Well, if you consider each transition producing an amount of electromagnetic noise regardless of the number of transitions per second, then it would make sense that there would me more noise density with 1gbps than 10/100mbps.
The fact that the circuits themselves don’t get affected by noise as much as 10/100 is different from the noise they can produce.
Nope. In SPDIF the clock must follow the source, and if there’s low frequency drift in the source clock them you can get audible jitter.
In fact, it’s both, more sensitive to noise, and the produces less noise. Go ahead and do some research on it.
Dude, the thread is about Optical SFPs. Not TOSLink. 
Not specifically to the optical portion of the SFP, to the electrical portions, i.e. the other end of the SFP and the rest of the Media Converter, as well as to prevent a cheap [PSU] from spewing noise back to the AC supply.
Not sure I follow. The ADOT has an external PSU doesn’t it? Or do you mean the regulators inside the Media Converter?
Thanks again.
Sorry, the ADOT does have external PSU but the innards (regulator board) seem pretty beefy compared with the 10GTek. I wondered whether this might be a factor in the sonic difference reported by some and the preference they’ve expressed for the ADOT. We agree it can’t be any difference in the digital domain, so it’s either (a) in the power supply element or (b) in the imagination. Try to resist! 
Yeah, Media Converter do vary quite dramatically on their internals. Here’s an example from when someone raised the Sonore Optical module (take a look at the picture, it’s the “audiophile tax” in action 
)
W.r.t the ADOT, I can’t recall exactly, but the last time I looked, it was just another Taiwanese or PRC OEM, there was nothing special about it.
By the way, how much regulation is designed onto the converter also depends on the sensitivity of the chips and other components within the unit, it’s not always about how much external noise is reduced or not.
Right I am miles away from the knowledge that is being shown here but despite that I have an optical break in my home network . This has been done purely to lose the noise that could be generated by my router from getting to the rest of my system . So for me it is a very simple set up TP-Link Router LPSU powered with ethernet connection wired to a small Optical converter unit that is LPSU powered in to a D-Link Unmanaged Switch which has SFP input . Rest of the network is ethernet wired to the switch . I had an old Cisco managed switch previously but I could never get the SFP input to work . Bought the D-Link off e-bay used for around £38 including postage and it worked straight after plug in . Two things it has achieved is that my network is now much more robust and the drop outs and issues I have had in the past have gone . Secondly on music through the network it is now really quiet and pure and things just sound more real . The difference was not as acute as when I added an external clock but that is another story . In my view using an optical break is a very good thing for music and does no harm to anything else on the network . I have noticed that 4K streaming via Amazon are now much clearer and the colours are more intense as well so perhaps this also helps in other areas .
Customized Grade B Duplex Typical 0.12dB IL OS2 Single Mode LC/SC/FC/MU 2.0mm Bend Insensitive Fiber Optic Patch Cable has arrived:
Torben
Below a comparison of:
Fosco: LC-LC Duplex 9/125µm Corning ClearCurve Single Mode Bend Insensitive Fiber Optic Patch Cable
and
FT: Customized Grade B Duplex Typical 0.12dB IL OS2 Single Mode LC/SC/FC/MU 2.0mm Bend Insensitive Fiber Optic Patch Cable
Torben
For what it is worth, I just installed an ‘on-the-cheap’ optical connection from my router to my Bartok with a couple of (very inexpensive) 10Gek media converters - they came with matching (removable) SFP’s and i bought 50ft run of a pretty turquoise fibe cable. ~$80 CND. I even splurged on an iFI 5v PSU. The setup works flawlessly for the cost of a few digital downloads. I can’t say it sounds any better than running Cat5e direct from my fibe modem (I know others have different experiences). Main advantage for me is that i can use the in-wall cat5e for the AppleTV without installing another switch behind the rack. So things are a bit neater now.
At risk of confusing by quoting a quote from another thread:
Did you pick the TFC1000MGA for any particular reason or could it just as well have been any other Trendnet SGP media converter? The visual contrast is indeed quite stark, but I wondered if there was anything specific which makes it stand out from the crowd.
I’ve had the TrendNet for a quite a few years (among others). IIRC, I originally picked for a specific reason, but I can’t quite recall for sure, but I think it was for the number of discrete components it had.
The innards of the TFC1000MGA look truly impressive compared to the very inexpensive 10Gtek I purchased. But to put things in perspective the TFC1000MGA is spec’d to work over a distance 120km vs the 100m of the Gtek!
For everybody interested in the TRENDnet TFC-1000MGA: there is a new version, the old one is discontinued:
DISCONTINUED PRODUCT
The TRENDnet TFC-1000MGA (Version v1.0R) has been discontinued. It has been replaced by the TFC-1000MGA (Version v2.0R) .
As is par for such things, I would expect v2.0R to also be significantly cost reduced in terms of internal components such that Anupc might not select it over alternatives now.
Maybe you are wrong again 
TFC-1000MGA (Version v1.0R)
Mini-GBIC slot accommodates Multi (SX) or Single-Mode (LX) fiber modules
100/1000Base-T Auto-Sensing / Negotiation Ethernet port
100Base-FX / 1000Base-SX / LX Mini-GBIC slot
Functions as a standalone converter or with the TFC-1600 fiber chassis
Lifetime Warranty*
TFC-1000MGA (Version v2.0R)
1 x 100/1000BASE-T RJ-45 port
1 x 100/1000BASE-SX/LX SFP slot
LLCF support (Link Loss Carry Forward)
SFP slot accommodates multi-mode or single-mode modules
Functions as a standalone converter or with the TFC-1600 fiber chassis
Auto-MDIX 1000BASE-T RJ-45 port supports full-duplex mode
NDAA / TAA compliant (U.S. and Canada only)
Lifetime Warranty*
The spec differences don’t tell you if they redesigned things so all the functions are on a single chip or just a few chips now.
But you assume:
I would expect v2.0R to also be significantly cost reduced in terms of internal components
On what base is your assumption?
Almost every time a company releases a “v2” of the same product, they cost reduce it and consolidate the component count.
It’s just standard industry practice.
If they wanted to highlight the new features, they’d revise the model number as well.
As an example, look at all the different revisions of hardware Linksys sold as the “WRT54G” router:
You’ll see the amount of flash memory and RAM actually went down over time as they continued to cost reduce it and switched from general-purpose CPUs to a SoC.
I see, your assumption is just sarcasm. And you don’t know actually what happened, in this case.
Bartok v2, Rossini v2, Vivaldi v2… all bad. Apex the same.
Sorry Bill .