IFi Defender +…the polish you need for dirty USB audio!

[voodooless]

This works perfectly and has been successfully tested on countless DACs.
For DACs with their own power supply, you only need to disconnect the positive supply voltage; the negative supply voltage is needed to detect the connection.
For a DAC that requires a 5V supply, the positive supply voltage from the host is also disconnected, and the DAC is powered with 5 volts via the Y-cable.

Small circuit boards that accomplish exactly this are also available on AliExpress.

I think this ifi device basically does this. If you look at the inside images, it’s just a bunch of transistors, probably to automatically switch between the supplies.

 

[Roland68]

I think this ifi device basically does this. If you look at the inside images, it’s just a bunch of transistors, probably to automatically switch between the supplies.

On the other hand, there are no components?
That would be really disappointing.

If the part costs more than €3-4 to manufacture, that would be ridiculous.

 

[Schlippwhip68]

Noise or ground loops via USB are relatively common when using a PC as the source. At least as far as I have seen. You can simply hear it, as soon as the amp/active speaker is on even with no music playing. The noise also often changes if you load the CPU or GPU and sometimes even if you move your mouse. On DACs with a well designed grounding scheme it shouldn't make it to the amp, but sometimes it just does.

Agreed, there is also a notable difference using different power supply's for the Defender itself. I tested 3 power supplies and by far the best outcome i.e. cleanest sound is with the iFi Power X.

 

[Roland68]

Agreed, there is also a notable difference using different power supply's for the Defender itself. I tested 3 power supplies and by far the best outcome i.e. cleanest sound is with the iFi Power X.

Assuming the audible differences are nearly as significant as you describe, then you're only treating the symptoms.
If you actually achieve an audible improvement with such simple means, then it's safe to assume that some of the information was already lost or severely degraded beforehand.
It's physically impossible to recover lost information, and you can't improve a digital music signal.

Perhaps it would be more sensible to invest the money in a technically sound and properly functioning source rather than in essentially useless accessories.

By far the most effective device for USB problems is the Intona USB 2.0 Isolator. This device was developed for industrial, measurement, and medical technology applications and is also used in the development and measurement of professional audio equipment, including, as far as I know, by RME.
In a technically sound hi-fi system with a USB source, the Intona USB Isolator is completely useless and offers no advantage, neither audible nor measurable.

Perhaps you should buy one of those and only return it if it doesn't make an audible difference in your system.

 

[phofman]

By far the most effective device for USB problems is the Intona USB 2.0 Isolator. This device was developed for industrial, measurement, and medical technology applications and is also used in the development and measurement of professional audio equipment, including, as far as I know, by RME.

IMHO that isolator was developed (and priced) way before single-chip USB2 highspeed isolators from TI and Analog Devices hit the market. Today the price of high-speed isolators has dropped to below 40USD from China. Yes, they are still very useful.

 

[Roland68]

IMHO that isolator was developed (and priced) way before single-chip USB2 highspeed isolators from TI and Analog Devices hit the market. Today the price of high-speed isolators has dropped to below 40USD from China. Yes, they are still very useful.

That's correct, but unfortunately, none of these inexpensive USB isolators can compete with those from Intona.
We use the Intona ourselves for measurements of our industrial products. I also thought that the newer, inexpensive USB isolators would work just as well, and ordered three different ones, including one of the ones recommended here. But the difference is truly night and day (not in terms of sound quality, but in industrial applications), and crashes during our measurements were unfortunately the result.

The devices do work in principle, but unfortunately not always adequately. There are also cases here in the forum where galvanic isolation of the USB connection via an inexpensive DDC and a Toslink connection to the DAC worked perfectly, but one of the inexpensive USB isolators didn't solve the problem satisfactorily.

I've lent out the Intona countless times over the past few years, and it has always functioned flawlessly.
From my experience in medical technology, I know that even more expensive competitor products don't always work perfectly, unlike the Intona.
At least, that's been my practical experience.

However, as I said, that's just treating the symptoms. It's always better to solve the underlying problem; then you don't need any of these devices.

 

[phofman]

Thanks. Is the problem with the actual data transfer, or perhaps inadequate power of the isolated DC/DC power supply, incapable of powering the attached USB device reliably? It would be rather surprising if TI and Analog Devices designed a faulty chip. The DC/DC functionality is unrelated to the actual isolator chip.

 

[Roland68]

Thanks. Is the problem with the actual data transfer, or perhaps inadequate power of the isolated DC/DC power supply, incapable of powering the attached USB device reliably? It would be rather surprising if TI and Analog Devices designed a faulty chip. The DC/DC functionality is unrelated to the actual isolator chip.

The problem isn't the power being transmitted, but likely the DC/DC filtering and digital transmission. However, this also depends on the specific problems and the application.

I know of cases where one of the well-known wall wart filters from the DIY forum produced perfect results (connected directly to the computer's +/- power supply of the USB output), but also cases where complete galvanic isolation via Toslink was necessary.

 

[phofman]

I wonder if it was the built-in DC/DC PSU, and using a standalone 5V PSU for the device would have fixed the problem. I would be really surprised if those new chips were having issues with USB2 highspeed when that is their only purpose.

 

[Roland68]

I wonder if it was the built-in DC/DC PSU, and using a standalone 5V PSU for the device would have fixed the problem. I would be really surprised if those new chips were having issues with USB2 highspeed when that is their only purpose.

It's not about speed issues.
In the case at hand, disconnecting the positive supply with an external power supply would likely have had the same result.
But again, there are several potential problems.

The two basic chip solutions from TI and ADI naturally have limited functionality and were developed as inexpensive standard applications. Of course, the implementation of the ICs and the DC/DC filtering are also crucial.
Therefore, there will likely be differences between the various implementations.

The Intona is a very sophisticated bidirectional isolator/reclocker/repacketizer at the physical level, which has been used for over 11 years in all areas of industry, medical technology, science/research, aviation, etc., and must function without any limitations. It also has a protection function, as devices with 4, 5, or 6 digits are often connected.
The firmware is also customizable if needed, and Intona has a head start of about 15 years in this area.

You would have to delve deeper into the subject, take your own measurements, and analyze the various different cases.
But what's the point if you can avoid these cases altogether?

Used Mac Minis and HP business desktops and notebooks in good condition can be found for €150-250. I've never had any USB problems with audio using these, and the use of an Intona USB isolator (or any other) is completely useless.

 

[LesterHiggins]

I have an Lenovo M920q PC. This little 1 Litre box of wires ‘n’ bits is my primary music server. It’s quite a cost effective solution. It is not the most quite or cleanest motherboard made and many have noted this with the Lenovo PC. In short the USB on this thing is dirty, noisy also impossible to dumb the power down but that is another issue.

I run an usb cable to the Douk Audio U2 Pro (2025) and optical to the dac. I ran coaxial to the dac but static crackle soon killed that idea.

The sound was clean, not run home and tell your mum impressive, but clean enough. In consideration of nullifying the static via the coaxial connection from the U2 to the dac I decided to grab an iFi defender + when I saw one at a very reasonable price. I had no idea what this thing would do realtime and only had theory based on what it supposed to do.

The iFi Defender arrived today and after some fiddling around got the thing pushed into place and the U2 Pro plugged in. I am using a cheap GaN 5v power supply from Amazon to power the Defender.

Everything a place and into Apple Music to pay a few select tracks with desired frequency highlights.

What a difference!

Big clean up operation complete and in one little device! Just how dirty the usb power delivery straight from the PC to the U2 is now clearly evident by just how effective the iFi Defender + performs. It cuts off the power from the usb that was feeding the U2 Pro and with the power supply plugged into the Defender now feeds the U2 Pro with that power.

The difference is literal night and day. The bass is much cleaner and more precise without all that dirty noise coming from the motherboard. The vocals have lifelike clarity and the high etc and all clean, clean, clean!

This is a significant upgrade to any PC based system imo and not just for ground looping, hum etc but also just for dirty usb!

I would recommend this little gadget to anyone using a PC and usb as their audio source without any hesitation. I don’t work for ifi or am sponsored to say this. This is purely my own equipment and experience and I have to say iFi, top job boys and girls!

*blows whistle* Flag on the play.
Improper use of "literal." 20 yard penalty. Repeat 1st down.

It does not change the time of day, it is not literally night and day, it is metaphorically night and day, but you don't have to say "metaphorically night and day," we already know "night and day" is a metaphor not to be taken literally.