USB galvanic isolation

Hello!

One of my most used setups at the moment is Spyverter + HF modded LimeSDR USB that I use with a laptop computer.
However, when my laptop is connected to a charger I experience severe noise issues in the lower part of the HF spectrum (as expected).
I have tried powering LimeSDR via a barrel jack from a battery pack, but it doesn’t change anything. I know that it is powered from the pack because I monitored the battery current being drawn and observed about 820-1100 milliamps being drawn from it.
I suspect that the noise gets its way into the device using either the power rails of the USB cable or the data pins themselves.
I have tried naively to cut the power pin in one of my tested working USB cables, but this way it seems that my host controller does not want to enumerate the device. I will investigate this a bit further with an oscilloscope in the future.

I have seen devices that galvanically isolate USB devices, however I fell very short on options that would provide USB 3.0 compatibility.

I will very much appreciate any suggestions and ideas. Thanks in advance.

Perhaps worth trying a ferrite choke on the USB cable?

Hello, Andrew! Thanks for the reply.

I do have a choke on the cable from the factory, but apparently it is not nearly enough to filter the noise all the way down to the hundreds of kilohertz, and it becomes apparent when trying to use Lime for HF.

So I have conducted a couple of experiments. First, I tried to supply my laptop with its usual charging voltage (19V) from a pack of 6 Li-Ion batteries instead of the charger. I wanted to know whether it is the charger that is introducing the interference or the internal battery management system in the laptop. To my pleasant surprise, plugging and unplugging the battery charger substitude did not add any noise to the recorded signal. So it seems that the USB power rails are filtered adequately on the way to the ADC somewhere. So it seems that my charger is just exceptionally noisy and I will have to box it in into a metal enclosure with some beefy LC filter on the output. I’ll report if it gets rid of the noise.

Secondly, I borrowed a USB2.0 optical isolator gadget from my buddy who had success in dealing with USB noise issues on his hobby CNC mill. Introducing the isolator had absolutely no effect on the recorded signal whatsoever, so for now I conclude that I should spend my funds elsewhere since USB 3.0 superspeed isolators are prohibitively expensive.

1 Like

If the laptop charger doesn’t cause noise until it’s plugged into the laptop, try putting the choke on the charge cable.

Yeah, I guess the factory molded one that’s on there is either a fake or just insufficient. Thanks for the suggestions - I’ll keep you posted on my findings.