Do Prosticks Die?

Hey team after ideas / help

My pi with FlightAware pro stick and flight aware antenna has up until a week ago worked perfectly.
Now, recently it keeps dropping Data Feed which it will loose for hours and only sometimes come back on if everything is rebooted.

I’ve just tried my old generic RTL SDR dongle and it works fine which kinda points to the prostick?

Anything else I should be looking at ?

Thanks

The heat that thing puts out makes me wonder about its longevity and whether heat generation causes performance and other problems. I found this thread about the prostick heat but can’t find posts correlating prostick heat generation with longevity, at least not yet.

ads-b-flight-tracking-f21/flir-images-rpi-2-and-pro-stick-t38426.html

Do you have the local port exposed to the internet (are you port forwarding)? If so, stop forwarding the port to the external internet, and the crashes will probably stop. This was the issue for me – remote machines were intentionally crashing the local webserver causing the tracker to go offline until rebooted. Now I use VPN instead.

Could be totally unrelated, but worth looking into.

samwathegreat

Can you SSH in to the unit when it stops reporting?

What you describe is often due to a failing power supply despite brand xyz and amperage rating being good.

I agree on suspecting the power supply – the Pro Stick draws a bit more power thanks to the preamp stage.

If you have a good, accurate, digital voltmeter, check the +5 rail at the Raspberry Pi, such as on the GPIO pins (carefully, you don’t want to short +5 to adjacent pins).

If it’s below about 4.8 volts, I’d replace wall wart and/or USB cable until the voltage measured at the Pi is between 5 and 5.2 Volts at the Raspberry Pi under your full load. If you’re running a Pi 3, close to 5 works, as the power supplies on the Pi 3 are better than on the 2 (better, but it’s faint praise).

Don’t worry about heat on the Pro Stick – if you can touch a component and it doesn’t burn your finger, it’s not that hot. Yes, running cool is better than running hot, but some things run warm. There are approaches to taking away that heat, such as by using Chomerics thermal gap filler material as a heat spreader/transfer to transfer heat to a cooler, larger surface. Aavid/thermalloy also makes some good thermal gap filler materials.

bob k6rtm

(a member of the original Macintosh team)

Most of the longevity studies on silicon chip were done on CPUs. CPU probably have a similar lifespan of the prostick considering they are running constantly.

Basically they found chips running a bit hotter than room temperature, 30 C, were estimated to last 25+ years.
Chips that run at 50C were usually around 10 years. <— Prostick is here
Chips that run at 70C+ (these are usually CPUs) were 5 years or less. ← Raspberry Pi are usually around 50-70C
These are mean time before failure so it is possible to have a 70C chip that runs for a decade.
If you are interested in more information you can usually find the papers through a google search for “CPU temperature lifespan”.

Basically, silicon chips are usually not the first thing to die.

The most common problems of electronic failures are capacitor failure, liquid damage, static electricity, dropping it.

Have you posted any stories on folklore.org?

not familiar with that one – I’ll take a look in my voluminous free time…
I was at Apple from October 1980 - March 1997

bob k6rtm

Everyone always blames the power supply but it could as easily be the cord from the wall wart to the pi … if it’s a bit of a long one you might get sufficient volt drop to make the Pi + ProStick unreliable.

If you use a long one to charge the phone it just takes longer.
on the receiver setup - it might just not work right.

That’s why I say to measure the voltage at the Pi GPIO pins – that’s what is getting to the Pi, combining the wall wart with the cable and its connectors.
Many moons ago I built a test jig to measure the voltage drop in USB cables under a 500mA load.
I cut up a bunch of cables!

bob k6rtm