• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Topping D90 Balanced USB DAC Review

Final

Member
Joined
Mar 19, 2018
Messages
49
Likes
75
I have had a laptop power supply kick all kinds of noise into HP/amps not even connected in the chain electrically (PC was wifi to streamer)- moved the supply as far away as possible- interference stops. Horrendous things!
Felt almost wrong to hit the Like Button:cool:
 

Final

Member
Joined
Mar 19, 2018
Messages
49
Likes
75
Do you notice any GPU noise when gaming or other heavy GPU usage? Or is this not part of your personal system usage? Watching Video's doesn't cause the GPU to work hard enough to generate noise.

What brand / model of power supply was the quiet one? I'm build a new computer based on the AMD Ryzen 5000 / Radeon 6000 parts, and need to select a power supply.

Thank you! :)
The noise was there without any heavy GPU usage. I have a Asus PC snd bought a new AC pose supply from them. It saus ADP -90YD B on the model number.
 

Harmonie

Major Contributor
Forum Donor
Joined
May 30, 2020
Messages
1,927
Likes
2,084
Location
France
Just wondering:
What's so special about the Supra 0.7m cable and it's noise blocker ?
Does it have Ferrite inside ?

I have Eupen power cables that have Ferrite powder for that noise blocking function, quite effective.

Eupen 2,5mm.jpg
 

Neto Rare

Member
Joined
Jun 6, 2020
Messages
83
Likes
54
Location
Socal
In fact Roon can stream native Tidal MQA to a Raspberry Pi 4, that's what I use. I'm running RoPiee on my Pi 4, I have no extraeneous noise at all, just great music.

This is why you don't post in the wee hours lol. I actually don't even use RP4 yet!

I use a combination of BubbleUPnP/Foobar (supports VST plugins) on my pc and use Mconnect lite on android to stream Tidal/Qobuz.
 

Veri

Master Contributor
Joined
Feb 6, 2018
Messages
9,596
Likes
12,036
Just wondering:
What's so special about the Supra 0.7m cable and it's noise blocker ?
They were just suggested, not actually verified to be worth the money or any better than one supplied with your DAC. I personally bought some singxer brand USB cables from shenzhenaudio and they have ferrites at both ends and are gold plated. They work.. just fine/great. No longer available though, guess you can only get them with purchase of a singxer product as a freebie, now.
 

hmscott

Addicted to Fun and Learning
Joined
May 17, 2020
Messages
604
Likes
493
The noise was there without any heavy GPU usage. I have a Asus PC snd bought a new AC pose supply from them. It saus ADP -90YD B on the model number.
Ah, just a noisy power brick, that's a common problem, those little power bricks are made by the zillions across many brands and models and a fault in manufacturing causing noise is something I have seen a number of times - not for audio, but for display and other laptop problems.

I was hoping it was a desktop power supply. I'm glad you found the problem, I'm sure it was a relief at the time.
 

hmscott

Addicted to Fun and Learning
Joined
May 17, 2020
Messages
604
Likes
493
Just wondering:
What's so special about the Supra 0.7m cable and it's noise blocker ?
Does it have Ferrite inside ?

I have Eupen power cables that have Ferrite powder for that noise blocking function, quite effective.
Very nice :)

I'm not sure what ferrite powder unformed into a toroid for magnetic effects does, perhaps it's simply a powdered iron shield? There are formed flat plates of ferrite designed for winding wire to effect as described in the "spoilers" and on the site I am quoting, perhaps the line of ferrite powder also acts in this way as distributed through twisted pair wires?
The ferrite cores are each designed to reduce at the source the noise as a field effect loss. That is why you have one at each end as the potential at the source (bi directional, one at each end) helps in the effect. That's is also why you can put them in series and parallel with windings to further reduce the noise - low pass filter works with a percentage of loss at the high frequencies which is additive. The ferrite core dimensions and density or iron in suspension along with whether the core wraps the cable, or the cable turns wrap the core, changes the potential effects.

Simply wrapping a snap on core cuts a bit of noise - in my case one ferrite core wrapped around the cable at each end drops the chirpy noises enough to suffice - it's low enough even in quiet passages to be a faint noise barely detectable.

The Supra cable doesn't have ferrite cores or ferrite powder that I know of, certainly not externally, and relies on shielding between power and data along with using twisted pair wiring independent for power and data. And, that's why the chirpy noises aren't filtered out at all and are too noisy to allow the RCA cables to be connected.

The Supra cable does have positive audible improvements to the signal I am hearing over the Tripp Lite cable with 2 ferrite cores. IDK if that is due to the effects of cutting the signal - low pass filters have some loss at the desired frequencies too - as I've been told it's best to only use 1 ferrite core at one end, IDK which end it is supposed to be at but I assume the source end - as 2 cores tends to affect the audio negatively. So what I might be noticing is a degradation of the signal with the Tripp Lite with Ferrite cores vs the Supra Cable which has no filtering effect in the design.
A short snippet of discussion on how ferrite cores work:
https://article.murata.com/en-us/article/basics-of-noise-countermeasures-lesson-8

"Therefore, the coil functions as a low-pass filter that blocks high-frequency current, enabling attenuation of high-frequency noise. Furthermore, the use of a ferrite core also provides an additional effect. When current flows to an inductor comprising a ferrite core, magnetic flux is generated in the ferrite core, and the current energy is converted into magnetic energy. However, when the current changes, this magnetic flux is converted back into current by electromagnetic induction. At this time, not all of the magnetic flux energy is returned to current energy, and some is lost as magnetic loss. (This is called "hysteresis loss.") As a result, part of the noise current passing through the conducting wires is lost as magnetic loss, reducing the energy. "

"The noise removal performance of ferrite cores varies according to the ferrite materials and shape.The magnetic permeability changes according to the ferrite materials, so the impedance also differs. In addition, the ratio between the resistance component and reactance component of the impedance also differs according to the materials. However, the materials of ferrite cores sold as noise countermeasures are blended specially for noise countermeasures, so there is no great difference in the characteristics regardless of the material selected. "

Here's a larger explanation with diagrams:
https://article.murata.com/en-us/article/basics-of-noise-countermeasures-lesson-8
<Noise countermeasure parts that can be used without changing the board>
Previously, we introduced noise countermeasure parts mounted on a board as part of the electronic circuit. This time we introduce noise countermeasure parts that do not require mounting on a board. (Although they are sometimes fixed to the board...)
As previously introduced, when commercializing an electronic device, it is necessary to check that the noise emitted from the device satisfies the EMI regulations. However, the final check cannot be performed until the device design is complete. Recently, experience concerning designs that do not emit noise is increasing, and various measures are being implemented beforehand to prevent the generation of noise, but of course the effects cannot be known until the final test. There is no problem as long as the noise level is within the regulation value as expected at this point, but the test results sometimes exceed the regulation value. When the delivery date is near, there is no time to change the board, so parts such as ferrite cores that enable countermeasures without changing the board come in handy.

<Ferrite cores are lumps of ferrite>
Ferrite cores are ceramic magnetic bodies consisting of ferrites (soft ferrites) processed into various shapes. Formerly, coils were often made by winding conducting wires around a ring-shaped ferrite core, so ferrite used for noise countermeasures is likewise called a ferrite core.

Ferrites include Mn-Zn ferrite and Ni-Zn ferrite according to the composition. Mn-Zn ferrite is conductive, so it requires insulation work, and Ni-Zn ferrite has better high-frequency characteristics. For these and other reasons, Ni-Zn ferrite is often used for noise countermeasures.

<Principle by which ferrite cores remove noise>
Ferrite cores come in various shapes, but most are ring-shaped. By passing conducting wires through the hole of the ring, the conducting wires and the ferrite core form a coil (inductor). This coil (inductor) is based on the same principle as that of an electronic part inductor, so the impedance increases together with the frequency as shown in Figure 1. Therefore, the coil functions as a low-pass filter that blocks high-frequency current, enabling attenuation of high-frequency noise. Furthermore, the use of a ferrite core also provides an additional effect. When current flows to an inductor comprising a ferrite core, magnetic flux is generated in the ferrite core, and the current energy is converted into magnetic energy. However, when the current changes, this magnetic flux is converted back into current by electromagnetic induction. At this time, not all of the magnetic flux energy is returned to current energy, and some is lost as magnetic loss. (This is called "hysteresis loss.") As a result, part of the noise current passing through the conducting wires is lost as magnetic loss, reducing the energy. The right side of Figure 1 shows the impedance characteristics of a coil with conducting wires passed through a ferrite core. The impedance of a normal coil consists mostly of the reactance (X) component, but when a ferrite core is used, the resistance (R) component becomes extremely large. This is a result of selecting ferrite materials suitable for noise countermeasures, and causes the noise energy consumption effect due to magnetic loss to account for a larger portion of the noise removal effect of the ferrite core than the current limiting effect of high impedance.

en-20120713-p1_img0001.png
Figure 1. Noise removal effects of ferrite cores<Ferrite core specifications and performance>
The noise removal performance of ferrite cores varies according to the ferrite materials and shape.The magnetic permeability changes according to the ferrite materials, so the impedance also differs. In addition, the ratio between the resistance component and reactance component of the impedance also differs according to the materials. However, the materials of ferrite cores sold as noise countermeasures are blended specially for noise countermeasures, so there is no great difference in the characteristics regardless of the material selected.

en-20120713-p2_img0001.png
Figure 2. Difference in performance due to different ferrite materials
The inductance increases (proportionally to the square of the number of windings) together with the number of times the conducting wire is passed through the ring core (number of windings). However, when the conducting wire is wrapped around the core, the winding start (entrance) and winding end (exit) come close to each other and have floating capacitance between them. High-frequency noise is conveyed via this floating capacitance area, which is a factor lowering the high-frequency performance. Therefore, in consideration of the target frequencies for noise reduction, the number of windings must either be increased to target the low-frequency range, or reduced to target the high-frequency range.

en-20120713-p2_img0002.png
Figure 3. Relationship between number of windings and impedance
In addition, the ferrite core dimensions affect the performance as shown in Figure 4. Therefore, a ring core with the smallest inner diameter and widest cross-sectional area possible should be selected.

en-20120713-p2_img0003.png
Figure 4. Ferrite core dimensions and performance
<Using a ferrite core as a common mode choke coil>
Due to their convenience, ferrite cores are often used on cables. However, these cables include interface cables, power cables, and other multiple conducting wires that run in parallel, so common mode noise is often an issue. Common mode choke coils are an effective noise countermeasure in these cases, and common mode choke coil functions can be achieved by passing the cables together through a single ferrite core. For example, in case of an interface cable that has multiple signal lines wired to a single end, it is difficult to wire common mode choke coils. However, common mode choke coil functions can be easily achieved by passing all of the interface cable wires together through a ferrite core.

en-20120713-p3_img0001.png
Figure 5. Using a ferrite core as a common mode choke coil<Various ferrite core shapes>
Thus far we have introduced ring-shaped ferrite cores, but various other ferrite core shapes have also been commercialized. These include cores with wide and thin shapes that match the shapes of flat cables and flexible printed circuits (FPC), and divided cores that are assembled around cables to eliminate the work of passing the cables through the cores. In addition, simple plate-shaped cores that are not rings are also provided. These plate cores are attached over ICs and other locations that emit electromagnetic waves, and aim for radio wave absorption effects by attenuating the electromagnetic waves passing through the plate core by the magnetic loss of the ferrite.

en-20120713-p3_img0002.png
Figure 6. Various ferrite core shapes
It's been more than a few years since I was fresh on the theory and implementation - and that was for other specific RFI / EMI problems - I'm mostly interested in buying a quick off the shelf solution for this specific instance rather than start a research project of scale.

That site also has more details and products, I may have not even gotten the best out of it yet.
Oh, and the difference in audio I am hearing with the Supra USB 0.7m cable I have not heard with my D90 MQA / A90 before, not with the pack-in cable, another nice audiophile USB cable, or the Tripp Lite with 2 built-in Ferrite Cores, and it's a nice enough audible improvement for me to keep the Supra Cable even if it didn't do anything at all for the USB GPU Birdies.

I'll sum up what I am hearing in a short sentence. The audio has atmosphere.

The other cables had various stage width depth and height, and the Supra cable has all of that as far out as I've ever heard my system with my long used headphones and IEM's, but what I am clearly hearing is reverberation between objects that gives the presentation "atmosphere" that I haven't experience before.

This effect provides a audible feeling of individual object separation and brings a connectedness of not only the objects but of the room itself reflecting the objects and providing more clues as to the location for each source sound - this adds a "presence" to each object.

And, this overall result is that I am hearing spatial separation of individual sounds that now have a spatial presence they didn't have before.

This quality has grown over the time from when I first put the cable in and is very clearly there after about 10 hours of listening - it wasn't there to start, in fact I didn't like the sound in the first hour or so as it was flat and closed in and overall was worse in every way from the Tripp Lite cable I had just swapped it with.

At this point I was already detailing the ebay return note I was going to send to return the Supra cable. It was that bad.

If it wasn't that short stretch of a few hours without GPU birdies I would have wrapped it up before letting it burn-in further. The audio levels came back in about 5-6 hours - I moved the Gain back to M then to L where it is now and I also noticed something in tracks I hadn't heard before - atmosphere. That "echoey trail" you hear illuminating objects in live performances, but in full, not just the loudest peaks but the whole effect.

So, I'm very happy with the Supra cable audio, and without the RCA cables connected to stop the USB GPU Birdies the audio is even better than before, I'm really surprised and happy with the upgrade from swapping the USB cable to the Supra.

For now I'm going to enjoy the music and try to not be concerned about the USB GPU noise, perhaps I will discover something, or it will come along, or be suggested that assures me a solution will get rid of the chirpy noises - I'm pretty sure the conduction through the RCA cables can be stopped, but I'm not sure if the GPU noise will be induced otherwise into degrading the signal on or through the motherboard. And, there I go starting to think about it again...
Previously I couldn't stand listening to the Blon BL-05 for more than a short time even after burning in for weeks, it sounded "weird" as compared to the sound of the Blon BL-03 or other well mannered IEM's. While the Supra cable was run in for 24+ hours I spent an entire day really enjoying listening to the Blon BL-05. That is a real and surprising improvement that turned an IEM I couldn't stand into one I am really enjoying.

Now I'm swapping between my headphones and what I am hearing that I wasn't hearing before swapping in the Supra cable is making me very happy. My Sennheiser HD 598 CS headphones which have always been "flat" and not all that enjoyable - more of a yeoman's service headphone, are now very enjoyable.

The KZ ZAX's now have almost too much information. They were very enjoyable to listen to before - very engaging but not too much, but now they are too busy - overwhelming... I hope I get used to them again as I was really enjoying them. So there can be effects positive and negative for such "improvements".

These positive effects will likely become the new normal, but I am hoping my previously less enjoyable headphones will continue to shine.
Thanks to everyone that contributed suggestions, I do appreciate your suggestions, they helped immensely - that Supra USB cable of 0.7m only cost $44 + shipping = $48.18, and well worth the expense for the improvement in the system audio I am hearing in all of my headphones and IEM's.

Update: After a little more digging on the Jenving site that makes the Supra USB cable I found this interesting Note about Ferrite Cores interactions with their cables:

"Note! Be careful in the use of ferrites. Not all ferrites are tuned correctly to be used with this high frequency data transfer cable without changing the electrical properties of the cable to the point where the data transfer is corrupted and/or slowed down. This is a general notification for all digital interconnects. We have encountered HDMI cables not coming even close to the original excellent performance, thanks to later added ferrites with the benign intention of reducing RFI."
http://www.jenving.com/products/view/usb-2.0-a-b-1001908381

That describes what I am hearing with the Tripp Lite cable audio results vs Supra. I'll ask them what they suggest for Ferrite materials and application with their cables. It may be they would need to do this and tune the Ferrite application with the cable to best effect and least detriment.

I would have assumed Tripp-Lite did this with their USB + 2 Ferrite Cores cable. Selected the Ferrite material and design tuned for their USB cable application, but in comparison to the Ferrite-less Supra cable the Trip-Lite cable with the Ferrite Cores has poor audio quality - but it does reduce the GPU chirpy noise.

Update: Jenving uses this specific solder - Almit SR-34 Super - to build their cables, and provides the solder to their customers too:
http://www.jenving.com/products/view/solder-tin-0.8mm-sn-ag-500g-1086000229

One thing leads to another, maybe I could use this on the RCA cables coming from the D90 MQA to A90 to interrupt the USB sourced GPU birdies coming through ground noise transfer?

It's price + shipping is a bit pricey, and comes from another vendor on Ebay, I'll ask the Surpra cable provider if they can provide the Supra AGS-10k for less...

Supra AGS-10K Hum Blocker - Ground Loop Humming $128 USD + $56.87 shipping.
https://www.ebay.com/itm/Supra-AGS-10K-Hum-Blocker-Ground-Loop-Humming/233648525191

ProductsGround Separators Supra Hum-Buster
http://www.jenving.com/products/ground-separators-89
The Supra AGS-10k is can be connected between the audio outputs on a computer/laptop and the audio inputs of a hi-fi amplifier to block any audible hum. Alternativly, it can be used between any two line-level items to avoid ground loop hum. Different to "common ground" separators, the Supra AGS-10k is of a high-end design using toroid transformers with wide frequency range and low distortion, even at the lowest frequencies.

The chassis is made from extruded aluminium, which helps prevent any RFI pick-up.

Specification:
  • Input impedance: 10 kohms
  • Output impedance: 10 kohms
  • Ratio: 1:1
  • Frequency Range: 15 Hz - 22 kHz
  • Size: 53 (H) x 96 (W) x 89 (L) mm
  • Weight: 440 g
1604320867542.png

Update: zendada responded and they can provide the Supra AGS-10k, I'm now waiting for them to list it on ebay so I can order it.
 
Last edited:

Final

Member
Joined
Mar 19, 2018
Messages
49
Likes
75
Ah, just a noisy power brick, that's a common problem, those little power bricks are made by the zillions across many brands and models and a fault in manufacturing causing noise is something I have seen a number of times - not for audio, but for display and other laptop problems.

I was hoping it was a desktop power supply. I'm glad you found the problem, I'm sure it was a relief at the time.
It sure was. I hope you find a good solution for your desktop.
 

Dclone

Member
Joined
Apr 30, 2020
Messages
53
Likes
46
Very nice :)

I'm not sure what ferrite powder unformed into a toroid for magnetic effects does, perhaps it's simply a powdered iron shield? There are formed flat plates of ferrite designed for winding wire to effect as described in the "spoilers" and on the site I am quoting, perhaps the line of ferrite powder also acts in this way as distributed through twisted pair wires?
The ferrite cores are each designed to reduce at the source the noise as a field effect loss. That is why you have one at each end as the potential at the source (bi directional, one at each end) helps in the effect. That's is also why you can put them in series and parallel with windings to further reduce the noise - low pass filter works with a percentage of loss at the high frequencies which is additive. The ferrite core dimensions and density or iron in suspension along with whether the core wraps the cable, or the cable turns wrap the core, changes the potential effects.

Simply wrapping a snap on core cuts a bit of noise - in my case one ferrite core wrapped around the cable at each end drops the chirpy noises enough to suffice - it's low enough even in quiet passages to be a faint noise barely detectable.

The Supra cable doesn't have ferrite cores or ferrite powder that I know of, certainly not externally, and relies on shielding between power and data along with using twisted pair wiring independent for power and data. And, that's why the chirpy noises aren't filtered out at all and are too noisy to allow the RCA cables to be connected.

The Supra cable does have positive audible improvements to the signal I am hearing over the Tripp Lite cable with 2 ferrite cores. IDK if that is due to the effects of cutting the signal - low pass filters have some loss at the desired frequencies too - as I've been told it's best to only use 1 ferrite core at one end, IDK which end it is supposed to be at but I assume the source end - as 2 cores tends to affect the audio negatively. So what I might be noticing is a degradation of the signal with the Tripp Lite with Ferrite cores vs the Supra Cable which has no filtering effect in the design.
A short snippet of discussion on how ferrite cores work:
"Therefore, the coil functions as a low-pass filter that blocks high-frequency current, enabling attenuation of high-frequency noise. Furthermore, the use of a ferrite core also provides an additional effect. When current flows to an inductor comprising a ferrite core, magnetic flux is generated in the ferrite core, and the current energy is converted into magnetic energy. However, when the current changes, this magnetic flux is converted back into current by electromagnetic induction. At this time, not all of the magnetic flux energy is returned to current energy, and some is lost as magnetic loss. (This is called "hysteresis loss.") As a result, part of the noise current passing through the conducting wires is lost as magnetic loss, reducing the energy. "

"The noise removal performance of ferrite cores varies according to the ferrite materials and shape.The magnetic permeability changes according to the ferrite materials, so the impedance also differs. In addition, the ratio between the resistance component and reactance component of the impedance also differs according to the materials. However, the materials of ferrite cores sold as noise countermeasures are blended specially for noise countermeasures, so there is no great difference in the characteristics regardless of the material selected. "

Here's a larger explanation with diagrams:
https://article.murata.com/en-us/article/basics-of-noise-countermeasures-lesson-8
<Noise countermeasure parts that can be used without changing the board>
Previously, we introduced noise countermeasure parts mounted on a board as part of the electronic circuit. This time we introduce noise countermeasure parts that do not require mounting on a board. (Although they are sometimes fixed to the board...)
As previously introduced, when commercializing an electronic device, it is necessary to check that the noise emitted from the device satisfies the EMI regulations. However, the final check cannot be performed until the device design is complete. Recently, experience concerning designs that do not emit noise is increasing, and various measures are being implemented beforehand to prevent the generation of noise, but of course the effects cannot be known until the final test. There is no problem as long as the noise level is within the regulation value as expected at this point, but the test results sometimes exceed the regulation value. When the delivery date is near, there is no time to change the board, so parts such as ferrite cores that enable countermeasures without changing the board come in handy.

<Ferrite cores are lumps of ferrite>
Ferrite cores are ceramic magnetic bodies consisting of ferrites (soft ferrites) processed into various shapes. Formerly, coils were often made by winding conducting wires around a ring-shaped ferrite core, so ferrite used for noise countermeasures is likewise called a ferrite core.

Ferrites include Mn-Zn ferrite and Ni-Zn ferrite according to the composition. Mn-Zn ferrite is conductive, so it requires insulation work, and Ni-Zn ferrite has better high-frequency characteristics. For these and other reasons, Ni-Zn ferrite is often used for noise countermeasures.

<Principle by which ferrite cores remove noise>
Ferrite cores come in various shapes, but most are ring-shaped. By passing conducting wires through the hole of the ring, the conducting wires and the ferrite core form a coil (inductor). This coil (inductor) is based on the same principle as that of an electronic part inductor, so the impedance increases together with the frequency as shown in Figure 1. Therefore, the coil functions as a low-pass filter that blocks high-frequency current, enabling attenuation of high-frequency noise. Furthermore, the use of a ferrite core also provides an additional effect. When current flows to an inductor comprising a ferrite core, magnetic flux is generated in the ferrite core, and the current energy is converted into magnetic energy. However, when the current changes, this magnetic flux is converted back into current by electromagnetic induction. At this time, not all of the magnetic flux energy is returned to current energy, and some is lost as magnetic loss. (This is called "hysteresis loss.") As a result, part of the noise current passing through the conducting wires is lost as magnetic loss, reducing the energy. The right side of Figure 1 shows the impedance characteristics of a coil with conducting wires passed through a ferrite core. The impedance of a normal coil consists mostly of the reactance (X) component, but when a ferrite core is used, the resistance (R) component becomes extremely large. This is a result of selecting ferrite materials suitable for noise countermeasures, and causes the noise energy consumption effect due to magnetic loss to account for a larger portion of the noise removal effect of the ferrite core than the current limiting effect of high impedance.

en-20120713-p1_img0001.png
Figure 1. Noise removal effects of ferrite cores<Ferrite core specifications and performance>
The noise removal performance of ferrite cores varies according to the ferrite materials and shape.The magnetic permeability changes according to the ferrite materials, so the impedance also differs. In addition, the ratio between the resistance component and reactance component of the impedance also differs according to the materials. However, the materials of ferrite cores sold as noise countermeasures are blended specially for noise countermeasures, so there is no great difference in the characteristics regardless of the material selected.

en-20120713-p2_img0001.png
Figure 2. Difference in performance due to different ferrite materials
The inductance increases (proportionally to the square of the number of windings) together with the number of times the conducting wire is passed through the ring core (number of windings). However, when the conducting wire is wrapped around the core, the winding start (entrance) and winding end (exit) come close to each other and have floating capacitance between them. High-frequency noise is conveyed via this floating capacitance area, which is a factor lowering the high-frequency performance. Therefore, in consideration of the target frequencies for noise reduction, the number of windings must either be increased to target the low-frequency range, or reduced to target the high-frequency range.

en-20120713-p2_img0002.png
Figure 3. Relationship between number of windings and impedance
In addition, the ferrite core dimensions affect the performance as shown in Figure 4. Therefore, a ring core with the smallest inner diameter and widest cross-sectional area possible should be selected.

en-20120713-p2_img0003.png
Figure 4. Ferrite core dimensions and performance
<Using a ferrite core as a common mode choke coil>
Due to their convenience, ferrite cores are often used on cables. However, these cables include interface cables, power cables, and other multiple conducting wires that run in parallel, so common mode noise is often an issue. Common mode choke coils are an effective noise countermeasure in these cases, and common mode choke coil functions can be achieved by passing the cables together through a single ferrite core. For example, in case of an interface cable that has multiple signal lines wired to a single end, it is difficult to wire common mode choke coils. However, common mode choke coil functions can be easily achieved by passing all of the interface cable wires together through a ferrite core.

en-20120713-p3_img0001.png
Figure 5. Using a ferrite core as a common mode choke coil<Various ferrite core shapes>
Thus far we have introduced ring-shaped ferrite cores, but various other ferrite core shapes have also been commercialized. These include cores with wide and thin shapes that match the shapes of flat cables and flexible printed circuits (FPC), and divided cores that are assembled around cables to eliminate the work of passing the cables through the cores. In addition, simple plate-shaped cores that are not rings are also provided. These plate cores are attached over ICs and other locations that emit electromagnetic waves, and aim for radio wave absorption effects by attenuating the electromagnetic waves passing through the plate core by the magnetic loss of the ferrite.

en-20120713-p3_img0002.png
Figure 6. Various ferrite core shapes
It's been more than a few years since I was fresh on the theory and implementation - and that was for other specific RFI / EMI problems - I'm mostly interested in buying a quick off the shelf solution for this specific instance rather than start a research project of scale.

That site also has more details and products, I may have not even gotten the best out of it yet.
Oh, and the difference in audio I am hearing with the Supra USB 0.7m cable I have not heard with my D90 MQA / A90 before, not with the pack-in cable, another nice audiophile USB cable, or the Tripp Lite with 2 built-in Ferrite Cores, and it's a nice enough audible improvement for me to keep the Supra Cable even if it didn't do anything at all for the USB GPU Birdies.

I'll sum up what I am hearing in a short sentence. The audio has atmosphere.

The other cables had various stage width depth and height, and the Supra cable has all of that as far out as I've ever heard my system with my long used headphones and IEM's, but what I am clearly hearing is reverberation between objects that gives the presentation "atmosphere" that I haven't experience before.

This effect provides a audible feeling of individual object separation and brings a connectedness of not only the objects but of the room itself reflecting the objects and providing more clues as to the location for each source sound - this adds a "presence" to each object.

And, this overall result is that I am hearing spatial separation of individual sounds that now have a spatial presence they didn't have before.

This quality has grown over the time from when I first put the cable in and is very clearly there after about 10 hours of listening - it wasn't there to start, in fact I didn't like the sound in the first hour or so as it was flat and closed in and overall was worse in every way from the Tripp Lite cable I had just swapped it with.

And, this is the weird part - it is the first time I've had a break-in period include a large audio level change. The A90 was set to M Gain 12 oclock with SendyAudio Aiva's - I was rocking out - and when I swapped in the Supra cable I had to switch to H Gain about 11 oclock. I was a bit perplexed, I thought something else had happen to attenuate the audio, but no I checked everything and the audio with the Supra USB cable was definitely much lower. There wasn't but a gap of a minute between swapping cables and continuing listening.

At this point I was already detailing the ebay return note I was going to send to return the Supra cable. It was that bad.

If it wasn't that short stretch of a few hours without GPU birdies I would have wrapped it up before letting it burn-in further. The audio levels came back in about 5-6 hours - I moved the Gain back to M then to L where it is now and I also noticed something in tracks I hadn't heard before - atmosphere. That "echoey trail" you hear illuminating objects in live performances, but in full, not just the loudest peaks but the whole effect.

So, I'm very happy with the Supra cable audio, and without the RCA cables connected to stop the USB GPU Birdies the audio is even better than before, I'm really surprised and happy with the upgrade from swapping the USB cable to the Supra.

For now I'm going to enjoy the music and try to not be concerned about the USB GPU noise, perhaps I will discover something, or it will come along, or be suggested that assures me a solution will get rid of the chirpy noises - I'm pretty sure the conduction through the RCA cables can be stopped, but I'm not sure if the GPU noise will be induced otherwise into degrading the signal on or through the motherboard. And, there I go starting to think about it again...
Previously I couldn't stand listening to the Blon BL-05 for more than a short time even after burning in for weeks, it sounded "weird" as compared to the sound of the Blon BL-03 or other well mannered IEM's. While the Supra cable was run in for 24+ hours I spent an entire day really enjoying listening to the Blon BL-05. That is a real and surprising improvement that turned an IEM I couldn't stand into one I am really enjoying.

Now I'm swapping between my headphones and what I am hearing that I wasn't hearing before swapping in the Supra cable is making me very happy. My Sennheiser HD 598 CS headphones which have always been "flat" and not all that enjoyable - more of a yeoman's service headphone, are now very enjoyable.

The KZ ZAX's now have almost too much information. They were very enjoyable to listen to before - very engaging but not too much, but now they are too busy - overwhelming... I hope I get used to them again as I was really enjoying them. So there can be effects positive and negative for such "improvements".

These positive effects will likely become the new normal, but I am hoping my previously less enjoyable headphones will continue to shine.
Thanks to everyone that contributed suggestions, I do appreciate your suggestions, they helped immensely - that Supra USB cable of 0.7m only cost $44 + shipping = $48.18, and well worth the expense for the improvement in the system audio I am hearing in all of my headphones and IEM's.

Update: After a little more digging on the Jenving site that makes the Supra USB cable I found this interesting Note about Ferrite Cores interactions with their cables:

"Note! Be careful in the use of ferrites. Not all ferrites are tuned correctly to be used with this high frequency data transfer cable without changing the electrical properties of the cable to the point where the data transfer is corrupted and/or slowed down. This is a general notification for all digital interconnects. We have encountered HDMI cables not coming even close to the original excellent performance, thanks to later added ferrites with the benign intention of reducing RFI."
http://www.jenving.com/products/view/usb-2.0-a-b-1001908381

That describes what I am hearing with the Tripp Lite cable audio results vs Supra. I'll ask them what they suggest for Ferrite materials and application with their cables. It may be they would need to do this and tune the Ferrite application with the cable to best effect and least detriment.

I would have assumed Tripp-Lite did this with their USB + 2 Ferrite Cores cable. Selected the Ferrite material and design tuned for their USB cable application, but in comparison to the Ferrite-less Supra cable the Trip-Lite cable with the Ferrite Cores has poor audio quality - but it does reduce the GPU chirpy noise.

Update: Jenving uses this specific solder - Almit SR-34 Super - to build their cables, and provides the solder to their customers too:
http://www.jenving.com/products/view/solder-tin-0.8mm-sn-ag-500g-1086000229

One thing leads to another, maybe I could use this on the RCA cables coming from the D90 MQA to A90 to interrupt the USB sourced GPU birdies coming through ground noise transfer?

It's price + shipping is a bit pricey, and comes from another vendor on Ebay, I'll ask the Surpra cable provider if they can provide the Supra AGS-10k for less...

Supra AGS-10K Hum Blocker - Ground Loop Humming $128 USD + $56.87 shipping.
https://www.ebay.com/itm/Supra-AGS-10K-Hum-Blocker-Ground-Loop-Humming/233648525191

ProductsGround Separators Supra Hum-Buster
http://www.jenving.com/products/ground-separators-89
The Supra AGS-10k is can be connected between the audio outputs on a computer/laptop and the audio inputs of a hi-fi amplifier to block any audible hum. Alternativly, it can be used between any two line-level items to avoid ground loop hum. Different to "common ground" separators, the Supra AGS-10k is of a high-end design using toroid transformers with wide frequency range and low distortion, even at the lowest frequencies.

The chassis is made from extruded aluminium, which helps prevent any RFI pick-up.

Specification:
  • Input impedance: 10 kohms
  • Output impedance: 10 kohms
  • Ratio: 1:1
  • Frequency Range: 15 Hz - 22 kHz
  • Size: 53 (H) x 96 (W) x 89 (L) mm
  • Weight: 440 g
View attachment 90901
Sorry to hear Your GPU noise is back. I have tried a lot of things to reproduce noise through my laptop-> Supra 10 meter D90 MQA-> Krell preamp without success. Maybe it’s the laptop? I use an old Acer Timeline from 2006 with Nvidia GPU and Intel i3 CPU. Only modification is a swap to SSD disc. I use the laptop mainly as Tidal and Spotify player, some emailing and Outlook calendar. No gaming or movielooking. And it is 8 meters away from the D90 (hence the 10 m cable). It connects to internet through 2,4 GHz wi-fi. Dead quiet. I do not use ferrites. And I have Absolutely no interest whatsoever in the Supra company (could have choosen any brand shielded in 10 m length). I do have different current purifiers (DC blocker, phase corrector, iFi purifier and dedicated powerline 25Amp to my system). Those things I had before the D90 and Supra cable. I really hope Your problem will come to an solution Keep trying! D90 sounds marvellous!
 

mocenigo

Major Contributor
Joined
Dec 8, 2018
Messages
1,274
Likes
1,034
Are you going to use it on a desktop? If so, I like the rotary volume control on DX7 Pro better. If it is in a stereo system with remote controlling it, then I would go for D90. It will be better for bragging points on measurements. :).

Not to speak that its sound is simply spectacular. So may be many other DACs on the market, but at this price point...
 

mocenigo

Major Contributor
Joined
Dec 8, 2018
Messages
1,274
Likes
1,034
I am curious to understand the difference between the DAC and PRE mode. Apart from bypassing the volume control on the AKM chip, does the PRE mode also bypass (part of) the output buffer?
 

mocenigo

Major Contributor
Joined
Dec 8, 2018
Messages
1,274
Likes
1,034
How does that bypass anything? I've never really gotten why it would be an advantage to have it on DAC mode vs preamp at 100% volume. How are they different?

It is for audiophools that
1) Need to have "unadulterated" DSD in their chain, so they are happy to have a potentiometer instead, and
2) "Know" that attenuation in the digital domain will not only NOT be bit perfect, but it will also "chop off" precious bits, so they prefer to have a potentiometer instead.

Necessary marketing, they are catering to both measurement freaks and pure-subjective guys. What counts is that they have a good product at a reasonable price. Just use it in "pre" mode and you can have less active or passive electronic components between your DAC and the speakers, or your headphones, and avoid the MQA version.
 

mocenigo

Major Contributor
Joined
Dec 8, 2018
Messages
1,274
Likes
1,034
It should bypass whole vol control section in DAC mode, I think. So it would mean max out. But that's only my guess. I am still struggling with DX7pro vs D90+A90 decision.

If you have not decided yet, I prefer to go for separates.
 

Harmonie

Major Contributor
Forum Donor
Joined
May 30, 2020
Messages
1,927
Likes
2,084
Location
France
Very nice :)
I'm not sure what ferrite powder unformed into a toroid for magnetic effects does, perhaps it's simply a powdered iron shield? There are formed flat plates of ferrite designed for winding wire to effect as described in the "spoilers" and on the site I am quoting, perhaps the line of ferrite powder also acts in this way as distributed through twisted pair wires?

The Supra cable doesn't have ferrite cores or ferrite powder that I know of, certainly not externally, and relies on shielding between power and data along with using twisted pair wiring independent for power and data. And, that's why the chirpy noises aren't filtered out at all and are too noisy to allow the RCA cables to be connected.

The Supra cable does have positive audible improvements to the signal I am hearing over the Tripp Lite cable with 2 ferrite cores. IDK if that is due to the effects of cutting the signal - low pass filters have some loss at the desired frequencies too - as I've been told it's best to only use 1 ferrite core at one end, IDK which end it is supposed to be at but I assume the source end - as 2 cores tends to affect the audio negatively. So what I might be noticing is a degradation of the signal with the Tripp Lite with Ferrite cores vs the Supra Cable which has no filtering effect in the design.

Frankly ?
I use these for power strips and not in signal cables, so that it's not directly sound related.
I also liked to fiddle around, terminating my own cables.
On the other hand, it doesn't induce noise.
As I stated elsewhere, I hate noise:
Hiss in active speakers
AC/diode noise in "cheaper power bricks".
Maybe I'm a "noise oriented hunter" - but sincerely, I'm very sensitive and focused on such noise so that the music I'm listening to has to be free of it.
 

hmscott

Addicted to Fun and Learning
Joined
May 17, 2020
Messages
604
Likes
493
Frankly ?
I use these for power strips and not in signal cables, so that it's not directly sound related.
I also liked to fiddle around, terminating my own cables.
On the other hand, it doesn't induce noise.
As I stated elsewhere, I hate noise:
Hiss in active speakers
AC/diode noise in "cheaper power bricks".
Maybe I'm a "noise oriented hunter" - but sincerely, I'm very sensitive and focused on such noise so that the music I'm listening to has to be free of it.
We are simpatico, I also have a pure signal preference :)

But, in times of need I can pick a signal out from the noise easily after many years of experience doing so. But, when I can arrange it properly to produce a better clean signal I do appreciate that and prefer it - high noise content can be distracting early on until our brain automatically filters it out for us.

Exchanging tips and tricks like this is very helpful as we each cannot think of or experience everything others do.

Thank you for your helpful suggestions and input. :)
 

bogi

Active Member
Joined
Sep 16, 2020
Messages
238
Likes
170
Location
Slovakia
Are there any affordable USB isolators that work and would be worth trying? Either USB => USB, or USB to Optical, or USB to AES / I2S / ?

Can the PI2AES be used as a converter between USB and USB / Optical / AES / I2S, rather than be the end node / player / renderer?
I suggest you to buy such things only if they can be returned in case they don't (sufficiently) help.
My candidates are iDefender (to cancel ground loop) and iSilencer (ANC to cancel other type of computer noise than ground loop).
iDefender works best with additional power supply, iPower is recommended. Older and cheaepr versions of these products may be okay too. But take care about the return policy before you buy anything.
 

Veri

Master Contributor
Joined
Feb 6, 2018
Messages
9,596
Likes
12,036
I suggest you to buy such things only if they can be returned in case they don't (sufficiently) help.
My candidates are iDefender (to cancel ground loop) and iSilencer (ANC to cancel other type of computer noise than ground loop).
iDefender works best with additional power supply, iPower is recommended. Older and cheaepr versions of these products may be okay too. But take care about the return policy before you buy anything.
Sorry but this all sounds like you sure drank the iFi kool aid. The only one that can *fix* ground issues is the defender one which cuts power and takes it from an external source (a phone charger would work fine, the iPower is not at all necessary).
 

mocenigo

Major Contributor
Joined
Dec 8, 2018
Messages
1,274
Likes
1,034
Anybody noticed a very soft popping sound a few seconds after music reproduction has stopped?
 

mocenigo

Major Contributor
Joined
Dec 8, 2018
Messages
1,274
Likes
1,034
some are saying you can let your software player convert all files to DSD and it helps the DAC Chip with the processing.... the specs of the DAC chip do show some degradation at the highest of the DSD files...

Yep. Audiophoolery at its highest.
 
Top Bottom