Quantcast
Channel: Raspberry Pi Forums
Viewing all articles
Browse latest Browse all 5044

General discussion • Re: likely failure points in rpi 3B

$
0
0
Hello there!

Long time user of rpi's --- I've got a 3b, multiple 4s, and now a 5 purchased just recently. They just work, haven't had any particular problems. Used them as light desktop replacements, with an XARCADE controller for some retrogaming, and a pi-hole, amongst other apps.

I've been using my 3b as a pi-hole for at least 4 years continuously (longer?) in an equipment rack in a 21C environment with UPS power, so it shouldn't be abused too much, random pulling of power, etc. The 3B was purchased on 2017, and I think running Raspbian 9(!!).

The pi-hole serves as my DHCP server, and as a dns-relay -- so it's a fairly critical device to having internet access at home. Especially with two teenagers in the house.

I've treated it like a dumb appliance. Maybe upgraded pi-hole software once or twice a year. I have had zero problems since installation. Recently, it locked up, with a solid green LED. Powered it off, back on again, everything came back. Two weeks go by, and it's happened again. Solid green light makes me think disk problems.

There's no heatsink, no active cooling. I ran a badblocks read-only test against the SD drive. I also created a large file with dd. Not seeing any errors. I pulled the power supply and hooked it up to an electronic DC load. I tested 100ma current through 1700ma, and it was above 4.75v the whole time. The supply is rated for 2.6A, but it does fall down outside USB spec on larger loads. I'm not testing the supply over time or doing anything intensive.

I'm running this headless, so no monitors, nothing attached via USB, connected with a copper ethernet cable.

Are the likely failure points the SD card, and the power supply? I popped open the case, no dust or anything inside, and replaced the power supply. I'll probably just replace the unit en masse. I'm inclined to put together a higher reliability supply because I can't stand these crappy quality wall-wart supplies.

Thanks
Historically, the 3B had an onboard power supply that had trouble keeping up with the CPU when it switched from idle to full load. The official solution was overvolt so the dips didn't cause crashes. My experience is setting the maximum CPU clock speed to 900 MHz was preferable. That's what I'd try in you case.

For more information, see

viewtopic.php?t=139712

Statistics: Posted by ejolson — Fri May 10, 2024 3:12 am



Viewing all articles
Browse latest Browse all 5044

Trending Articles