• 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!

Singxer UIP-1 USB Isolator Review

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,563
Likes
238,997
Location
Seattle Area
This is a review and detailed measurements of the Singxer UIP-1 USB Isolator and "Purifier." It is on kind loan from a member. The UIP-1 costs fair bit at US $259.

The box while utilitarian is a step above budget category:

Singxer UIP-1 – USB isolation Purifier Audio Review.jpg

The connectivity is simple enough:
Singxer UIP-1 – USB isolation Purifier Connectors Audio Review.jpg

For operation the UIP-1 needs an external power supply as you can see from the connector on the left. Strangely, none came with it! I am assuming they are thinking (perhaps correctly) that the market for this is audiophiles would be using fancy linear power supplies and such. For my testing, I used my Keysight/Agilent lab power supply at 7 volts.

The guts show an off-the-shelf USB isolator:

UIP1-3.jpg


The white band is a required design practice to fully isolate the "dirty" and clean sides of the USB bus. The number (maxim 4930?) seems to be rubbed off however, as is the case with the IC to the left that is taking data from it. All in all looks like $30 worth of parts.

USB Isolator Audio Measurements
Recently I tested the Meizu Hifi USB-C dongle which showed high sensitivity to USB bus noise. As much as I hate acknowledging anyone knows more than me about testing :), the owner made a good suggestion of testing that DAC with the Singxer so that is what I did. Here is our dashboard with straight USB Connection:

Meizu Hifi dongle audio measurements.png


If you look to the right, we see a sea of spikes which collectively knocked down the SINAD by good many dBs as distortions were below -120 dB. If we route the USB connection through the Singxser UIP-1 we get this improved picture:

Singxer UIP-1 – USB isolation Purifier Audio Measurements.png


There are still noise products but now their levels are below that of the harmonic distortion. Our SINAD therefore improved by a significant 8 dB. It would have improved more had the UIP-1 not inserted its own mains noise on the left of our 1 kHz tone. Those peaks are now higher than the harmonic distortion so pull down SINAD by themselves.

Fortunately our hearing is not very sensitive to low frequency noise so the trade off seems good. Also note that despite my Keysight power supply being made for precision electronics work, in the past I have found that it leaks some amount of mains noise. So a bit of that may be due to the power supply. One thing you can't escape though: anytime you add another power supply to the chain, you increase your chances of getting new ground loops or mains leakage you did not have before. This is why there is almost no mains noise when I tested Meizu Hifi dongle by itself above.

For a better look at the spectrum I ran the jitter test:

Singxer UIP-1 – USB isolation Purifier Jitter Audio Measurements.png


We see what we already know. The UIP-1 reduces the noise floor (blue) and removes fair bit of spikes. But also introduces its own as higher spikes to the left.

Listening Tests
Ever since I received my new custom built gaming PC for my main workstation, the DX3 Pro which I use as daily DAC and headphone amp, has been produces pretty bad noises when no music is playing. It got so bad that I disconnected the USB and went with Toslink optical from the motherboard. That took care of the problem nicely. I thought I could test the UIP-1 and see if it could eliminate the USB noise. Alas, no success because the noise has completely vanished! Such is the vagaries of mains leakage, ground loops and such.

I listened anyway and there is no audible difference whatsoever. So please don't waste your money on such gadgets if you don't have clearly audible issues.

Conclusions
Many of these tweaks and USB regenerators do nothing. I am glad the Singxer does work and does reduce USB noise. Such functionality would not be needed of course with any quality USB DAC as even $99 ones produce exceptional performance using identical USB (noisy) connections from my PC. Should you have a nice DAC but still run into noise problems as I was with my DX3 Pro, and you can trial one of these boxes, I would be fine with it. Just don't tell me it improved the sound if you were not hearing any noises before! :)

So no recommendation but no opposition either to Signxer UIP-1 if used in proper application.

------------
As always, questions, comments, recommendations, etc. are welcome.

My muscles are aching from lifting and carrying all the speakers I am testing these days. They say a nice warm, daily massage may do wonders for it. Need some money to hire private sessions so please help me maintain my health by donating using: https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/
 

Xyrium

Addicted to Fun and Learning
Forum Donor
Joined
Aug 3, 2018
Messages
574
Likes
493
I like Maxim...

So, is this to famous galvanic isolation folks speak of?
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,563
Likes
238,997
Location
Seattle Area
I like Maxim...

So, is this to famous galvanic isolation folks speak of?
Yes. There is a resistor that cross-connects the ground from left to right side so it is not a perfect isolator.
 

urfaust

Active Member
Joined
Sep 25, 2018
Messages
113
Likes
59
Location
France
Interesting, so in the case of dongles being size limited, they don't have the circuitry or just harder time to do proper basic filtering?
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,563
Likes
238,997
Location
Seattle Area
Interesting, so in the case of dongles being size limited, they don't have the circuitry or just harder time to do proper basic filtering?
Correct but may also be poor layout/grounding.
 

Tks

Major Contributor
Joined
Apr 1, 2019
Messages
3,221
Likes
5,496
The folks who came up with Toslink were geniuses. Shame the standard never really saw much updates in the consumer sphere (I know it still has serious use in holding the network of the internet together).
 

bequietjk

Senior Member
Joined
Jan 4, 2019
Messages
462
Likes
474
Would LOVE to see the Singxer F-1's measurements. Always great to see Singxer up in here :)
 

Mittomen

Member
Joined
Nov 14, 2018
Messages
29
Likes
31
Alas, no success because the noise has completely vanished!
How come??? Can you please do your woodoo on my two old PCs? Both have dirty usbs, while the "Work" MacBookPro's usbs are dead silent. Since the dac is a Sabaj Da-3, it'd be silly to filter it with a box which costs double then the dac itself. Using balanced out (floating ground) solves usb noise issues.
 

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
15,981
Likes
36,175
Location
The Neitherlands
groundloops can be broken in several ways.
I am sure other (real) ground loop breakers will do the same.
The added mains hum may also have been another groundloop.
It really is only needed when there are groundloop issues.

How would the Meizu dongle react to say a wyrd (which does not isolate but only reclock and add clean 5V).
This way one can test if the power supply is the culprit of the noise or a groundloop.

Good to see testing these kind of de-crapifiers is done in a more logical/practical way.
 

MC_RME

Addicted to Fun and Learning
Technical Expert
Audio Company
Joined
May 15, 2019
Messages
870
Likes
3,607
Two theories:

- the isolating chip should be the Silanna ICE08USBB, used since years in various products, and well-known for its compatibility problems.

- these problems can be fixed by adding a USB hub chip on the device side, which most probably is the one to the left in the photo.

No theory:

- which then occupies USB resources. The Intona isolators are the only ones existing that work like a straight cable/wire, no Silanna and hub in there.

- in real-world operation customers should not be forced to use an additional power supply. It is possible to transfer the B-port (computer side) power to the A-port (the device), clean and galvanically isolated. But not full power, usually below 500 mA. But that is more than enough for many USB bus powered devices, and already overkill for devices which just need 5 V for USB recognition. Not seeing this here is disappointing.

Amirm wrote: There is a resistor that cross-connects the ground from left to right side so it is not a perfect isolator.

On the photo I see no resistor, just two unpopulated footprints? Such a resistor would make the whole unit senseless. Same goes for capacitive coupling, it must be in the picofarad range to prevent coupling of unwanted noise. This capacitive coupling is usually part of the isolator chip itself, plus the galvanically isolated power supply (if there is one, missing here).
 
Last edited:

Veri

Master Contributor
Joined
Feb 6, 2018
Messages
9,597
Likes
12,039
It's interesting in a way that, with some USB filtering it reaches 108dB SINAD, more than the ~101dB SINAD of the Meizu PRO :)

Thanks, @amirm
 

ZolaIII

Major Contributor
Joined
Jul 28, 2019
Messages
4,152
Likes
2,445
It's interesting in a way that, with some USB filtering it reaches 108dB SINAD, more than the ~101dB SINAD of the Meizu PRO :)

Thanks, @amirm
Well it would helped the Meizu Pro too and probably equally. After all it's the same design and not a very good one as shown.
 

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
15,981
Likes
36,175
Location
The Neitherlands
On the photo I see no resistor, just two unpopulated footprints? Such a resistor would make the whole unit senseless. Same goes for capacitive coupling, it must be in the picofarad range to prevent coupling of unwanted noise. This capacitive coupling is usually part of the isolator chip itself, plus the galvanically isolated power supply (if there is one, missing here).

Amir can check this with a multimeter. Maybe the screens (being connected to ground planes via the usual resistor and parallel capacitor) can still make a connection somehow (via casing or other path ?

No idea... as said a multimeter can show the resistance to be there or not via another path than the separate planes.
 

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
15,981
Likes
36,175
Location
The Neitherlands

ZolaIII

Major Contributor
Joined
Jul 28, 2019
Messages
4,152
Likes
2,445
Looks like the 'purifiers' Amir tested recently. I assume it does not isolate the ground connection so most likely does nothing against groundloops. The tested device here does.
It has two rows of resistors + two triodal tiristors, multi lv capacitors, 5W signal relay & coil to it. It's rather complex full design with quality components used, even PCB is high quality multi lv one. Unfortunately USB port used doesn't look anything extraordinary. Hire estimated cost of components alone is around 30$.
https://www.audiophonics.fr/img/cms/Images/Produits/0-10K/9056/9056-elfidelity-axf100pro-inpage2.jpg
 
Top Bottom