• 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.

USB high-speed isolators comparison tests

Joined
Aug 30, 2018
Messages
9
Likes
0
#1
Will we have a test comparing several USB high speed isolators that includes more affordable units?

The only one I could find is the one comparing the Intona to the Regen, and neither is quite affordable, in the sense that they cost close to a Topping D50.

As there have been issues about low frequency distortions in the D50 usb XMOS implementation, probably related to that interface between XMOS and 9038, which the Intona seems to cure, I wonder about other isolators, like the one from HiFimeDIY. for instance. Do they improve on that too or not?

There might be others around that I do not know about

I'm sure many people might be interested on such test, instead of knowing about new DACs.
 

solderdude

Major Contributor
Joined
Jul 21, 2018
Messages
2,162
Likes
3,484
Location
The Neverlands
#2
https://hifimediy.com/high-speed-usb-isolator-480Mbps < linky to the HiFime

There is also a cheaper version (to max 24/96) 12Mbps version

would be interesting to test but difficult to show the possible effect of breaking a ground loop. Amir would have to create a groundloop first and the effect it could have is highly DAC and situation dependent.
At least Amir could confirm it does not make things worse.
 
Joined
Aug 30, 2018
Messages
9
Likes
0
#3
My test would be connecting all isolators in the same way, as he did with the Intona and the Regen.

And I am interested in full DSD resolution, up to 768KHz. For 24/96 I stick with coaxial.

Something to help with units like the Topping D50 and still keep costs down.
 

solderdude

Major Contributor
Joined
Jul 21, 2018
Messages
2,162
Likes
3,484
Location
The Neverlands
#4
What would be the point of testing isolators and not test the ability to isolate ?
Sure there will be galvanic separation but one also would like to measure/test how much LF and HF injected voltages/currents are attenuated.

The regen is not an isolator, the intona is.
I expect both to be 'transparant' for USB as would the HiFime be as well.

What the intona and hifime do and the regen does not is isolate.
When one wants to test the isolation a groundloop has to exist otherwise it is merely a test to see if USB signals pass (which they probably will)

Isolators will only 'help' DACs when they actually break a groundloop and one gets rid of weird noises in the audio this way.
Therefore it is pointless to test without such a groundloop IMO and one needs a DAC that is affected by groundloops in the first place to show it does what it is designed for.
When you just want to reclock and get a cleaner +5V there are other solutions (which don't do much in most cases).
 
Last edited:
Joined
Aug 30, 2018
Messages
9
Likes
0
#5
Was the groundloop broken for the Intona test?

I would like to see what happens with the low frequency distortion spikes using other isolators, like the HiFime. That much I would be ready to pay.

And I'm curious about this one too:

https://electronics-shop.dk/?id=1038&google=1

Just 34 euro. It seems to be an isolator.
 

solderdude

Major Contributor
Joined
Jul 21, 2018
Messages
2,162
Likes
3,484
Location
The Neverlands
#6
It's 12Mbps only... you would be using the coax if I'm not mistaken ?

I am quite certain the intona was not specifically tested for groundloop breakage but only Amir knows for sure.
It was tested in the same vain as the other 'USB improvers', just to see if it adds extra noise or removes some.

To test what it actually is designed for (actual isolation) is another thing.

IME the isolators using small isolation transformers can break AF groundloops but do little to nothing for HF garbage. These only attenuate HF by just a few dB.
The ones using optical isolators inside are better at this but generally have more jitter. Not that such matters for modern USB DACs. That is IF they also have excellent 5V isolators as well which usually is done with small transformers (switchers) but may need extra care for HF filtering.
 

Veri

Major Contributor
Patreon Donor
Joined
Feb 6, 2018
Messages
2,087
Likes
1,893
#7
It's 12Mbps only... you would be using the coax if I'm not mistaken ?

I am quite certain the intona was not specifically tested for groundloop breakage but only Amir knows for sure.
It was tested in the same vain as the other 'USB improvers', just to see if it adds extra noise or removes some.

To test what it actually is designed for (actual isolation) is another thing.

IME the isolators using small isolation transformers can break AF groundloops but do little to nothing for HF garbage. These only attenuate HF by just a few dB.
The ones using optical isolators inside are better at this but generally have more jitter. Not that such matters for modern USB DACs. That is IF they also have excellent 5V isolators as well which usually is done with small transformers (switchers) but may need extra care for HF filtering.
Supposedly the Intona breaks ground loops ('galvanic isolation').
 

solderdude

Major Contributor
Joined
Jul 21, 2018
Messages
2,162
Likes
3,484
Location
The Neverlands
#8
It will break ground loops so does the hifime device, the Regen won't.

The problem is that one can measure it (no resistance measured between input and output USB sockets) but when connected to a proper setup where there are no ground loop issues the device 'does' nothing. In the sense that it just transports the USB signal.
There will be a slightly different background noise spectrum most likely or drown in the noise when properly designed.

What the hifime and intona add here (isolation) is not needed in a normal setup.
In setups where there is a groundloop and the DAC or connected amp in question (or the PC) have currents running through the USB connection and via the groundplane(s) into the audioline such a current COULD (depends on how the wiring/PCB inside the devices are made) become audible.
This could be currents drawn from the USB device or the PC or coming or going to the analog part (via the enclosure and safety ground or mains filters).
In order to test devices like these isolators one would have to create a potentially unwanted situation like applying a 100Hz tone between the analog input RCA screen on the amp and protective earth and then see what is measured with and without the USB isolator in the chain for instance.
How much effect this will have (what shows up in the analog signal) thus depends on the DAC used, audio cables used and amp used and how they are fed. It thus differs per setup.
A D10 can be perfectly quiet in one setup and prove problematic in another.

That is why I think isolators need to be tested for proper data transmission and for ground loop isolation with a repeatable and calibrate-able setup.
When doing EMC testing the way it is done is to inject a current via either a capacitor and resistor network (differential mode) or via a clamp (common mode) and then measure on the USB input cable and output cable (depending on where the inductance takes place.
This requires dedicated equipment (that makes sweeps) with predetermined and monitored currents/voltages and knowledge as well as grounded test site with the equipment under test placed at a certain distance from ground.
Difficult unless you have an EMC lab.
 
Top Bottom