to it, a 48V phantom power condenser mic in to the interface with line output directed to the PC. Powered monitors are driven by output from the audio interface.

I thought I was dealing with a bad interface (Focusrite Scarlett 8i6) because there was so much hum and scratchy noise, dependent on computer activity. Like when the mouse was moved there was an audible scratching coming from the speakers, or when the screen changed the pitch of the background noise changed. Moving the volume knob made a whole lot of static. I blamed the PC's onboard sound but using a USB-stereo output made no real difference.

I tried a Soundcraft Notepad 8-FX and a Yamaha AG06mkII and they had pretty much the same background noise issues. (The Yamaha can't actually output the mic to the PC, so wasn't suitable anyway - but did make less background noise, is USB-powered rather than using an external supply). The only thing that cured the noise problem was the iDefender. I could unplug it right now and connect the Scarlett direct to the Mac and I know the noise would be right back.

I believe it avoids a ground loop. My power is correctly wired and earthed according to the tester I have, which does show if ground is not connected.  I would love to know exactly what is going on with it but I have to say not all of this company's products are snake oil. Not talking golden ear audiophile stuff here, the difference is so obvious I could probably show it off by pointing a phone camera at the computer.

[thm_w]

So at the risk of everyone pointing at me and laughing, I bought one of their products a few weeks ago:

https://ifi-audio.com/products/idefender-plus/

It works. I use a USB audio interface connected to a Mac via USB, with balanced inputs from a PC next to it, a 48V phantom power condenser mic in to the interface with line output directed to the PC. Powered monitors are driven by output from the audio interface.

I'm not sure whats inside that thing as it supports USB 3 and just says "breaks the earth".
Are you supplying 5V power externally or just using it alone?

If you just need USB 2.0 isolation, which is useful with a lot of test equipment, you can get these for $6: https://www.aliexpress.us/item/3256802773510544.html

[aeberbach]

Yes I have an external power supply for it. When it is not powered it reduces noise somewhat - I can still hear it but not well enough to record with the phone. When it is powered I can't hear the buzz. Here's the difference it makes, absent then present and powered:

https://youtu.be/KaHYt8Xl_Es

https://youtu.be/qB6Ga6CrsrA

I have a USB 2 isolator for test gear, the audio interface is not recognised when connected through it.

[themadhippy]

hum busters from your local car audio place often  work wonders with noisy computers,there also cheaper than a pair of di boxes and dont require power,the only down side is they aint balanced .

[mansaxel]

hum busters from your local car audio place often  work wonders with noisy computers,there also cheaper than a pair of di boxes and dont require power,the only down side is they aint balanced .

They also sound "meh".

Much as I'd love to love transformers, and our locally produced Lundahl ones are the only ones worth bothering with apart from Jensen, they are heavy, expensive and are an audio compromise.

This in all likelyhood is a Pin 1 problem (Muncy, JAES June 1995). I'd try some extra bonding between chassis in the various components.

As for the OP product, well, yeah. People who fail to understand the layers of abstraction and timing isolation between data transport via IP packets in Ethernet frames and decoded compressed digital audio usually fall for this kind of shit.

[Nominal Animal]

Typical ethernet ports include galvanic isolation via transformers.  The product in the initial post is just an unneeded extra transformer.  You can build yourself one using a standard MagJack and a male RJ45 connector.  Makes absolutely no sense to me, though.

USB isolators can be divided into four categories:

• USB 2.0 LS/FS (low speed/full speed), 1 Mbit/s or 12 Mbit/s, typically using a DIP switch.
  These are commonly based on Analog Devices ADuM3160/ADuM4160 chip, specifically designed for the purpose.  In addition to the isolator chip, you only need four 0.1µF=100nF X7R/C0G/NP0 bypass capacitors and four 24Ω resistors.  The only "downside" in using these is that they need isolated downstream supply for the downstream port.
  The USD $10 isolators off eBay are exactly this, and work quite well.  The only difference in them is the isolated DC-DC converter; their efficiency and maximum output varies, but is typically as low as 200mA maximum.  Also, the cheapest isolated DC-DC converters they use have a typical 10% minimum load, under which their output voltage (5VUSB) may go up to 10-12 V.  None I've seen have a minimum load resistor or a 5.1V Zener, not to mention a 5V LDO, to stop that overvoltage.  Note that this is not a problem with the isolator, but only on the isolated DC-DC converter used; and that varies.
   
• USB 2.0 HS only (high-speed, 480 Mbit/s).
  Expensive ones use a pair of FPGAs implementing the upstream and downstream USB interfaces, with a digital data interface in between.
   
• USB 3 (SuperSpeed).
  A pair of FPGAs implementing the upstream and downstream USB interfaces, often using similar transceivers as used in 10GbE.
   
• USB 2.0 LS/FS/HS (full USB 2.0 support).
  Dedicated AD ADuM3165/3166/4165/4166 or TI ISOUSB211 isolator chips.  Basically "next generation" to the dedicated USB 2.0 LS/FS isolator chips.
  ADuM3165/3166/4165/4166 require an additional 24 MHz crystal or clock input.  ISOUSB211 can benefit from 3.3V and/or 1.8V external regulators on one or both sides to reduce the heat generated in its internal LDO's.  Two-layer PCB design is possible, but for reliable HS (480 Mbit/s), the differential impedance should be controlled to 45Ω±10% or so (but check the USB specs, this is just off the top of my head).  Easy and cheap with current PCB prototype manufacturing for us hobbyists!  But again, the choice of the isolated DC-DC supply is a critical factor; the good regulated 5V @ 500mA ones cost about the same as the isolator chips themselves.

I have not used any of the FPGA ones, but their USB host/device implementations are often tested with only one type of device (usually USB Audio).  The el-cheapo eBay USB LS/HS isolators have served me very well, but I've been careful to check the DC-DC supply in each one.  I've been waiting for sensible ADuM3165/3166/4165/4165/ISOUSB211-based isolators to become available, but apparently have to build my own.  I'm only a hobbyist, and the post-regulation of the 5V isolated output is giving me pause.  Or go with TI Webench and a discrete isolated switcher to 6V with post-LDO to rock-solid 5V? 

[thm_w]

ADUM3165 $30: https://www.aliexpress.com/item/1005007044460567.html

https://www.aliexpress.com/item/1005006744005556.html - claims usb 3, seems unlikely
https://www.aliexpress.com/item/1005001933476089.html
https://www.aliexpress.com/item/1005005184945508.html

[madires]

Typical ethernet ports include galvanic isolation via transformers.  The product in the initial post is just an unneeded extra transformer.  You can build yourself one using a standard MagJack and a male RJ45 connector.  Makes absolutely no sense to me, though.

What is often overlooked is that screened ethernet cables plus RJ45 plugs with metal shield can create a ground loop, as the screen/shield is usually connected to the ground of a device. If you connect two PCs via an ethernet switch you get one ground connection via mains earth/PE (via the PSU), and a second one via the screened ethernet cables. However, you don't need to get a fancy LAN isolator. Simply use an ethernet cable without shielded plugs.

[SiliconWizard]

Typical ethernet ports include galvanic isolation via transformers.  The product in the initial post is just an unneeded extra transformer.  You can build yourself one using a standard MagJack and a male RJ45 connector.  Makes absolutely no sense to me, though.

What is often overlooked is that screened ethernet cables plus RJ45 plugs with metal shield can create a ground loop, as the screen/shield is usually connected to the ground of a device. If you connect two PCs via an ethernet switch you get one ground connection via mains earth/PE (via the PSU), and a second one via the screened ethernet cables. However, you don't need to get a fancy LAN isolator. Simply use an ethernet cable without shielded plugs.

Yes. Something often overlooked. Same potential issue with a USB isolator if the shield is connected on both sides, which is rarely not the case.

[Nominal Animal]

What is often overlooked is that screened ethernet cables plus RJ45 plugs with metal shield can create a ground loop, as the screen/shield is usually connected to the ground of a device.

Very true.  I do have S/FTP Cat6 cables myself –– extremely useful in EM-noisy environments, and when you have to pull an Ethernet cable next to mains wiring –– but as I rarely use them, I hadn't thought about that.

Same potential issue with a USB isolator if the shield is connected on both sides, which is rarely not the case.

In the USB isolators I have, the shield is connected to the local ground (I believe via a capacitor and a resistor in parallel in one.

Unlike Ethernet (especially MagJacks with the built-in transformers), where only the signals (pairs) are isolated, in the USB isolators using the dedicated chips there is no common ground plane at all: upstream and downstream have their separate isolated ground planes, bridged only by the isolator chip and the isolated DC-DC converter.  This means that the USB cable having shield connected to ground pin is perfectly okay.  (This is also clearly indicated in the datasheet Applications Information, so it is very unlikely that even the cheapest isolators get this wrong.)

A possible issue I can see, however, is if the isolator has a metal chassis, and the USB connectors' shells touch that chassis.  It would completely defeat the entire isolation.


There are USB 3.x SuperSpeed isolator chips, like Advanced Photonics' APISU30/31, APISLCU30/31, but I haven't yet had access to any (expensive!).  At least the APIS ones also need an USB 2.0 ADuM3165/3166/4165/4166/ISOUSB211 isolator for full USB support, too.

[MrMobodies]

Just noticed this:

Following the unfortunate theft of our Facebook account, we’ve been left with no option but to start a new one which you can visit by clicking here. Please follow and share to help us rebuild our community and, of course, be cautious of any messages unusual posts or requests coming from our old account.

Joke: I bet they clicked on one of those "Sponsored" posts on Facebook that redirected them where the destination URL is hidden and wrapped in the Facebook redirect URL, they thought it was a post it wasn't, they couldn't see the destination URL, and they possibly had a case session hijacking after clicking and being redirected to a hacked website impersonating like the BBC with one of those false news articles.

It remind me of a theme similar to that of a stock market scam in the name of MarketMaster and 60op back in 2019 with the a registration form "Change your life today!" and button under it "NEXT", like they are laughing at whoever clicks that button.

[mansaxel]

There are optical USB isolators, complete with multimode or single mode line interfaces. When you absolutely don't want any TEMPEST leakage from your USB smart card reader. I have seen and used these.

Anyway.  In an earlier reply I referred to the Muncy paper.  It explains why ground loops form, why they ruin your day, and how to defeat, not the loop, but the interference. 

In general, lifting and isolating grounds might achieve an audio frequency solution, but at the cost of a RF problem.  EMI/RFI emission and tolerance standards, if they are to be followed, and one should! in many situations rule out isolation in favour of a ground/screen bonding strategy.  Studying and understanding this can be very beneficial for the electronician.