Powernumpty

Do the slow waterfall refresh thing. If you have a PC/MAC run the slow waterfall on that, set it up to show the noise best on the waterfall, leave it running as long as you can, while you walk around or stress equipment, E.G. run a heavy graphical render on a PC, CPU stress test, .

Part of the reason I haven't been on here for ages is that I got fairly obsessed with tracking noise down, got the general level low (fibre lines through the house, CAT7 where it made a noticeable difference) lots of grounding, ferrites and careful cable routing, only to then have any one of my neighbours negate the gains, instantly with a Chinese battery charger or bad PC PSU. I should have changed my username to Sisyphus.

The first step is really to characterise the noise and look for changes during the day. If you can locate it quickly that is great, (walk round with an AM battery radio) but if you at least know it changes on light level, temperature, people being awake in the house, TV on, then that narrows it down. All wires in the house are antennas, those DC lines from power supplies need to be coiled up, and ferrite clipped. I reduced my GPS lead pickup by using a short GPS antenna and a car type repeater from the roof.

If you have a mobile browser running with a waterfall, keep moving around poking and changing things until nothing makes a difference anymore. Even before you locate the particular problem, I bet you'll find one or more background source to reduce.

If you have a cable tracer, the type that uses an audio tone, remind yourself of the noise, then go locking for that pickup, or radiation, on any wiring.

Ok, sorry, now I've listened to the WAV it sounds like SMPS.

I had a neighbour with a 1200W computer supply that would put that across the entire HF, he had a noisy supply but would then put that gaming PC in a glass fronted case with extra long white DC lines for appearances.

For ruling out sources in your own home, I used to start a very slow waterfall plot on one RX channel, and another more normal speed. Then go round turning things off/on seeing if that changed anything. Make changes at set times so it is easier to identify trends in the slow plot.

The slow plot can help identify things like "Car A pulls up, one minute later noise starts", "house lights opposite go out, noise disappears", "Garage light off, noise reduces".

Obviously if you have any sort of null on your antenna you can use that to narrow down the direction but that sounds like switch mode. Other things to test are CAT7 cables, and rerouting the GPS antenna, in case there is any pickup on the shields. Little clip on ferrites on the switcher supplies for everything normally helps general noise floor.

tower-research.com

"Quantitative Trading and Investment Strategies"

Thanks for the IP.

Audacity can also be set to record over a set threshold.

Preferences - Sound Activated Recording.

I wonder if it possible to flush any residual flux out from under the BGA afterwards?

https://www.dfrsolutions.com/hubfs/DfR_Solutions_Website/Resources-Archived/White-Papers/Reliability/Review-of-Models-for-Time-to-Failure-Due-to-Metallic-Migration-Mechanisms.pdf