@Anupc - I am not sure we are talking about the same filter. I am not sure what you mean by filter?
I am talking about broadband line filter.
Torben
Care to explain what you believe a “Broadband line filer” is?
For clarification, I’m currently using a Ubiquiti Unifi switch (the main switch for my study) into a Network Acoustics switch running over 10m of Single Mode fibre (both switches have SFP ports).
1 Like
Dusting off a few relevant posts as my Trendnet TFC-1000MGA duo arrived today from the US. I’m going to send them straight back as the mains plugs have a pin missing 
@BillK was spot on. Nowhere near as impressive a build of discrete components as @Anupc 's v1.0.
1 Like
I only see 1 major change, the removal of a choke - which arguably was for chassis hot-insertion. And a couple of minor changes, like the through-hole Capacitor replace by an MLCC, and through-hole Crystals replaced by Surface-mount ones.
The most important aspect, the discrete Ethernet Port magnetics remains, although they seemed to have changed supplier for that part, but the specifications look identical.
V2 looks to be more about manufacturing modernisation/efficiency rather than cost-cutting.
3 Likes
Moving to more surface mount components and more manufacturing efficiency is cost cutting.
The Realtek RTL8367SB-CG used for the V2 module is also a bit cheaper than the RTL8213M used on the V1 module.
Cost cutting does not mean poorer performance per se, but rather the cost savings must have been great enough for them to bother to spin the board as none of the obvious components on the V1 board come up as discontinued, the other reason boards usually get respun.
2 Likes
Thanks both; I think you’re in violent agreement! Cost-cutting and efficiency are basically the same thing viewed through different lenses; you could argue they may have different primary drivers if you wanted to do the hair-splitting thing but I don’t think there’s much to be gained.
@Anupc obviously knows enough about circuit boards to recognise a choke and @BillK obviously knows enough about circuit board production to use terms like spinning. Got to admire an expert. 
Suitably reassured here. Listening to follow.
1 Like
I just discovered this product: TeraDak T-S211 (SC)
https://kamaudio.com/TeraDak-T-S211-media-converter
With a TCXO or OCXO clock. The OCXO version is 100$ on top. Not sure what the benefit would be.
Torben
As discussed elsewhere, yet another product which helps to reinforce the myth that ethernet clock accuracy can have any impact on the quality of digital data which emerges.
Having said that, it doesn’t seem to have attracted the audiophile tax so might be bargaintastic…
Will take a closer look. Thanks
Testing USB 2.0 to barrel connector type N cable - USB A DC 5V (OD: 5.5mm, ID: 2.5mm) barrel power connector + Anker Powerbank 20,000mAh with my FMC from Startech.
Works just fine So the FMC can be powered by a power bank
Torben
@torbenrick I have wondered about this solution to provide true isolation from the optical break by removing the ground plane route for noise to travel. I have so far not been enamoured by using FMC and in many cases found the sound to be worse than using copper ethernet but it would be interesting to hear what you found by using battery to power the FMC.
@torbenrick I’m interested in what attracted you to the idea of a battery to power your FMC(s). Is it the ground plane thing which @NickBacon mentions?
I have read a few articles suggesting that supercapacitors have offer many of the benefits of battery power supply but are better at providing fast bursts of current when required. FMC’s of course are not exactly current-hungry so maybe this generality doesn’t apply. Still, I’d be interested to hear more about what inspired you to explore this idea.
Best wishes
@TheFlash + @NickBacon - Complete galvanic isolation is the optimum when it comes to “Ethernet sound”.
That only works if you really operate the FMC from a battery, rechargeable battery or power bank. Perfect decoupling of signal and power supply. BUT: If there is no interference, there is nothing to eliminate and therefore no changes in the sound.
That is the reason for me testing it.
My Startech FMC has only a consumption of 5VDC/0.4A. My Anker Power bank 325 has: Two standard USB A Output: 5V = 2.4A
From the FMC I than use 0.5 Meter Prysmian Draka UC900 Super Screen 27 CAT.6A S-FTP/PIMF (cable type CAT.7) - RJ45 Hirose TM31 Cat.6A connector to the DAC, All other cables in my setup are UTP.
Torben
1 Like
Thanks Torben,
Sounds interesting. I know you’ll be using your ears for the test; do you have any way of also measuring the difference? My ears are all I have!
Good luck and report back 
Have been doing some comparison/test with the below setups:
I could not ear any difference in the to setups - same SQ.
If you haven’t done any noise reduction, the Power Bank probably would be a good an cheap solution.
Torben
Thanks Torben.
Did you have the Powerbank disconnected from the mains when you did your listening tests?
How long can you use it for when disconnected before it needs to be charged again? I guess this is what would put some people off, even if a cheaper solution: the need to connect, charge, disconnect, use/discharge… and repeat.
Thanks
Nigel
Yes - I don’t have it connected to mains when I am using it.
The power bank that I have (20.000 mAh) need to be recharged after roughly 50 hours. I have 2 of them - one “playing” the other recharging.
Torben
PS:
Any suggestion for a GOOD USB Power Cable - Type N Barrel (OD: 5.5mm, ID:2.5mm) that can be bought in Europe.
So far I have only found this from China:
50 hours is a fair outing, and having two of them also guarantees you won’t end up being musicless while waiting for a sole Powerbank to charge. Good plan.
Do you have a rough calculation of how much your two power setups cost? I suspect that if the sonic performance is the same, many will prefer the lack of hassle with your SMPS setup - unless the 2 x Powerbank option is significantly cheaper.
Re USB - DC barrel cable, if your present cable is both foil-shielded and braided, it’s difficult to imagine improving on it. Sure, there are more expensive USB cables out there but you’re only transporting power on yours, not data, so I’d stick.
The power bank solution with 2 power banks and cable is roughly 90 EURO and the SMPS solution is roughly 480 EURO. This is without FMC.
To secure complete galvanic isolation from the mains I will keep the power bank solution.
Torben
1 Like
@torbenrick That back to back comparison is interesting and thank you for spending the time doing it. Can I just ask for additional validation of the result, have you had the FMC powered any other ways where you did hear differences? Also, have you tried taking out any of the components in your mains powered solution? Eg what happens if you take out either the DC ferrites or the mains ferrites? Are they both essential? Indeed are all the components essential on the power side of that example for it to be as good as the battery?
Also have you tried a copper wire ethernet connection without the FMC (I ask that because I come from a background of so far only hearing inferior sound when adding in a FMC and so far I prefer copper ethernet. I have tried the FMC connected direct to the final device and I have also tried adding a FMC just before an Innuos PhoenixNET switch. I have also tried connecting a Dela S100 switch by optical to a Cisco 2960 and much preferred copper ethernet between the S100 and the Cisco. I guess I just want to make sure that there isnt anything else that might be making it so that you are not hearing any difference between your two power solutions with the StarTech FMC.
1 Like