March 4th, 2025

Pixel Problems, Thermal Testing, and Dusty Desert.

And a missed blog post.

First Things First.

Where on earth was last week’s post? Why did I miss it? Do I not have enough content? Did the chickens eat my computer?

No.

Actually, I got sick again. Thankfully I’m mostly fine now. And I even had some content sprout up in the midst of being sick.

Why, Google? WHY?!

I have mentioned before that I have a Pixel phone. Pixel 4 at one point, now a Pixel 6. My sister has one too. A Pixel 4 XL. She’s been using that phone longer than I used my Pixel 4, thanks to the fact that it didn’t have odd power issues like my Pixel 4. So far, so good. Until last week.

The cameras stopped working. Straight up quit, with no warning signs whatsoever. Upon opening the camera app, the phone displayed a message saying there was something wrong with the cameras, and only allowing selfie mode to be used. Odd. I rebooted the phone, as per the message, and still no dice. I even went to the extent of pressing where the camera connector was. Nope. Nothing. Absolutely nothing. She relies on the cameras on her phone quite a bit, so it needed to be fixed. One replacement Pixel 7 ordered later, and hopefully the phone would arrive, and I would have a phone to try to repair and make a blog post out of. The phone arrived today, but a couple hours before the phone showed up…

The old one started working again.

I don’t know exactly what would have caused this, but there are a few things that could have caused it to self heal.

My first suspicion is that the cable got seated just right. If this it the case, then that is fantastic and means that not much needs to be done unless it breaks again.

My second one is that the part is dying and will do it again in the future.

My third and most unlikely suspicion is that the phone broke to encourage the purchase of a new device, only to decide that it didn’t want to be replaced with another used phone at the last minute.

Either way, I find the whole thing very odd. I know someone else with the same model of phone, and the cameras on their phone died as well, though the cameras never came back from the dead. Maybe I should try to get my hands on that phone and see if I can get the cameras to work on it again.

She ended up getting the Pixel 7, a nice upgrade for her since the cameras are really good.

However, that was not without its problems too. The seller seems to not have very thorough testing, as the device was slightly worse than “fair” when I got it. The charger did not go in right, the phone was slightly bent, and the screen had burn in. I managed to clean out the charging port (91% isopropyl alcohol, a toothbrush, and one of those plastic teeth flosser things with a pointy at the end) , and it works properly now. I also managed to bend the phone back almost to its original shape. The only issue is the burn in, specifically on the bottom, where the navigation pill is, and the top of the phone, where the time and notifications are. It’s not too big of a deal, but the seller should test for these kinds of defects. Especially considering that I bought this from a place that buys used electronics.

Thermal Nonsense.

I have made previous indications in this blog that I am a bit of a nut when it comes to CPU benchmarks and cleaning computers. I especially enjoy redoing the thermal paste on computers that have 10+ years old paste, as it is quite satisfying to remove the old and crusty paste. I’ve been using Arctic MX-4 on all of my devices, and so far it seemed fine.

I had repasted my Shuttle SK43G last September. It has a AMD Sempron 3000+ CPU in a AMD Socket A motherboard, and as a result is direct-die cooled, unlike most modern desktop CPUs with an integrated heat spreader (or IHS). I ran Cinebench 2003, my personal favorite for these older machines, and got a score of 247. Pretty good. Fast forward to this January. It was cold outside, and I was stuck inside. I re-ran Cinebench for the fun of it. 231 was the score. This was on a fresh boot, the machine had not been used for some time prior to this. I decided to let it sit outside in the 40F weather, cool down, and try again. Same result. So I let it cool the rest of the day, and stuck the machine out the door, fans pointed into the cold, and ran it again. Same result.

I ended up shelving the idea of that, and decided to revisit the problem. I have reason to think the problem could be any of the following:

1: The thermal paste degraded, and as a result the CPU runs slower. This could be the case, but HWInfo shows that the CPU is running fine, and is being cooled below the 50C max the CPU should experience. I also ran the fans at full blast in a cold, outdoor, environment in the night, and had the same 231 score.

2: The first time was a fluke. This could be quite likely, but I don’t know why the CPU would get that high of a number and never get it again. I don’t recall overclocking the machine at any point in time either. If I did overclock it, I would have remembered doing so.

I decided that the first was the most likely. Some research shows that Arctic MX-4 suffers from “pump-out”, especially on direct-die like my Shuttle, and maybe even some laptops, which is quite worrying considering I have used it on a few laptops, and if they aren’t being cooled right I will have to fix them. I will have to buy some of the thicker, and supposedly pump-out resistant thermal paste to test it out on the Shuttle. I did run Prime95 on the Shuttle for a good 30 minutes after the repaste, so it’s possible the heat from that degraded the paste as well

I think I may attempt to do a test using a more modern laptop that I have around here. Generally speaking, most laptops are direct-die cooled, and as a result have a chance of pump-out. I also have a Socket A tower, though the Abit KD-7 board powering it needs a serious recap due to nearly every capacitor being bad. I also need to find a quiet fan for it, as the previous owner wired the CPU fan directly to the power supply, causing it to sound insanely loud. I assume they did some overclocking with it, as it has quite a beefy heatsink on it as well. The CPU on it is direct-die cooled as well, and could be another point of testing if I can fix the board.

Dust.

Yesterday there was a nice dust storm. Turned everything red inside due to covering up the sun’s light for a bit. It is still quite windy today.

Keyboard Layout Switch Update.

I switched over to the Colemak layout some time ago. I have been using it, and have noticed the hand strain is less than QWERTY, but I still get some. It is a switch that I think is beneficial for me. I also reconfigured my keyboard to use its layout so I do not have to deal with changing layouts in Windows. I did leave Caps Lock alone as I am a heavy user of that key. I did notice that I do not tend to use the right finger on the right key for some reason. That’s going to be hard to train out of my brain.

Lent is Nigh

Today is Shrove Tuesday, or Fat Tuesday if all you think about is food. Which means that tomorrow is Ash Wednesday. Don’t forget your extra Lenten penance, and don’t forget that St. Joseph’s Day (March 19) and St. Patrick’s Day (March 17th) are perfectly reasonable days to eat treats on if you give those up for Lent.

Desk Nests.

Agathae decided my desk looked like a ideal nesting spot. She seems to show up, sit for a while, then leave without laying anything.

Agathae on my desk.

She then told the nest hunting Dominica about it. I heard Agathae make some kind of noise, prompting Dominica to jump on the desk and find the spot.

Dominica on my chair. Not the desk. Oops.

Or not. It took some time, but she finally got it right.

Dominica in nest on desk

On top of that, she found some nesting material, which was promptly confiscated.

Dominica holding a piece of foil in her mouth.

Chickens have once again proven they are smart.