JDS Synapse & Neutron V1 USB Isolators Review

[Jarrett]

Absolutely worth it these devices. The HS01 solved about 75% of the problem I was having with a computer in the chain. Later, upgrading from the Topping E30 to the E30 II, inadvertently bumped that up to 95%. Can't explain why. Anyway, these devices will get you close to a balanced system if it's too pricey to move on to one.

 

[johny_2000]

Given that these isolation devices are fairly small and require only a few components, why don't DACs include an isolation circuit on their USB input.

It might not work for a dongle, but their should be enough space on a desktop DACs PCB to include it.

Mass-produced High-Speed (480 Mbps) integrated USB isolation chips have only become available in the last couple of years:
Texas Instruments ISOUSB211, November 2021, $8~$13/pc
Analog Device ADUM4165 (ADUM4166), March 2022, $9~$17/pc
Analog Device ADUM3165 (ADUM3166), May 2022, $6~$12/pc

I believe that many of the current devices on the market were developed in earlier years than these dates, so the electronic components were not available at that time.
We may see changes in future devices developed, especially those aimed at the highest performance (and cost) and audio analyzers.

The previous generation of USB isolators were based on discrete components using FPGAs with custom firmware, optocouplers, drivers, etc. and were much more expensive (~$250 and bulky) to manufacture, let alone integrate into a DAC.

 

[pablolie]

Mass-produced High-Speed (480 Mbps) integrated USB isolation chips have only become available in the last couple of years:
Texas Instruments ISOUSB211, November 2021, $8~$13/pc
Analog Device ADUM4165 (ADUM4166), March 2022, $9~$17/pc
Analog Device ADUM3165 (ADUM3166), May 2022, $6~$12/pc

I believe that many of the current devices on the market were developed in earlier years than these dates, so the electronic components were not available at that time.
We may see changes in future devices developed, especially those aimed at the highest performance (and cost) and audio analyzers.

The previous generation of USB isolators were based on discrete components using FPGAs with custom firmware, optocouplers, drivers, etc. and were much more expensive (~$250 and bulky) to manufacture, let alone integrate into a DAC.
View attachment 412724

I also may suspect that, with the growth of USB bandwidth and it's ability to supper hub on hub device multiplication, the chances of having crappy devices in there grows. I make a point of never ever getting the cheapest device added to my USB hub, and always stay aware of why stuff is there.

 

[johny_2000]

Actually, when I first started, it was much wider gap. But as soon as I grounded the device and then ungrounded it, it settled to lower value you see in the review. Leakage due to capacitance can be variable that way with the cap getting charged up/discharged.

Correct. Even these USB isolators also have some capacitive coupling. In general, many of them use cheap built-in DC/DC converters that have ~1000 pF of capacitive coupling from input to output. On the other hand, the Intona USB isolator has ~7 pF of capacitive coupling. This can be significant in measurement applications.

@MC_RME mentioned this on the RME community forum:
The effective amount of reducing leakage current and ground loop effects is given by the coupling capacity between USB 'input' and 'output', in pF (picofarad). While the chips itself stay far below 10 pF, the layout of the PCB, but even more the typically added power module (to transport input power to the output galvanically isolated) can add significant capacity. For example, the Topping HS02 has 1 nF (1000 pF) coupling capacity due to the included Mornsun DC/DC converter, making it not suitable for laboratory measurement tasks. Intona does the same, but use their own, discrete solution. The result is that the (much more expensive) Intona reaches below 10 pF (100 x less) - that's a different class, obviously.

It is difficult to say what kind of DC/DC power converter is used in the JDS Synapse, as it seems to be made of discrete components and an isolation transformer (similar to Intona's approach). Therefore, the coupling capacity is unknown. But it is definitely based on TI's ISOUSB211 chip.

This could be just a textbook (datasheet) implementation:

 

[TRDK]

Have been using the Neutron V1 USB Isolator for some time now and must say it works just fine.

Here the setup:

 

[192kbps]

HS02 1 nF
DSD TECH SH-G01L 37 pF
Neutron HiFi Isolator V1 22-24pF
Hifime High-Speed USB Isolator v2 71 pF

from MC_RME
@johny_2000

 

[dmitrykos]

Neutron HiFi Isolator V1 22-24pF

20-22pF

 

[Ajax]

Amir was able to respond in a professional way, as I would expect of him. He didn't take offence to my mild ribbing.
Perhaps you should learn from him.

It certainly appears he took no offence, and if as you say it was simply "mild ribbing, then I apologise and stand corrected, although it did not read that way to me.

Anyway, carry on Jeeves.

 

[Antagron]

I had a basic amp (Pioneer A400) that once picked up transmissions from a fire engine across the street. Always found that very strange.

I once lived near a tall AM radio antenna outcropping. I could hear the radio playing on the house’s pipes or wiring from within the walls. I’m glad I moved. That was a lot of electromagnetic radiation!

 

[AnalogSteph]

I get, what sounds like GPU whine, over HDMI. Any suggestions on available products or otherwise to tackle this?

The HDMI connection may facilitate the ground loop currents, but your actual problem is going to be downstream. HDMI galvanic isolation is $$$ to $$$$, so I'd prefer to tackle the problem someplace else. We'd have to know a lot more about the setup to give proper advice.

 

[Lambda]

Thanks Amir!

Is it a myth or no that ferrite beads can isolate a USB cable for 1$ ?

Not a myth. i tested this.
Under some conditions they work well into the audion band.
( if you use a lot) But its still way cheaper then a proper isolator.

https://audiosciencereview.com/foru...-measurement-and-comparison.27937/post-970936

 

[pkane]

Not a myth. i tested this.
Under some conditions they work well into the audion band.
( if you use a lot) But its still way cheaper then a proper isolator.

https://audiosciencereview.com/foru...-measurement-and-comparison.27937/post-970936

Ferrite beads offer no isolation, let's be clear about that. A proper isolator (as cheap as $30) offers a proper galvanic isolation and device protection.

 

[Esprit]

Hifime High-Speed USB Isolator v2 71 pF

Is this the most effective?

Hifime Hi-Speed USB Isolator - new 2023 version

The Hifime High-Speed USB Isolator is designed specifically for audio and supports USB Hi-speed mode that is required by many Audio devices. It will prevent ground loops, prevent interference and protect your equipment against power surges.

hifimediy.com

 

[dmitrykos]

Is this the most effective?

Hifime Hi-Speed USB Isolator - new 2023 version

The Hifime High-Speed USB Isolator is designed specifically for audio and supports USB Hi-speed mode that is required by many Audio devices. It will prevent ground loops, prevent interference and protect your equipment against power surges.

hifimediy.com

The lower the pF value (coupling/parasitic capacitance) the better. Here is the same list from @192kbps sorted from best to worst by this parameter:

Neutron HiFi Isolator V1 20-22pF
DSD TECH SH-G01L 37 pF
Hifime High-Speed USB Isolator v2 71 pF
HS02 1 nF (1000 pF)

 

[davmol]

I had a lot of audible noise introduced to my Topping D50III DAC over USB, which was quite disappointing. Using a separate usb power cable and data line helped reduce it a bit but did not eliminate it. Interestingly, the noise only came when using work my Dell laptop and there was no noise introduced from my personal Lenovo laptop, so it's with trying other computers before buying one of these, if that's a resolution available to you.

I'm now solely using my Lenovo so no more USB noise issues for me, but it's good to know there is a solution out there should the issue appear again in future.

 

[CedarX]

I see the confusion. There are two classes of isolators/reclockers, filters that are being sold to audiophiles:

Class 1: they claim that even if you are not hearing any noise from your system now, that these products improve its sonic fidelity by widening soundstage, blacker blacks, more "analog sound," sound that is more like live music, etc. Nothing in this review validates such claims. And previous tests have shown the same in taking a perfectly working system and showing that such devices do not change their output.

Class 2: these are products that are aimed at people who have real problems. They have ground loops or "hear" their computer working as the noise profile changes with system activity. This is a serious problem that needs resolution. Many people report them and we scramble to make suggestions to solve these problems. One solution as stated in the review is to isolate the PC using optical Toslink. The other, is using these USB isolators.

So all depends on company claim and what customers buy them for. That defines how I test and recommend it.

I think this is an important post (thanks Amir! )

First of all, the exact same device can be Class 1 or Class 2, depending on what the OEM/Distributor claims about their product. Class 1/2 has probably more to do with the company--what they claim--than the product... but Amir reviews product not companies, so I think a Class 1 / Class 2 product assessment is the right approach.

Second, I don't think Class 1 / Class 2 is always a clear cut... A product may have features (claims...) that fall in either categories, for example:

• This claims from JDSLabs about Synapse: "Its ISOUSB211 enhances performance by cancelling USB signal loss and optimizing the USB timing diagram, while supplying super clean, transformer isolated power to your DAC, mixer, interface, or other UAC2 device". Is this a Class 2-type claim or Class 1? What does this "optimizing the USB timing diagram" mean (honest question... I don't know !!!)? It doesn't mention any particular problem about USB timing diagram that Synapse will solve. Class 1?
• Neutron statement about error correction: "Isolator V1 does not propagate errors in USB signal from the Host side. Therefore your USB Client is always protected from any unexpected misbehavior caused by erroneous USB signal." So, if I send a DSD stream to a DAC not supporting DSD, Isolator V1 will take care of it? Class 1?
• Neutron claim about the noise barrier "full isolation of the power and USB data signals" appears reasonable. Except when you use the auxiliary power input (micro-USB): my understanding is that V1 does nothing to isolate that power source... Deceiving? Class 1 or Class 2?

I believe that, for these products, voting 'Great' should be for an absolute Class 2 product, voting 'Poor' should be for a broken product and/or Class 1 product... and everything in between. I voted 'Fine'... But a few months ago, Neutron previous description of V1 "a must-have for audiophiles" (was changed) would have yield a 'Poor' vote from me.

 

[dmitrykos]

• Neutron statement about error correction: "Isolator V1 does not propagate errors in USB signal from the Host side. Therefore your USB Client is always protected from any unexpected misbehavior caused by erroneous USB signal." So, if I send a DSD stream to a DAC not supporting DSD, Isolator V1 will take care of it? Class 1?

Isolator's IC is replaying/reconstructing USB signal, so on Isolator's output you get the same USB data but re-clocked and formed by the IC. What you are saying is not correct, and it was never stated that Isolator V1 can tackle non-USB signal (it is an USB device and thus it is assumed implicitly that you connect it to USB port, not to something else). Instead of erroneous though description should use weak signal, it is what was meant initially.

Neutron claim about the noise barrier "full isolation of the power and USB data signals" appears reasonable. Except when you use the auxiliary power input (micro-USB): my understanding is that V1 does nothing to isolate that power source... Deceiving? Class 1 or Class 2?

External power is provided to isolated side of Isolator, directly to its USB output. In this case it is an option for a user to tackle power hungry devices (which consume >400 mA) and have isolated USB port. External power must be isolated, it is mentioned in the User Manual as well as usage scenarios with and without external power. By Full means that Isolator is isolating Data and Power of the incoming USB signal. There are Data-only isolators in the market, so beware - they will not isolate the connected device from the ground loop if connected device is using power via VBUS pin of such device (so they do not isolate USB signal Fully).

I think you've gone too deep with your nitpicking with these statements, and I hope I managed to clear the confusion

 

[dmitrykos]

• This claims from JDSLabs about Synapse: "Its ISOUSB211 enhances performance by cancelling USB signal loss and optimizing the USB timing diagram, while supplying super clean, transformer isolated power to your DAC, mixer, interface, or other UAC2 device". Is this a Class 2-type claim or Class 1? What does this "optimizing the USB timing diagram" mean (honest question... I don't know !!!)? It doesn't mention any particular problem about USB timing diagram that Synapse will solve. Class 1?

Nothing is wrong with this statement either. This diagram can be found in the specification of ADUM4165 (ADUM4166) IC for example, see Page 15. There is Eye/Time Diagram of reconstructed USB signal. It is similar to Isolator V1's statement about erroneous/weak signal reconstruction. Therefore this IC solves the problem of the weak signal. Which can happen for example due to a long USB cable, or cable with rusty connectors causing additional noise or attenuation which is corrupting Eye Diagram, or user put ferrite beads on non shielded cable and so on. On output of isolator device you get a perfectly looking USB signal - job done.

 

[Esprit]

Neutron HiFi Isolator V1 20-22pF
DSD TECH SH-G01L 37 pF
Hifime High-Speed USB Isolator v2 71 pF
HS02 1 nF (1000 pF)

Where does JDS Synapse fit on this list?

 

[dmitrykos]

Where does JDS Synapse fit on this list?

Has to be measured and specified by the producer in the product details, so currently unknown.