• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). Come here to have fun, be ready to be teased and not take online life too seriously. We now measure and review equipment for free! Click here for details.

Intona USB 3.0 Isolator Review

Final

Member
Joined
Mar 19, 2018
Messages
41
Likes
50
Thought I'd share my personal experience:). I had a serious grounding issue with my old Asus PC. Heavily distorted sound when streaming from the USB output into an Accuphase DAC 40. When connecting headphones directly to the sound output of the PC I could hear the same issue. I tried an IFI micro iUSB.30 to no avail. The Intona 2.0 did the trick. The noise was gone, totally silent. Therefore I am a bit puzzled that Amir had trouble recreating such a hostile invironnment. In my case I also ended up investigating further. I bought a new power supply from Asus and the Intona was no longer needed. An expensive product Intona, but ' it may come in handy in some cases.
 

Veri

Master Contributor
Joined
Feb 6, 2018
Messages
7,911
Likes
9,574
Therefore I am a bit puzzled that Amir had trouble recreating such a hostile environment. In my case I also ended up investigating further. I bought a new power supply from Asus and the Intona was no longer needed. An expensive product Intona, but ' it may come in handy in some cases.

Funnily enough noise is typically there when you don't need it, but when you want to reproduce it for testing it's suddenly difficult to find. Lol.
 

laidick

Active Member
Forum Donor
Joined
Apr 22, 2018
Messages
148
Likes
96
Funnily enough noise is typically there when you don't need it, but when you want to reproduce it for testing it's suddenly difficult to find. Lol.
Very similar to software bug in production environment... Very hard to reproduce the same in test environment too. Lol
 

solderdude

Master Contributor
Joined
Jul 21, 2018
Messages
9,780
Likes
21,126
Location
The Neverlands
I am a bit puzzled that Amir had trouble recreating such a hostile invironnment.

It is not difficult to test for this but it requires different test equipment (which is fairly expensive) intended for EMC testing as well as special couplers/injectors.
Amir tests for audioperformance and not for EMC aspects.
 

KSTR

Major Contributor
Joined
Sep 6, 2018
Messages
1,440
Likes
2,985
Location
Berlin, Germany
Creating, for example, a "ground loop" lab test scenario is actually extremely easy, with an USB isolator in place. All you need is a (somewhat potent) function generator running off of batteries or hooked at a mains isolation transformer. In Amir's case, just use the transformer-isolated analog output of the AP to inject balancing current on a DUT and its associated cables (USB in, RCA/XLR out).
By this, one can inject shield current on any part of the chain, notably the DAC under test itself to test how good its internal layout is.
 
Last edited:

mansr

Major Contributor
Joined
Oct 5, 2018
Messages
3,727
Likes
8,410
Location
Hampshire
It is not difficult to test for this but it requires different test equipment (which is fairly expensive) intended for EMC testing as well as special couplers/injectors.
Amir tests for audioperformance and not for EMC aspects.
Accurately quantifying the performance of such a device is indeed non-trivial. Creating a ground loop with some random noise, enough to tell whether the device does something, doesn't usually take much effort.
 

solderdude

Master Contributor
Joined
Jul 21, 2018
Messages
9,780
Likes
21,126
Location
The Neverlands
Yes, one can rig up a simple test and inject some signal to see if something happens at the analog out.
However, this only says something about the fact that there is some influence within a certain range.
It's nothing 'standard' though and also HF sweeps might be in order plus it would be nice if measurements were adhering to a standard of sorts.
For this, to really characterize and be able to compare calibrated results you need different test equipment to get meaningful results.

I find it similar to test if an amp is on by touching RCA input pins and check if you hear a hum.
It tells you the amp is on and functioning but not much specific.
I don't think that's what ASR (Amirs measurements) is all about.
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
34,989
Likes
130,333
Location
Seattle Area
I can certainly make a huge project out of this and make it happen. But that is not the goal of the review. Most people who buy these boxes think they improve sound fidelity even when they don't have any audible noise. For the core functionality of the device, I am sure it does what it says it does so doing a ton of work to confirm that is out of scope.
 

mansr

Major Contributor
Joined
Oct 5, 2018
Messages
3,727
Likes
8,410
Location
Hampshire
Yes, one can rig up a simple test and inject some signal to see if something happens at the analog out.
In my experience, no rigging or injection is required. If I connect a USB DAC to my desktop computer with the analogue output going to an ADC connected by USB to a laptop on AC power, I get lots of noise, like this:
1583435085437.png


With the laptop running off battery, it looks like this:
1583435152957.png


If an isolation device had the same effect, I'd say it was working.
 

solderdude

Master Contributor
Joined
Jul 21, 2018
Messages
9,780
Likes
21,126
Location
The Neverlands
Is it doubtful that the isolator wouldn't work as expected ?
I would think it would be very easy to obtain an EMC report from the manufacturer (if one knows how to read and interpret).

I see your point of course and would agree that as an extra test it could be relatively easy to add a test where a certain (fixed level) signal was injected to see how much of it would be found in the analog output.

Haven't seen any other reviewers test for this though. In some cases a groundloop can be very audible and annoying (and sometimes not easy to get rid of when one doesn't understand the 'path' the crap takes)
 

DDF

Addicted to Fun and Learning
Joined
Dec 31, 2018
Messages
565
Likes
1,069
If an isolation device had the same effect, I'd say it was working.

Switching the laptop to a class 2 power supply without earth ground (ie 2 prong, not 3) breaks the ground loop from the laptop AC and gets rid of the ground noise much more cheaply. $40 (Canadian, probably much cheaper US) will buy you a good brand 3rd party class2 laptop supply.
 

mansr

Major Contributor
Joined
Oct 5, 2018
Messages
3,727
Likes
8,410
Location
Hampshire
Switching the laptop to a class 2 power supply without earth ground (ie 2 prong, not 3) breaks the ground loop from the laptop AC and gets rid of the ground noise much more cheaply. $40 (Canadian, probably much cheaper US) will buy you a good brand 3rd party class2 laptop supply.
Sure, that particular setup is easier to fix by other means. It's still a useful test case.
 

DDF

Addicted to Fun and Learning
Joined
Dec 31, 2018
Messages
565
Likes
1,069
Sure, that particular setup is easier to fix by other means. It's still a useful test case.

Agree with the test case.
Posted the Class2 PSU info to hopefully save some people considering this $300.
 

KSTR

Major Contributor
Joined
Sep 6, 2018
Messages
1,440
Likes
2,985
Location
Berlin, Germany
Switching the laptop to a class 2 power supply without earth ground (ie 2 prong, not 3) breaks the ground loop from the laptop AC and gets rid of the ground noise much more cheaply. $40 (Canadian, probably much cheaper US) will buy you a good brand 3rd party class2 laptop supply.
While a 2-prong laptop SMPS brick will break the proverbial 50/60Hz "ground loop" (when both grounds, of the laptop and the DAC+subsequent gear, are bonded to slightly different PE potentials) it will still inject a lot of nasty balancing current because the brick's output ground is connected to the rectified mains via a 2.2nF EMC cap. One can easily end up with less hum but more buzz, notably if the DAC etc still have GND at PE potential. With longer RCA cable runs, a class-II power supply may not be sufficient, it just replaces one noise mechanism by another one which often is less benign. Actual testing in-situ is required as one cannot be sure whether there will be an actual improvement or not just by going from 3-prong to 2-prong.
 

DDF

Addicted to Fun and Learning
Joined
Dec 31, 2018
Messages
565
Likes
1,069
While a 2-prong laptop SMPS brick will break the proverbial 50/60Hz "ground loop" (when both grounds, of the laptop and the DAC+subsequent gear, are bonded to slightly different PE potentials) it will still inject a lot of nasty balancing current because the brick's output ground is connected to the rectified mains via a 2.2nF EMC cap. One can easily end up with less hum but more buzz, notably if the DAC etc still have GND at PE potential. With longer RCA cable runs, a class-II power supply may not be sufficient, it just replaces one noise mechanism by another one which often is less benign. Actual testing in-situ is required as one cannot be sure whether there will be an actual improvement or not just by going from 3-prong to 2-prong.

Class 2 breaks connecting DAC ground through laptop USB ground back to laptop earth ground. My set up (laptop->USB dac) had all sorts of buzz, laptop process dependent noise, time variant high frequency noise currents. No 50/60 Hz hum. Changing to class2 brick on the laptop completely rid the system of all noise, and measured noise is extremely low.

Class2 brick is a very viable solution and I'd hate for others to forgo the option and spend $300 on an Intona when $25 brick could very well do.
 

KSTR

Major Contributor
Joined
Sep 6, 2018
Messages
1,440
Likes
2,985
Location
Berlin, Germany
Class2 brick is a very viable solution and I'd hate for others to forgo the option and spend $300 on an Intona when $25 brick could very well do.
That's why I said test before buy. If a new supply fixes it, all is well ;-)
 

DDF

Addicted to Fun and Learning
Joined
Dec 31, 2018
Messages
565
Likes
1,069
That's why I said test before buy. If a new supply fixes it, all is well ;-)

My motivation for replying is I think your comment could create a false sense of risk about class2 supplies, and for the less initiated, could drive them towards costly less effective band aids instead.

Any brick design could be poorly executed and couple HF noise or EMI onto the secondary DC. A poor design isn't a special feature of class2 bricks. I agree, don't buy buy a poor, noisy PSU and if so unlucky, return it. Nothing uniquely applicable to class2 bricks in that risk assessment.

There's a great deal of focus in the audio community on these USB gizmos and almost no discussion about the significant tangible benefits low cost, safe, proven class2 bricks bring to audio PCs, which effectively solves the problem at the root cause (not try and clean it up later).
 

KSTR

Major Contributor
Joined
Sep 6, 2018
Messages
1,440
Likes
2,985
Location
Berlin, Germany
Any brick design could be poorly executed and couple HF noise or EMI onto the secondary DC. A poor design isn't a special feature of class2 bricks. I agree, don't buy buy a poor, noisy PSU and if so unlucky, return it. Nothing uniquely applicable to class2 bricks in that risk assessment.
I see you didn't understand what I was talking about. It's not the quality of the DC output (where you're comment is fully correct), it's about the common-mode mains leakage current, the mains coupling.

With a class-II SMPS we have the mentioned rather large coupling capacitor directly accross the transformer plus the fact that the coupling isn't going to L or N, rather to a rectified waveform wrt PE. This is were the difference is to traditional transformer class-II supplies. While those also have inter-winding capacitance (typically way less than than 1nF except big toroids), they couple to some voltage between L and N directly and hence the leakage current is mainly 50/60Hz.

With a 3-prong design, the coupling typically is a direct connection to PE and the mains leakage current is thereby effectively zero with reference to that PE. Note, though, class-I doesn't require the output to be bonded to PE, it's all about the input side. The output side can be quasi-floating, with only a capacitive connection to PE. A capacitive connection to PE is much better than a capacitive connection to rectified mains and is the preferred solution as there is only a weak path for 50Hz/60Hz whereas most of the buzzing frequencies are shunted to PE.

https://www.mouser.com/catalog/additional/SL Power_TE_Externals_General_Overview_AppNote.pdf
 

DDF

Addicted to Fun and Learning
Joined
Dec 31, 2018
Messages
565
Likes
1,069
I see you didn't understand what I was talking about.

Really? As an EE "because the brick's output ground is connected to the rectified mains via a 2.2nF EMC cap " isn't too difficult a concept.

My point was that you're significantly overstating this risk and ignoring the far greater benefits.

I thought there would be general interest in potentially saving over $300, but judging by the response, maybe not. Lesson learned.
 
Top Bottom