USB Power isolation

[beerandmusic]

Would this device from china work to isolate the 5V noise from usb bus?

Could an engineer from this site say if there anything in the description that would help determine if it would work for DAC purposes or not? (especially about the chip ADUM4160 or the speed rate?)

 

Low Noise USB Isolator Short Circuit Protection ADUM4160 HiFi JTAG Isolation | eBay

 

[h=2]Description[/h] This module is designed based on ADUM4160, a Full/Low speed 5kV USB isolator. It's a good choice for appilcations that isolation is needed between PC and peripheral device. This module features a low noise power that is very suitble for Hifi devices. In additon, overcurrent protection circuit is included. When short circuit happens in your device, the power supply will be cut off to protect both your device and this module. A red LED will light when overcurrent condition occcurs.

 

Features :

 

 

• USB 2.0 compatible
  
• Low and full speed data rate: 1.5MBps and 12Mbps, selected by jumper
  
• Bidirectional communication
  
• On-board LDO regulator for low noise applications
  
• Wide power input range: +6V to +24V
  
• Downstream port overcurrent protection( This function is very useful for protecting your device from damage when short circuit happens)
  

[JohnSwenson]

There are many such boards on the market, the critical part is that it is only good for full speed mode (the fastest speed of USB1.1), in order to use it you will need a DAC which runs at full speed, many modern DACs just run at high speed which means they will not work with this board.

 

96/24 is the fastest speed you can get with full speed so if a DAC maxes out at 96/24 it might well be a full speed DAC. To really perform well this board is going to need a very good low impedance power supply.

 

Such a board as this can help in some circumstances, it can also make things worse in others. The isolator chip is directly driving the output which means it will actually have higher jitter than what is coming in, but might wind up with better signal quality. It's a toss up.

 

John S.

[beerandmusic]

There are many such boards on the market, the critical part is that it is only good for full speed mode (the fastest speed of USB1.1), in order to use it you will need a DAC which runs at full speed, many modern DACs just run at high speed which means they will not work with this board.

 

96/24 is the fastest speed you can get with full speed so if a DAC maxes out at 96/24 it might well be a full speed DAC. To really perform well this board is going to need a very good low impedance power supply.

 

Such a board as this can help in some circumstances, it can also make things worse in others. The isolator chip is directly driving the output which means it will actually have higher jitter than what is coming in, but might wind up with better signal quality. It's a toss up.

 

John S.

 

thanks for the info....so we should look for a "high speed 480mbit" usb isolator when shopping? maybe dlna or optical is the way to go

[Pepsican]

thanks for the info....so we should look for a "high speed 480mbit" usb isolator when shopping? maybe dlna or optical is the way to go

 

Have you looked at the Regen from Uptone Audio? It does the job plus a few more things.

[beerandmusic]

Have you looked at the Regen from Uptone Audio? It does the job plus a few more things.

 

Yes, that is what got me looking...but it's not available until july, and i was just curious if there was something less priced out there. I understand it does 2 things, but the major thing for sound improvement is removing the noise due to the 5v bus, and just want to find/try something less expensive. If i don't find anything when a regen is available i may try it, but it wouldn't hurt to see if i can hear any improvement compared to just the noise eliminator. I found schiit wyrd, but its $99...hoping for something half that price.

[JohnSwenson]

thanks for the info....so we should look for a "high speed 480mbit" usb isolator when shopping? maybe dlna or optical is the way to go

 

The only USB high speed isolators that exist are very expensive (thousands of dollars) not 50 dollars.

 

John S.

[JohnSwenson]

Yes, that is what got me looking...but it's not available until july, and i was just curious if there was something less priced out there. I understand it does 2 things, but the major thing for sound improvement is removing the noise due to the 5v bus, and just want to find/try something less expensive. If i don't find anything when a regen is available i may try it, but it wouldn't hurt to see if i can hear any improvement compared to just the noise eliminator. I found schiit wyrd, but its $99...hoping for something half that price.

 

The REGEN's primary purpose is to regenerate the USB signal as close a possible to the DAC. Why this is important is that I found that the signal integrity is important to the USB receiver in the DAC, the lower the SI the more noise the receiver generates. Getting the cleanest signal you can get into the DAC cuts down on the noise in the DAC, not by blocking the noise from the computer, but by cutting down on the noise the USB receiver generates inside the DAC itself.

 

The clean power to the 5V is a side benefit for those DACs that use it. The primary purpose is feeding a very clean USB signal into the DAC.

 

The Uptone site has some technical articles that go into all the gory details if you are interested.

 

John S.

[seeteeyou]

USB 2.0 Hi-Speed Isolator withstands up to 2kV_RMS

 

http://intona.eu/en/products

 

Mass production has already started. The product is expected to be shipped at the end of September 2015. Target net price is 209 €.

[Pepsican]

USB 2.0 Hi-Speed Isolator withstands up to 2kV_RMS

 

http://intona.eu/en/products

 

Mass production has already started. The product is expected to be shipped at the end of September 2015. Target net price is 209 €.

 

First, the Regen including world wide shipping is about €15,- cheaper. Secondly, seems this thing does Galvanic isolation. Many DACs have that build in already. And it also does not resolve the issues around bad power and jitter generated in the DAC as crap can still leak into the DAC via the signal path.

[beerandmusic]

First, the Regen including world wide shipping is about €15,- cheaper. Secondly, seems this thing does Galvanic isolation. Many DACs have that build in already. And it also does not resolve the issues around bad power and jitter generated in the DAC as crap can still leak into the DAC via the signal path.

 

I just read these responses, thanks john and pesican, and i will read details on uptone, but wondering if it can be put in "layman's terms" without things like galvanic isolation, and john's response "REGEN's primary purpose is to regenerate the USB signal as close a possible".

 

i can kind of picture the digital signal, 5v, and noise as three different signals....when john says to regenerate the digital signal, i see that pretty much the same as removing the noise and the 5v (and my understanding is the noise is a result of the 5v signal, so if you remove the 5v, you should be essentially be removign the noise too?, leaving the digital signal, and then adding a cleaner 5v later (if needed). So i am kind of confused by "regenerating". isn't that the same as removing the 5v and noise and making the digital signal available to the dac? or does it actually reproduce the digital signal? It would seem like we would want the least "reproduction" as possible to eliminate errors? Am i totally in left field or what? I would hope it could be explained better in "layman's" terms.

[beerandmusic]

ok, i just read on uptone and it actually does create a new signal via a "hub chip"...

i am "guessing" the 5v and noise is isolated first, and then the digital signal is replicated via the "hub chip", and then it is further processed from there? And the purpose of replicating the signal after the 5v and noise is isolated, is because this hub chip has better characteristics....if i understand correctly, then that seems to make sense.

[Distinctive]

You may also consider to go down the supercapacitor route.

I have this one on order for my Hydra-Z: Hydra ZPM | Audiobyte

I am not sure about the proprietary part, but if it supplies 12VDC and 5VDC I guess it could be used in an interoperable manner.

[Pepsican]

I just read these responses, thanks john and pesican, and i will read details on uptone, but wondering if it can be put in "layman's terms" without things like galvanic isolation, and john's response "REGEN's primary purpose is to regenerate the USB signal as close a possible".

 

i can kind of picture the digital signal, 5v, and noise as three different signals....when john says to regenerate the digital signal, i see that pretty much the same as removing the noise and the 5v (and my understanding is the noise is a result of the 5v signal, so if you remove the 5v, you should be essentially be removign the noise too?, leaving the digital signal, and then adding a cleaner 5v later (if needed). So i am kind of confused by "regenerating". isn't that the same as removing the 5v and noise and making the digital signal available to the dac? or does it actually reproduce the digital signal? It would seem like we would want the least "reproduction" as possible to eliminate errors? Am i totally in left field or what? I would hope it could be explained better in "layman's" terms.

 

Ok, in layman's terms. The longer the signal needs to travel, the more messed up it gets. Every USB receiver (in laptops, PCs, TVs, cameras, you name it) therefore has to do a clean up exercise. That applies too to the USB receiver in the DAC. When the USB receiver in the DAC does that clean up, it creates its own noise, which in turn affects the DAC chip itself. Oops... What Regen does, is take over the job of the USB receiver in the DAC, thereby preventing the feedback into the DAC chip. It also reclocks the digital signal. And while is is doing that, it also stops the noisy 5V from the USB source and replaces it with a stable, almost noise free 5V.

 

In comparison, the Jitterbug only does the last job, not the first job, which is arguably the more important one.

[Forehaven]

Ok, in layman's terms. The longer the signal needs to travel, the more messed up it gets. Every USB receiver (in laptops, PCs, TVs, cameras, you name it) therefore has to do a clean up exercise. That applies too to the USB receiver in the DAC. When the USB receiver in the DAC does that clean up, it creates its own noise, which in turn affects the DAC chip itself. Oops... What Regen does, is take over the job of the USB receiver in the DAC, thereby preventing the feedback into the DAC chip. It also reclocks the digital signal. And while is is doing that, it also stops the noisy 5V from the USB source and replaces it with a stable, almost noise free 5V.

 

In comparison, the Jitterbug only does the last job, not the first job, which is arguably the more important one.

 

So would this be beneficial on a FW dac like my LIO 8, or only is effective on USB dac's?

[Pepsican]

So would this be beneficial on a FW dac like my LIO 8, or only is effective on USB dac's?

 

It's for when you use the USB-in port on your DAC.

[JohnSwenson]

There are 4 wires and a shield in the common USB implementation used for audio. Two of the wires form a differential pair for the data, it's a bi-directional bus, sometimes data is going one direction and other times it is going the other direction. There is a ground wire and a 5V wire.

 

The 5V wire is used to power a device if it doesn't have it's own power. It doesn't have anything directly to do with the data. The ground wire is used as the ground when powering a device, but is also used in conjunction with the data wires in a few special cases. Noise on the 5V wire is not directly related to the signal.

 

The signal integrity I talked about above is a property of the signal on the data wires, it has nothing to do with the 5V wire. The SI is comprised of the ramp time of the signal edges, amplitude of the signal, noise sitting on top of the signal and jitter of the edges. The computer determines this originally and then it gets degraded by running through a cable.

 

The REGEN, reads the data on the bus and retransmits it but with very low jitter and low noise.

 

The 5V and GND wires are not directly related to this process (the 5V line is not even used in the REGEN), but can affect things indirectly. In most cables the data and power wires are sitting in the same bundle without any shielding between them. Thus it is possible for noise on the power and ground wires to add noise to the data signals. Theoretically this shouldn't be a problem because the data is a differential pair, any noise that is the SAME on both D+ and D- will get ignored by the differential receiver. But cables are not perfect, and power/GND wires are rarely completely symmetrical with the data wires, thus noise can be picked slightly differently by the two wires causing increased noise in the receiver.

 

This noise picked up by the data wires from the power/GND wires is what these dual head cables are trying to fix, by running the power in a separate cable they cut down on the noise injected into the data wires.

 

Now if your DAC actually uses the 5V line to actually power it's circuitry THEN the noise on the 5V line comes directly into play. This is why there is separate ultra low noise regulator in the REGEN, to give really clean power power to the DAC IF it uses it. For DACs that don't use the 5V wire, that regulator in the REGEN is not used for anything.

 

John S.

[JohnSwenson]

USB 2.0 Hi-Speed Isolator withstands up to 2kV_RMS

 

Intona ? Products

 

Mass production has already started. The product is expected to be shipped at the end of September 2015. Target net price is 209 €.

 

Thanks for posting that, I was not aware of this product. It looks like they actually did what I had been describing this would take.

 

Since it is completely powered by the upstream 5V power, they must be using a switching isolating regulator to drive the downstream side. (a high frequency oscillator driving a transformer, then diodes and caps on the other side). We don't know anything about how good the signal is coming out of this other than it is isolated from the input. I hope they good job of it.

 

John S.

[Superdad]

USB 2.0 Hi-Speed Isolator withstands up to 2kV_RMS

 

Intona ? Products

 

Mass production has already started. The product is expected to be shipped at the end of September 2015. Target net price is 209 €.

 

AFAIK, that device will be the first affordable in-line USB high-speed isolators. What is interesting about it to me is that the the company's main area of expertise seems to be programming and licensing to other various FPGA cores for DSP. There are many much larger firms that do that, but Intona appears to be focused on audio-related IP.

 

As mentioned by me in some other thread here at CA, John and I had started looking into licensing a high-speed USB hub core for an FPGA as that is what would be required if we were to do an uber-REGEN that was galvanically isolated. An FPGA, two power supplies, low-jitter clocks, expensive GMR isolators, ultra-low-noise regulators (same as the ones in the current REGEN). But after hearing the updated "amber" REGEN we set that whole project aside.

 

Like John said, the Intona device will accomplish galvanic isolation, but it is unlikely to be optimized for highest signal integrity and impedance matching. Plus I was disappointed when I read it is bus powered. Still, I'll probably order one when they ship. Might be a nice thing to put in front of a REGEN (with the REGEN still being the device directly attached to the DAC).

 

Ciao,

--Alex C.

[One and a half]

Galvanic Isolation is really difficult and there are still issues....

 

The 5V rail is a nuisance at times. With the audiophillieo 2 for example, I had the 5V powered all the time from an external source. If there was an interruption to the USB input (a disconnection), the S/PDIF output once established again would hang until the 5V was powered off and reapplied again.

I also tried the same approach with the MDP-3 where the function of the 5V is such that it notifies the DAC that a signal will arrive shortly rather than power the USB front end in the DAC. If the 5V is always powered, there's a either a mismatch in timing of when the 5V is applied and the data flow arrives. If the time is excessive, say 5s or more, the DAC loses lock and there's no music.

 

When the 5V is part of a standard USB transmission using a cable as is, there's no problem.

 

Would the Regen produce the 5V all the time, or would it coordinate with the USB received signal?

[Superdad]

Would the Regen produce the 5V all the time, or would it coordinate with the USB received signal?

 

REGEN supplies the 5V all the time. Would be complex and possibly cause other problems to have it follow presence or absence of 5V from the cable.

[One and a half]

REGEN supplies the 5V all the time. Would be complex and possibly cause other problems to have it follow presence or absence of 5V from the cable.

 

Yup the 5V rail is a nuisance.

[sandyk]

Yup the 5V rail is a nuisance.

I agree with your deleted remarks too. Sometimes with the +5V disconnected in the PC end's USB-A plug, it MAY be able to fool the system with something like a 2.2K resistor between + and - on the plug.