Just finished a little experiment that I hope some will find helpful If considering using it with a phone. Plugging the device straight into the phone is definitely going to result in audible interference from the cellular radio when it is running (not typically an issue when cellular is off and phone is only using WiFi), as Woo themselves and others here have pointed out. This is separate from the hiss, which is present at a low to moderate level in all of the following conditions. I did this test with my Focal Celestee and a 4.4 mm Apos Flow 5 ft. cable.
1) I connected to my Raspberry Pi 4B that I use as a Roon endpoint with Ropieee. Hiss is there, but no significant interference the sounds like wireless digital transmission.
2) I connected to my Raspberry Pi 4B with a Topping HS02 galvanic isolator in the middle - no difference from 1.
3) I connected to my iPhone 15 Pro Max directly through USB-C - extremely obvious and bothersome cellular radio noise.
4) I connected to my iPhone through the Topping HS02 with the Tube Mini at the same proximity on the table - cellular radio noise greatly attenuated.
5) While still connected to iPhone through the HS02, I moved the phone over the Tube Mini and the noise increased, but not to anything like the degree present when plugged into the phone.
6) I moved the Tube Mini and HS02 as far from the phone as possible and had almost full attenuation of cellular noise while browsing, streaming, and reloading websites.
7) Back on the Raspberry Pi, I moved the phone over the Tube Mini though it was not physically connected and loaded some websites - some cellular radio noise was present, much less than when plugged directly into the phone.
This has led me to conclude that the cellular radio noise enters the Tube Mini by 2 mechanisms.
1) Through direct connection with the phone, presumably through the power transmission through USB-C as opposed to the data, as the HS02 greatly reduced the severity of the sound. Just trying to think of other differences - there was more total wire between the two with the HS02 attached, but I don’t think this was the reason for improvement. Proximity was not a factor here as the phone and Mini were kept at the same physical distance on the table.
2) Through RF interference transmitted through the air. I believe that this is a smaller part of the overall issue than number 1, but it is definitely a factor. It seemed that I needed to increase the distance from the Mini as well as from the HS02 - as keeping the phone close to the HS02 still seemed to result in some noticeable interference. When both devices were moved far enough away, I did not notice any interference (while still connected to the phone through the HS02)
_______
I did an additional test while I had the items handy and plugged into my iPad Pro 6th gen w/ cellular via USB-C. Once again the galvanic isolator helped to reduce the cellular interference, however, it was not able to completely stop it, unlike with the iPhone. This was true even when the tube mini and HS02 were plugged into a 6 ft cord and kept as far back as possible. Also notable, I do believe the bass response for really big bass drum hits / loud sub-bass synths is a tiny bit better on the iPad. Someone had mentioned above that it may have a higher allowable power output via USB than the phone, but the difference is pretty small to my ears.
_______
***EDIT*** I picked up an in-line USB power tester and compared the scenarios with extra power available and when connecting to different devices. There was no difference in any scenario when the Tube Mini was powering my HD600 through a balanced connection. I tried connecting to USB A and C ports on Raspberry Pi, with and without auxillary power and in all cases the Tube Mini was drawing 0.118-0.130 A at 5 V which puts its power consumption around 0.65 W. This is well under the 0.500 A that iPhone can output, so I now really doubt that this is making any difference. Placebo effect is a powerful force, and I think that may be what I was hearing below. I will leave my initial impressions, but I am no longer confident that extra power is beneficial. ***END EDIT***
Just as I was about to hit post on this, one more test idea hit me. I remembered seeing in the manual for the HS02 that the port that is not being used for data could be used as a power source, so I got a 5V / 2.1A wall wart and dug out a high quality shielded USB A to B cable. I hooked up the HS02 via USB-C to the iPhone and via USB-C out to the included adapter to the Tube Mini. It stayed powered even when the phone was unplugged in this configuration, and once the phone was connected, I could hot plug and unplug the USB-B power connection with no interruption in playback, which allowed for immediate A/B testing. On the track 15 Step from In Rainbows by Radiohead, the bass response was improved when connected to additional power. On Starboy by The Weeknd, I noticed improved slam / intensity of bass and less distortion of vocals over the repeated sub-bass hits. My theory is that, at least with the relatively easy to drive Celestee, it is clipping ever so slightly off of phone power alone, barely noticeable, but full extension is possible with the additional power connected. Maybe this would be more noticeable on a harder to drive headphone.
Sliding a little deeper into the rabbit hole, just to be totally absurd, I dug out a BRZHiFi 5V/25W linear power supply that I had picked up on eBay for my Pro-Ject Pre Box S2 Digital before I sold it. When I hooked this up instead of the wall wart for the additional power, I heard… no discernible difference at all from the wall wart providing the extra power, but hey, there’s nothing wrong with being a little over the top in this hobby…
_______
One last little bit of info I gathered. When hooking up the Raspberry Pi to the HS02 again via USB-A to B, and this time connecting the USB-C input on the HS02 to the Zendure Supertank Pro via the second USB-C port on that device. The Supertank Pro’s display shows a 0.2 W power draw when the Tube Mini is not connected, and when it is connected it shows a draw of 1.3-1.75 W, variable with volume (with 1.75 appearing only at obviously unsafe listening levels). This would confirm that the HS02 is, in fact, using the input that is not being used for signal to receive some power. Subtracting it out, this would suggest that the Tube Mini may be drawing as much as 1.1-1.55 W. I’m not sure if this wattage is reported relative to the 5V being requested by the Tube Mini or an internal higher voltage on the power bank, but a reasonability check would suggest not, as the Tube Mini is supposed to output 1W into 8 Ohms, so with the only slightly higher impedance Celestee, it would make sense that it would be putting out a little less than a Watt and chalk the difference up to the energy used by the DAC and losses due to efficiency. Also confirmed same power use amount / behavior by connecting the iPhone through a USB C to A adapter and a USB A to B cable into HS02 and USB C from Supertank to HS02, so not using more additional power when connected this way than when connected to Raspberry Pi’s USB port.
Interestingly, it would appear that this power is supplementary, as I had a very long (15 or 20 feet) USB A to B cable with resistors that I had previously used to try to troubleshoot a problem connecting a device to my computer. When I attempted to play music by connecting this to the Raspberry Pi and the HS02, the music was extremely grainy and muffled (not a subtle audiophile effect, obviously not functioning correctly), and this did not improve with connecting the extra power (though the Supertank still showed a 1.6 W output, so maybe it was an issue of data transmission). It was fixed by switching to the short USB A-B cable included by Topping with the device.
________
Overall, I think that using the Tube Mini with a phone with cellular enabled is not ideal, however, if you are pulling data through WiFi or are listening to locally stored songs, you are only likely to notice interference when receiving a call or send / receiving text messages (or using cellular data for browsing, etc. if listening off of your internal storage). The power may be slightly limiting, and this may be more of a factor on headphones that are less sensitive than the Celestee.
This was done pretty quickly, and it is possible that there may be other variables I didn’t consider or some interference that I just didn’t hear. In other words, YMMV, but I hope that this info is helpful for anyone who may have a use case where a device with cellular connectivity is all that is available and it is not practical to go into airplane mode. If you are going to attempt this and you are able, I would strongly suggest adding a Topping HS02 and a 6 foot USB cable to your kit to allow for electrical and physical separation respectively, and if you want to max out quality, use a charger connected to the unused input on the HS02 to make up for the difference in power if using an under-powered output device like an iPhone.