USB 3.0 Radio Frequency Interference on 2.4 GHz Devices (2012) | Hacker News Reader

USB 3.0 Radio Frequency Interference on 2.4 GHz Devices (2012)

intel.com

144 pointsmarcopolis51 comments

TLDR; USB3.0 high speed signals when not properly shielded can cause RF interference with RF devices operating under 5Ghz due to the broadband nature of the signal.

@Mods, could be titled with [2012]

I have one USB3 128GB flash drive—manufactured right on the cusp of single-package 128GB NAND becoming available—where whenever I plug it into any device, it slow's that device's wi-fi to a crawl, and multiplies tenfold the jitter for any connected Bluetooth input peripherals. None of my other USB3 devices do it. I always figured it was a grounding problem or something.

put tinfoil around it.

So, if you put tinfoil around it and attach the tinfoil to a data line, does that make it more disruptive?

Sounds like a great device to keep around. A Wifi jammer unregulated by the FCC, along with plausible deniability for using it!

. . . Would you be interested in selling it? :P

I have had this annoying problem and had no idea what was going on until I stumbled upon forums somewhere after many Google searches. My Anker USB 3.0 hub would cause my bluetooth keyboard/mouse receivers to become unresponsive at random times (in retrospect at times when the hub was placed close to the Macbook Pro). I placed the hub at some distance from the MBP and voila the issue went away!

I wonder how hubs downgrade usb 3 uplink with usb 2 devices

Usually USB 3 hub just contains separate hub for USB 2 devices so they don't use USB 3 uplink at all.

I'm glad to see that problems with Anker USB 3.0 Hubs happen not only for me. In my case, the Wifi became unusable once the Hub was connected. Quite a shame for a nice looking, affordable device to have a breaking issue with shielding.

Nice looking, affordable, doesn't break stuff.

Pick two.

The graphs show an interesting quirk of RFI. In figure 2-1, the theoretical USB 3.0 emission spectrum has a zero at 5.0 GHz (equal to its signaling rate). In figure 2-2 the measurement has a significant narrowband spike at 5.0 GHz. Never believe zeros in spectra, there's usually something lurking under them. 2nd harmonic distortion probably causes it here.

Good timing - just last night I had my speakers plugged into the NUC make weird noises every time I scrolled using a USB mouse receiver plugged into a USB 3 port. I knew it was interference but couldn't have made the USB 3 connection without reading this.

I don't think mice use USB3. It could be some other interference.

Actually, he is probably correct, but I think in his case the opposite is happening from the article.

With USB 2.0 and before, USB ports would have input filtering so that this RF energy would be attenuated before entering the system. With USB 3.0, there is so much speed (5 Gbit/s!) that a 2.4 GHz transmit signal is actually inside the bandwidth of USB 3.0, so it can't be filtered out as easily.

So, depending on the cabling inside the computer and the design of the motherboard/system, it is definitely possible that some of this energy from the nearby transmitter is leaking out into other areas (ground planes, power planes, etc.) where it would be coupled into the audio path in USB 3.0, but attenuated with USB 2.0.

Though USB 3.0 works by adding new wires, and there is no reason to remove the filtering on the wires/pins that carry 2.0

As another comment said - the culprit might not be USB 3. It could possibly be your monitor? I had an old monitor (not CRT) that made strange noises when the whole screen refreshed, such as when scrolling. I could possibly see that causing interference as well.

Actually I can tell fairly certainly from the direction the noise comes that it's not the monitor - the noise comes from speakers (which are separate and apart from the monitor) and it starts when I press the scroll wheel on the mouse - I need to see if a wired USB mouse causes that or not.

I think the other commenter thinks your monitor is causing the interference which is causing your speakers to create sound. The trick would be to turn off the monitor and see if it still results in sound when scrolling, or just scroll to the bottom of the page and keep scrolling without having the monitor change in any way.

Yup I ended up installing the wireless mouse "unifying receiver" on a USB extension cable so that it would work. Since I have a spectrum analyzer I could confirm that interference from the USB3.0 port was swamping the wireless signal from the unifying receiver.

That's a great idea - have a extension cable lying around, will give it a try!

My MS wireless receiver is on a USB 2.0 extension cable, plugged into a USB 2.0 hub, plugged into a USB 3.0 hub, plugged into MacBook's USB 3.0 port.

I tried it without the USB 2.0 hub, but it was flaky.

If it's a skylake CPU, chances are it has nothing todo with USB3. There are some coilwhine issues in the power delivery system when the GPU is entering and exiting power saving modes and is heavily plaguing the recent Dell XPS 13 model amongst others.

I'm guessing it's not really worth RMA'ing my 9350 for that coil whine issue. The problem where the Dell WiFi driver causes screen flickering might be, though.

I talked to a very nice service rep about the coilwhine and he said there wasn't much point because all of them suffer from it in varying degrees and they would only do something about it if it was very audible and then only if RMA'ed within the first 2 months or so. His suggestion was to disable the GPU power states but that sort of defeats the whole point of an ultrabook. It's a sad issue because it really detracts from an otherwise excellent laptop. I have not heard about that specific wifi issue though, have you updated to the latest firmware? I know there's been quite a lot of versions with various fixes.

Nope, it's a Haswell NUC. Skylake is problematic though, yes. Got a desktop Skylake Dell Precision and it was weird to see all that unreliable behavior - USB bootable disk sometimes would work and sometimes not. The computer itself would sometimes fail to start back up after it had crashed trying to suspend etc. It was fast with NVMe boot disk and 64GB RAM but it's a bag of pokes!

I'll have to read through it, but I sometimes shut my wireless mouse and put it in its cradle when downloading a large file over my WiFi-tethered mobile phone, place the phone in a 45-degree angle instead of flat on my desk. No scientific reason, just guessing it puts it more in direct line with the WiFi antenna in my notebook and the tower.

It is my sole internet connection here in SE Asia, and this seems to help, but maybe I am wish-biased!

Wireless input devices usually talk in the 2.4GHz and 5GHz bands of wifi. They are allowed to, since the spectrum is basically open for anything (AFAIK) but they effectively jam wifi when in use. So this is intentional while the USB3.0 cables jamming wifi seems unintentional.

Wireless mice usually don't talk very much. Mine runs for about a year on a single battery (the spec says up to 18 months). That's MUCH too little power to be able to effectively jam anyone's wifi.

Reduce its capacity by a small fraction while the mouse is moving, yes, effectively jam, no way. That would need much more power than a wireless mouse has.

Your phone's antenna probably runs along the side of the screen, same with your laptop's. You want polarisation to match for best signal strength so it's best to have your phone as close to the same angle as your laptop screen as you can get it.

I also heard about this problem with Seagate 2.5" USB 3.0 external drives. Their cable included was not good shielded and it was causing problems with WiFi. Once the cable was exchanged with a good shielded one the problems are gone.

well shielded :-)

How did you specifically shop for shielded cables? Are they advertised as such, or did you just look for something more expensive and hope for good build quality?

They are (or should be) all shielded. You just need one that's correct.

I have a Seagate USB3 drive and it is giving me bluetooth problems. Its cable is very thin for a USB3 cable. A shielded USB3 cable typically is as thick as CAT5 Ethernet.

The problem is, it's hard to tell how thick a cable is looking at the pictures on Amazon. I need a good quality 1 ft long USB3 cable and I don't know how to find it. Both my short (1 ft) cables are not shielded.

you cannot know it really. The Seagate one is really thin, every "normal" USB cable you buy should be thicker. Btw. I used an old USB 3.0 cable from a broken Western Digital 2.5" drive, it was thicker.

This is what all that tedious EMI testing and certification is supposed to prevent. Although I think it's much laxer for "conducted" emissions ..

I found out about this issue while purchasing an AC spec router last year:

https://community.netgear.com/t5/Nighthawk-WiFi-Routers/Netg...

The Asus RT-AC68U did not have proper shielding on their USB 3.0 hub (at least in their launch hardware revision) which lead to interference.

Is there an easy way to test for this? I just bought an AC router (TP-LINK AC1900) and I had planned to use the USB port.

I would assume running USB 3.0 read/write tests with radio on vs off would provide a good benchmark.

...does this explain why my bluetooth mouse or trackpad becomes unresponsive on my 2012 MacBook Air?

I haven't noticed a correlation with USB disk access and the mouse going limp, but now i'll be looking out for it.

The study says the interference is there if the disk is being accessed or not. Wrapping it in tinfoil seemed to make a difference though.

Apple actually has a support article on the topic[0].

[0] https://support.apple.com/en-us/HT203729

This is almost a daily annoyance for me. I have a WD Passport Ultra HDD at work, and any time I plug it into the USB3.0 ports on my Dell E6430, my mouse (Microsoft Wireless Mouse 4000) gets laggy.

I remember listening to FM radio with either my smartphone or a nokia candybar, when I was in the bus or tram, the radio "cut" for 2 or 10s or even longer.

I think I remember it happened on both devices.

Certain carriers were worse for this. When I had a Nextel phone, my PC speakers would give me advance warning of calls ahead of the phone ringing.

Found this out the hard way when I bought a little USB 3.0 hub off Amazon. You can plug it in and watch ping lose its lunch, then unplug it and the wifi returns to normal. (On a 2015 MBP).

Would this work the other way around too? I.E. use SDR to simulate a connected device over USB 3?

To cripple WiFi/Bluetooth in the area? Sure, but at that point, you might start to get in trouble with the FCC (in the US) or other regulatory authority.

Does anyone have pointers to good shielded usb 3 cables of different lengths?

Found out about this the hard way. I was confused why my Wi-Fi was slow. Disconnected USB3.0 external drive and bam, fast network again. Now I just have this drive as far away from the computer as possible, works fine.

I had problems sometimes with 2.4Ghz Wi-Fi, I permanently switched to 5GHz Wi-Fi which solved the problem. It may have been the USB3 hub I was using, I'll do more test.

A better solution would probably be to buy a well-shielded USB cable for the drive, then.

Reply on news.ycombinator.com