Review and Measurements of Schiit WYRD USB Filter

[trl]

I wonder what chip is U6, perhaps the USB reclock?

I do like the 10A TO220 transistors, I guess these are responsable for the +5V power rail.

 

[trl]

[...]
However if you are using a desktop PC.... other than maybe a front USB port (which implementation can vary per case) I would say that motherboard USB ports are going to be generally clean.

I choose the front USB on purpose to compare it with the Wyrd now, thanks for noticing.

 

[Jimster480]

View attachment 21958


View attachment 21960

View attachment 21961

I wonder what chip is U6, perhaps the USB reclock?

I do like the 10A TO220 transistors, I guess these are responsable for the +5V power rail.

I cannot read any numbers on it so.... its just guesswork.
But considering that it gets a signal directly from the crystal there its definitely possible.

 

[trl]

Measuring the +5V through a 1" long USB plug/cable:

Desktop's front USB port noise
(pretty noisy)
​

Wyrd USB port noise
(identical with my scope probe shorted with GND)
​

Measuring the +5V rail directly on ODAC's power plug pins:

ODAC connected to desktop's front USB port noise, not playing
(200uS/div. scale instead of 20uS/div. from the pics above)

ODAC connected to Wyrd, not playing

ODAC connected to front USB port, playing 1 KHz sinewave

ODAC connected to Wyrd port, playing 1 KHz sinewave
​

Interesting noises appearing every 1ms; seems to be generated by the ODAC itself and are present in both measurements (computer's USB port and Wyrd).

Wyrd is indeed delivering an extremely low-noise and stable +5V voltage, unmeasurable with my scope (signal is identical with my scope's probe when short-circuited with ground), so it's internal shunt regulator is doing it's job perfectly!

Now it's up to anyone to use it wisely, either for powering a DAC/transport, either by powering a...RaspberryPi, or simply to use it in DYI projects as a reference +5V power supply.

L.E.: No crocodile clips were used in the above measurements, just few mm long spring/wire.

 

[mansr]

I wonder what chip is U6, perhaps the USB reclock?

Judging by the pin connections, it's an SMSC/Microchip USB2412 hub chip. $1.43 in bulk from Mouser.

 

[amirm]

Interesting noises appearing every 1ms; seems to be generated by the ODAC itself and are present in both measurements (computer's USB port and Wyrd).

That is the timing of USB frame so shows up in some DACs.

 

[gvl]

For $50 bucks a used Wyrd is a useful gadget to keep around, I keep 2

 

[Jimster480]

Measuring the +5V through a 1" long USB plug/cable:

View attachment 21964

Desktop's front USB port noise
(pretty noisy)
​

View attachment 21965
Wyrd USB port noise
(identical with my scope probe shorted with GND)
​

Measuring the +5V rail directly on ODAC's plug's pins:

View attachment 21967
ODAC connected to desktop's front USB port noise, not playing
(200uS/div. scale instead of 20uS/div. from the pics above)


View attachment 21970
ODAC connected to Wyrd, not playing


View attachment 21972
ODAC connected to front USB port, playing 1 KHz sinewave


View attachment 21973
ODAC connected to Wyrd port, playing 1 KHz sinewave
​

Interesting noises appearing every 1ms; seems to be generated by the ODAC itself and are present in both measurements (computer's USB port and Wyrd).

Wyrd is indeed delivering an extremely low-noise and stable +5V voltage, unmeasurable with my scope (signal is identical with my scope's probe when short-circuited with ground), so it's internal shunt regulator is doing it's job perfectly!

Now it's up to anyone to use it wisely, either for powering a DAC/transport, either by powering a...RaspberryPi, or simply to use it in DYI projects as a reference +5V power supply.

L.E.: No crocodile clips were used in the above measurements, just few mm long spring/wire.

What about your rear USB connections?

 

[trl]

I see no audio differences between front USB and Wyrd, I don't think there should be on the rear USB, right? Or am I missing the point perhaps?
My rear USB ports have been measured and posted in the Archimago's blog, per above link; I can measure them again if you want, but the AC ripple & noise was between the front USB and Wyrd. BTW, I found a very clean +5V: Lenovo 65W power adapter with USB hub.

 

[monkey4054]

Any merit to more expensive solutions like the IFI MICRO IUSB3.0 or IFI Nano IGALVANIC3.0 USB Galvanic Isolation etc.? Or are they just as pointless as the Wyrd when going from a PC to a competent DAC?

 

[Jimster480]

Any merit to more expensive solutions like the IFI MICRO IUSB3.0 or IFI Nano IGALVANIC3.0 USB Galvanic Isolation etc.? Or are they just as pointless as the Wyrd when going from a PC to a competent DAC?

It would be just as pointless for any competent DAC. Basically nothing that is actually good has a bad USB implementation to the point that it will benefit from a device like this one.

 

[monkey4054]

It would be just as pointless for any competent DAC. Basically nothing that is actually good has a bad USB implementation to the point that it will benefit from a device like this one.

Right, I thought so. So basically all these USB "de-crapifiers" and conditioners/filters etc. are snake oil essentially?

 

[trl]

Any merit to more expensive solutions like the IFI MICRO IUSB3.0 or IFI Nano IGALVANIC3.0 USB Galvanic Isolation etc.? Or are they just as pointless as the Wyrd when going from a PC to a competent DAC?

Theoretically, less noise coming from power input means lower background noise on DAC's output, but these days the internal power regulators from most DACs are quite good and spending money on a Wyrd or anything else similar might not actually improve anything at all.

However, in my case, I had connected PLAYMATE directly, then via Wyrd to one of my laptops and background noise decreased while using Wyrd with about 1-2 dB for frequencies above 5-6 KHz. However, ARTA provided the exact same numbers for THD+N, also my ears where unable to make a difference too, so...not quite an audible nor measurable improvement.

In iFi's case, if speaking about a galvanic isolation for input power, then you might give it a try if you can borrow one from a friend, then see if RMAA & ARTA are providing different numbers for SNR and THD+N. Theoretically, galvanic isolation will remove any possible ground loops, so in some cases might be an option, but not sure this would be your case or not.

P.S.: Given iFi's price is comparable with DX7s on Massdrop I don't think I can recommend it. Also, D50 DAC costs much less then iFi Galvanic and is a very competent DAC.

 

[trl]

Right, I thought so. So basically all these USB "de-crapifiers" and conditioners/filters etc. are snake oil essentially?

It really depends on what issue you're facing currently with your DAC & PC. Like I said, a ground loop might be removed with a "de-crapifier" device, but not sure anyone ever measured any difference before & after using a "de-crapifier". Like I said, most DACs have already built-in power regulators and sometimes rail splitters to accommodate power for the internal circuitry, so the power you inject from the USB port is filtered inside the DAC. Even with my old ODAC V1 I couldn't measure any differences, but maybe I'll give it another try on a noisy USB port from my PC (I think I already did that once, but I'll try it again).

 

[amirm]

Any merit to more expensive solutions like the IFI MICRO IUSB3.0 or IFI Nano IGALVANIC3.0 USB Galvanic Isolation etc.?

I have an iFI iUSB 3.0 coming for testing. For now, for well designed DACs there would be no benefit.

 

[Jimster480]

Theoretically, less noise coming from power input means lower background noise on DAC's output, but these days the internal power regulators from most DACs are quite good and spending money on a Wyrd or anything else similar might not actually improve anything at all.

However, in my case, I had connected PLAYMATE directly, then via Wyrd to one of my laptops and background noise decreased while using Wyrd with about 1-2 dB for frequencies above 5-6 KHz. However, ARTA provided the exact same numbers for THD+N, also my ears where unable to make a difference too, so...not quite an audible nor measurable improvement.

In iFi's case, if speaking about a galvanic isolation for input power, then you might give it a try if you can borrow one from a friend, then see if RMAA & ARTA are providing different numbers for SNR and THD+N. Theoretically, galvanic isolation will remove any possible ground loops, so in some cases might be an option, but not sure this is would be your case or not.

P.S.: Given iFi's price is comparable with DX7s on Massdrop I don't think I can recommend it. Also, D50 DAC worth much less then iFi Galvanic and is a very competent DAC.

yeah yeah I mean these isolators cost as much as my DAC.
I paid $270 for my DX7 when it just came out.

 

[trl]

Attaching 4 pics with ODAC without vs. with Wyrd, while being connected to a noisy front USB port. Unable to see a real difference, but feel free to check the attached pics. Two are -72dB and two are -105dB (dithered, so noise raises when 20 KHz approaches). I like that even with -105dB audio files there's still enough headroom till the background noise of ODAC.

 

[bequietjk]

To add to my last post, here is Jason Stoddard on head-fi with his brilliant marketing back in 2014 telling the story of how Wyrd came about:

View attachment 18945

Say one thing on the website, but another in a forum.

The above by the way is textbook invalid audio testing.

I find myself doing this a lot when i mess with my audio chain

 

[formdissolve]

One use case for this product is to allow for extended runs of USB 2.0 cable from your source. The spec for max USB 2.0 cable length is set at 16 feet (I've had luck with 20, but after that I get dropouts, but different cables may be better). I have a headless Volumio Raspberry PI in a different room and use a Wyrd that I got from Schiit direct for $49 plus shipping in between, and no longer have dropouts at nearly 40 feet distance. I'm sure any old cheap powered USB 2.0 hub would work just as well, but this product found life in this use case! I'm going to be redoing my audio setup so it probably won't get used anymore, but for now it's great.

 

[trl]

Makes lot of sense in using Wyrd as a repeater, especially that an active USB cable starts at 70 USD. After all, it will act as an USB hub to regenerate the digital signals, but will also provide an extremely clean +5V output for the guest equipment.