HA 2026.4.1 - Raspberry Pi 4b fried after update?

Before I write anything else I need to emphasize that I can’t for sure know what happened during the period in which it updated or what’s at fault, I’m not directly blaming HA but it seems so weird and I wanted to report this just in case (sidenote - I literally had a HWA-2 in the mail while this happened and even if I could afford to replace this Pi I don’t know if I have the appetite to continue with HA now. I thought that this would be a better long-term option to something like a HomePod Mini and now I deeply regret going this route :pensive:).

I had a HA setup going for about a week and (2026.4.0) and apart from beginner hurdles it worked well and I was relatively happy with it. The device itself is a Raspberry Pi 4B 2GB that was collecting dust, and it was powered using a UGREEN USB-C charger through a certified Lightning cable that ive used to charge other things (it is verified, not a fake cable). No problems.

A day or two ago I get the update for 2026.4.1 and so I update it, the app complains that a process is already running (something involving core itself which seems weird), so after trying and failing a few times I try to restart the device, and though it reports an error I don’t remember (I think it was once again complaining about a subroutine called core running), it appeared to restart after becoming unresponsive so I leave it to it’s business. I get back to it later in the day several hours later and I try to connect through the local website and app and nothing happens, but it’s late and I want to sleep so I decide to leave it.

I wake up later that morning and inspect the device when I have the time try to figure out what happened. The device seemed particularly hot compared to what I expected and I noticed the Ethernet port lights weren’t blinking on the router like they would, but I also remembered that Pi 4s had a reputation for being at least a little spicy so this heat didn’t strike me as unusual at the time. I unplugged it, tried re-plugging it in, then tried reinstalling HA OS on the SD card twice. Nothing. Other things I try at this point:

  • Tried using different ethernet cables.
  • Tried using different chargers (including the official Raspberry Pi 4 USB-C charger).
  • Tried testing the router ethernet ports to make sure they still worked using other devices.

I leave it for another day as I don’t have the time for it.

I return to it this afternoon and on inspection find that the case has melted around the power connection and that the “Dialog” chip near the USB-C power connection appears blistered, I get the impression that the chip got fried at some point during this whole process. Pictures are included below:

(I’m a new user so I can’t post a second image of the melted plastic case, but it’s melted in the area around where the chip is, not where the USB-C power is)

For additional details:

  • The charger was connected to a surge protector.
  • The charging products aren’t knockoffs and haven’t damaged anything else I own in the years I’ve used them.
  • No lightning-based weather, no apparent or reported power surges happened in the area.
  • I used an official RPi 4b case and SD card.
  • The weather in this area has remained cool the entire time.
  • While it wasn’t put in a place with perfect ventilation, it wasn’t sandwiched between devices emitting lots of heat.
  • Tried installing Raspberry Pi OS with pre-configured WiFi details, cannot SSH, never connects to my router.

I don’t have advanced soldering skills or a station so there’s no way I can fix this, I just need somewhere to report this as it’s been so disheartening and I need a sanity check on what happened here. Thanks in advance.

Random hardware failure is random.

You failed to install the cool app update on April 1! Seriously, a software initiated hardware failure is rare, and blaming a software update when millions of other Raspberry Pi’s have successfully updated is a very long bow to draw. Attempting repair is uneconomical, even if you could source the individual parts, the internal PCB layers are most likely damaged. Accept fatality is final. Move on.

Commiserations, and possibly a cooling fan or heatsink for your next incarnation.

Check your UGREEN USB-C charger power supply voltage before you plug it into something else. It may have been affected or the cause of your woes. The signs that the Pi was on the way out were already there when tasks became unresponsive.

You mentioned dust. Encrusted thick layer, or just old? Contributing factor?

Your SD Card may have data you can restore from. I would treat that as a disaster recovery process, making a one-off clone copy of the card before discarding it. Don’t trust it for future use, even if it appears to be superficially ok. Random data loss is infuriating if you discover all your existing backups are affected too.

Consider waiting a short period if considering a replacement, as it appears the memory shortage price peak has arrived and prices should start falling soon.

1 Like

Yeah, hardware failures like that are frustrating!

Assuming you’re looking at buying a new host device for HA, consider expanding beyond the RPi line. They were once the best price/performance deal available. No longer. Lots of threads on this subject here, so I won’t repeat. Buy any small form-factor computer, or a used laptop or whatever. Note that HA now sells dedicated hardware like the “Green,” too.

I hope @IOT7712 knows something I don’t know about memory prices. This sure doesn’t seem to be a good time to be buying. Would be great if they start to come down.

@CaptTom I appreciate your considerate response given the other reply. I’m aware it’s exactly the wrong time to buy anything with RAM so I’ll be holding off for the time being.

I did have a brief look at the Green, but I’m gonna be honest the other responses on this thread make a pretty uncompelling case for the device if I ever need further assistance. I’ve never directly blamed HA, as said at the bottom I just needed a sanity check on what could have gone wrong here. Thanks though.

Retail memory prices may take a little while to come down as the more expensive stock is sold. Market pressure will take care of the rest, the windfall profits up front now becoming trailing losses that may have to be absorbed. Witness the 3Gb Pi just announced, as a delayed response to chip prices. Hopefully they only have small runs in the pipeline for what will shortly become a product very hard to sell.
Even second hand equipment prices have surged and will come back to their natural level as new prices put downwards pressure on them.
Wait and watch, unless you need a replacement right now.

I have no replacement device so I’ll just go without a home automation system for now and hope things get better, thanks for the info though :pray:

There are lots of bargains. The high RSM price do not seem to be affecting the price of older computers. For example you can buy and entire 2014 vintage mac Mini for $125. This is run cirlces arout a Pi4 and it comes with a nice case (milled from a solid block of aluminum and has an internal power supply and fan. The Mac now runs Linux

But my cheapest nw “server” is a 12-year-old PC notebook a freain gave by because it can not run Windows 11. It has 6GB RAM and an Intel Core Dual inside. I installed Proxmox and run it with the lid closed. You can find these computers for well under $100.

These older computers from 10+ years back are more then you need for HA and are cheaper than the RAM that is installed in them

1 Like