Eliminates the electrical and common-mode noise in the Ethernet by galvanically isolating the music streamer from the noisy network. Works without power supplies and can be placed between the music streamer and the switch, and/or between the WiFi router and the switch, bringing the CAS sound quality to the next level.
Traditional isolation utilizes a pair of electrical-optical converters which convert electrical signals to optical signals and then back to electrical signals. These conversions aim to isolate the electrical noise and common-mode noise (CM noise) existing in the Ethernet connections. However, this solution is complicated, expensive (better to have LPS) and occupies space, the effect is not significant. Indeed, this solution can isolate the majority of noise from the rest of the Ethernet but during the last stage of optical-electrical conversion, it introduces noises to the streamer again, and it is also susceptible to CM noise as well.
This Ethernet Isolator refers to the working principle of balanced isolation transformers, it galvanically isolates the streamer from other equipment in the network. It does not only eliminate the CM noise, but also creates an isolated and clean Ethernet for the streamer to work on, and therefore delivers better sound quality.
Each Ethernet Isolator possesses 2 individual channels, which can be used concurrently (between the WiFi router and the switch, and between the switch and the streamer), or connected in series (only applicable between the switch and the streamer). Connecting the channels in series yields a better isolation effect.
The most obvious effect after using the Ethernet Isolator is darker (quieter) background, reduced harshness and better bass response. On the other hand, the sound become more musical and analogue, and listener will feel less fatigue.
*The isolation transformers inside the chip need to take around 10 days to break in
Dimensions (mm): 79(L)x70(W)x24(H)
Data Rate: 1000Mbps x1, 100Mbps x1
Traditional isolation utilizes a pair of electrical-optical converters which convert electrical signals to optical signals and then back to electrical signals. These conversions aim to isolate the electrical noise and common-mode noise (CM noise) existing in the Ethernet connections. However, this solution is complicated, expensive (better to have LPS) and occupies space, the effect is not significant. Indeed, this solution can isolate the majority of noise from the rest of the Ethernet but during the last stage of optical-electrical conversion, it introduces noises to the streamer again, and it is also susceptible to CM noise as well.
This Ethernet Isolator refers to the working principle of balanced isolation transformers, it galvanically isolates the streamer from other equipment in the network. It does not only eliminate the CM noise, but also creates an isolated and clean Ethernet for the streamer to work on, and therefore delivers better sound quality.
Each Ethernet Isolator possesses 2 individual channels, which can be used concurrently (between the WiFi router and the switch, and between the switch and the streamer), or connected in series (only applicable between the switch and the streamer). Connecting the channels in series yields a better isolation effect.
The most obvious effect after using the Ethernet Isolator is darker (quieter) background, reduced harshness and better bass response. On the other hand, the sound become more musical and analogue, and listener will feel less fatigue.
*The isolation transformers inside the chip need to take around 10 days to break in
Dimensions (mm): 79(L)x70(W)x24(H)
Data Rate: 1000Mbps x1, 100Mbps x1
Last edited:
FAQ
1. IEEE 802.3 already specifies the isolation requirement is 1500Vrms at 50Hz to 60Hz for 60s, why is there a need for further isolation?
a. The standard only specifies how many volts that the ports / connections can withstand, the purpose of voltage isolation is solely a safety consideration which protects people from electrical shocks, but it does not specify how much noise should be isolated.
b. The below table shows CMRR and crosstalk of a commonly used chip in switches, there is still a room for isolation improvement. Adding an Ethernet Isolator can significantly improve the rejection ratio.
c. EMI / RIF noises are not necessarily generated / picked up at the switch, they can be picked up by the long Ethernet cables which act as antennas. Longer and looped Ethernet cables are more susceptible to noise pick up. Isolating the noise at the last mile (just before the streamer) is more effective.
2. Why are there two channels in the Ethernet Isolator?
a. Surprisingly a slower Ethernet connection sounds slightly better than a faster connection (this is also generally true for other audio equipment), with two channels you can choose which channel to go with. Alternatively, you can place the 100Mbps isolation between the switch and the streamer, and concurrently place the 1000Mbps isolation between the WiFi router and the switch to improve the overall Ethernet noise level.
3. What Ethernet cables are recommended?
a. Cat 8 shielded cables (STP) are recommended. Don’t worry about the ground loop problem as one side of the ports are made with insulated plastic RJ45 ports. However, for the connection to the streamer, a very short STP cable (e.g. 0.5m) is recommended.
1. IEEE 802.3 already specifies the isolation requirement is 1500Vrms at 50Hz to 60Hz for 60s, why is there a need for further isolation?
a. The standard only specifies how many volts that the ports / connections can withstand, the purpose of voltage isolation is solely a safety consideration which protects people from electrical shocks, but it does not specify how much noise should be isolated.
b. The below table shows CMRR and crosstalk of a commonly used chip in switches, there is still a room for isolation improvement. Adding an Ethernet Isolator can significantly improve the rejection ratio.
c. EMI / RIF noises are not necessarily generated / picked up at the switch, they can be picked up by the long Ethernet cables which act as antennas. Longer and looped Ethernet cables are more susceptible to noise pick up. Isolating the noise at the last mile (just before the streamer) is more effective.
2. Why are there two channels in the Ethernet Isolator?
a. Surprisingly a slower Ethernet connection sounds slightly better than a faster connection (this is also generally true for other audio equipment), with two channels you can choose which channel to go with. Alternatively, you can place the 100Mbps isolation between the switch and the streamer, and concurrently place the 1000Mbps isolation between the WiFi router and the switch to improve the overall Ethernet noise level.
3. What Ethernet cables are recommended?
a. Cat 8 shielded cables (STP) are recommended. Don’t worry about the ground loop problem as one side of the ports are made with insulated plastic RJ45 ports. However, for the connection to the streamer, a very short STP cable (e.g. 0.5m) is recommended.
Attachments
A
This will defeat the shielding of the cable as shields need both ends connected to function. With one end connected the shield turns into a 1/4 wave antenna and voids your CE approvals, both radiating in all directions and coupling in your cellphone signal. This device cannot be CE compliant and is therefore likely to be technically illegal to sell.
Actually there are unshielded Ethernet cable available in the market, and the plugs are made of plastics...
You're right. But as explained, the standard does not specify how much noise to be isolated. The effect of adding an Ethernet isolator is quite significant. I have been testing the Ethernet isolation for almost two years before deciding to make it available to the public. I have distributed a few prototypes to friends, and sold a few to local buyers, their feedback conincides with my findings.
That type is fine for it's use up to 1 Gbps. The problems start when you single end earth a shielded cable, which turns the shield into an efficient antenna.
Long ago I worked on approval of a shielded cable LAN, where one end was grounded and one end connected to ground via a high voltage capacitor. This allowed isolation at 50/60 Hz while keeping the shield working at radio frequencies. Not as good as proper grounding at both ends but enough to pass.
What a load of piffle. The data is either received or lost in which case there'd be a drop-out. IMO your product is bogus.
If there is packet loss from noise, the packets are resent and recovered by standard network protocols. If interference is so bad that packets cannot be recovered you will have drop-outs. Use Wireshark to see it. "What's Wireshark?" he asks.
The noise is not added to the data stream. Simple as that.
Please people do not buy into this nonsense. Do an experiment: Without anything playing but everything turned on, unplug the RJ45 connector. Hear any noise difference?
I run over 50m of unshielded Cat5e and also use Wi-Fi to stream audio around my property. There is no noise attributable to the network and IMO this product can do NOTHING to achieve any measurable difference in sound quality.
The noise is not added to the data stream. Simple as that.
Please people do not buy into this nonsense. Do an experiment: Without anything playing but everything turned on, unplug the RJ45 connector. Hear any noise difference?
I run over 50m of unshielded Cat5e and also use Wi-Fi to stream audio around my property. There is no noise attributable to the network and IMO this product can do NOTHING to achieve any measurable difference in sound quality.
Again, not talking about the noise effect on "data", but the sound quality. Please note there is nothing to do with "data". Neither there is data gain or loss, it is the high frequency noise the exists on the Ethernet which couples into the audio path.
Regarding the "noise", it is not the zzz, crackling or hum noise, it is the high frequency harness. Not trying to convince you but it does make a difference with the isolator plugged in between the switch and the streamer.
I fully understand that it's hard to believe in Ethernet isolation, I don't mind if you think this is a nonsense or bogus, it's up to you. But please bear in mind that I'm not talking about IP packets or data, this gadget has NOTHING to do with data integrity, but high frequency attenuation only.
I'm not replying to this thread anymore, thanks everyone for your input and comments!
P.S. A slow blow fuse and fast blow fuse has a different sonic characteristics on power amps (but less obvious on low power devices such as CD players), you can test it with the same power ratings.
I'm not replying to this thread anymore, thanks everyone for your input and comments!
P.S. A slow blow fuse and fast blow fuse has a different sonic characteristics on power amps (but less obvious on low power devices such as CD players), you can test it with the same power ratings.
Instead of contradicting your own claims made both here and on your YouTube, you could rise to the challenge put to you by TNT. Instead you make another absurd claim and run away. LOL!!!
Last edited:
Yes, show your measurements instead of your "hope fact"....
a) measurement in receiving unit without isolator
b) measurement in receiving unit with isolator
Present wideband analysis please, i.e. both audio range as well as HF - complement with distortion +N and multitone analysis.
If you don't have these at all, my take is that you base your claims on hope and/or guesses.
//