Asus Chromebook Autopsy - LFC
Вставка
- Опубліковано 28 лип 2024
- Not sure what happened to this Chromebook in the end. It looked as simple as a blown mosfet at the USB Power Delivery control chip, but after resolving that, we only uncovered a short-circuit deeper in the system (PCH? Aux CPU power rail?).
Check out Adamant IT 2 for the Pod Cast and More: / adamantit2
I stream on games Thursdays and Sundays, 7pm UK: / nethesem
Join the Discord right here: / discord
Support the channel directly at: / adamantit
Follow us on Twitter at: / adamantit
Check out our Instagram at: / adamant.it - Наука та технологія
Before anyone suggests it - remember that you can't inject power on more than 1ohm. A) The resistance means you'll need too much power and will likely damage something. B) Over 1ohm is not a "true" short, it's probably low-resistance through silicon (CPU/PCH/GPU, etc) so injecting on that is also a bad idea.
Sometimes economy isn't the only rule. One of my largest quarrels with Apple is -- all those "disposable" iMacs, what are they doing to the environment when they get put in the skip out back? Also, the budget end of computing often serves those who simply can't get better equipment, and are forced to rely on stuff like this that's arguably eWaste right out of the box, but is a viable computing platform of at least _some_ sort and is arguably better than Nan's mid-2000s "Vista Certified" Dell whatever that somehow still hasn't caught fire and died.
I'd like to see this one running again. It matters to someone, computing just kind of is that way. Plus, like I said, money is not the only consideration -- and if you want to see the sort of crop returns you get reaping the eWaste recycling industry, take a trip to a little town in China named Guiyu. (Warning: yikes.) Besides, everyone likes a feel-good video, and it doesn't take much to get folks to root for the underdog, either.
I, for one, want to see you go fullmetal Rossman on this :) let's save a hopelessly cheap Chromebook!
As long as you remain at the below 1volt territory,it's rather safe.You can limit the current on you bench psu.
as long as voltage is under 1 volt, it will not kill the CPU. and you can limit the current and increase it up to 5A if needed.
@@dimitrismaster V=IR. If there is 1.4 ohms of resistance, you can only inject 0.7A whilst remaining under 1V. That _might_ show up faintly on the thermal camera, but most likely will not be enough to detect. To inject a decent 3A, you will pumping 4.2V into the board. No ways around it.
@@man_eating_monkey One can stay on the safe side of things by removing the cpu coils.Or the pch.
You gave that a more than a valiant attempt. Thank you for letting us be a part of it.
That was engaging and instructive. One thing on which you might have elaborated was why after you'd found that super-hot component you spent you time probing around it rather than removing or replacing it immediately.
An overheated component doesn't necessarily mean a faulty one, It could be the cause of something else around it.
Fix or no fix, just interesting videos to watch. It's educational.
I like to leave the success or not to the end, it spoils the enjoyment for me, there is always hope. Great video.
That was interesting to see, i think it was worth the time just for the content.
That g5m chip at the end is probably some sort of linear power supply, if the output of those are shorted the power supply itself gets hot instead of the shorted component in the output, power injection there and the thermal camera could've possibly shown it.
Great video Graham, keep up the good work!
Cheers!
great video yet again graham, loved the interruptions and the almost vinyl scratch repeating of your sentence. This job would be easy if it were not for customers lol
“This Guy is shorted straight through” would make a good T-shirt
Thanks for sharing this. I feel I learned more from this than from other successful videos. Would have loved to see you go into greater detail for the sake of learning. But I understand it doesn't make much sense economically to go any further.
I've worked on this board before. I had a similar issue with the GSM linear voltage regulator overheating (due to the SlimPort IC drawing too much current). The SlimPort power pin 24 (if I remember correctly) seems to short out on these models a lot. Check the SlimPort power pins for a 10 Ohm short at 3.3V. I replaced the SlimPort IC and the chromebook worked for a month before shorting out again. I gave up after that
You thinking too much about replacing mosfets with same exact model. If we're talking about switching multi phase power supply then yes, is best to keep all of them same type and characteristic as designer of the board choose (to keep everything balanced, to be sure driver specification are good enough to maintain gate switching fast enough... )but for case like this one where it only open or close to enable power rail all you need to be sure is that voltage rating is fine for usb-c charging use and every mosfet you can find in that package will have high enough current rating. Long story short any trashed high end GPU from spare part box will provide you with enough mosfets to fix 20 of these.
This is good to know!
I'm glad I'm not the only one who falls into these frustrating rabbit holes.
big thanks for your informative videos
Nice video. I liked it. Thank you
Worth repairing if you do in-house repairs at a School as we cant just buy new laptops (even cheap £100 Chromebooks) unless it was planned for as all the money is allocated/accounted for so great to see this type of repair.
Fast becoming Alice with that rabbit hole. Loving to know if it's a hobby but appreciate stopping with your business .
great teaching video
USB-C for power delivery is a very bad way to do it. Elegant on paper, but too many points of failure, i.e. connector, controller, discretes, etc.
Yea, probably a much better way to do it like they've always done it. It's kinda like the PFC circuit.
That is great advice thanks💯👍🏻
Mixed view on the video, some really good diagnosis, some poor inefficient efforts. I run a similar repair business and was screaming at the screen, "you have the schematic, use it", the same thread on Bad Caps also has a BoardView for this PCB, so you could have traced it that way.
I had the same Chromebook come my way this week, same problem you state, just not economical for board repair but I took a look anyway. The 5 pin chip on your board that got hot is PU8304, it was an LDO type PSU IC, the FET you removed was PQ8313. LDO's only get hot if faulty or its output is shorted, removing PU8304 would determine if the LDO was shorted or it was on the output. With this information you would have traced further (which could still have been futile depending on what it was).
You can inject power into less than 1ohm, you just need to limit the current on the PSU so that the power isn't too high. In this instance the LDO is a low current device (typically less than 100mA), so a current limit of 100mA would have been fine to inject as it would be the same as what the LDO was doing.
I appreciate that this is an old video so this information wont help now, it is more about improving the process to find a fault, you don't want to be in a potion of randomly removing components when you have a schematic and boardview on hand.
Keep up the good work!
You nearly got there. The chip you took off last is probably as you said feeding a critical part of the circuit, so whatever it is feeding may be the fault. The EC is shutting down the charger as the last chip you took off is missing and not powering whatever it should be powering, but it does initially charge before shutting off, so progress - something was learned....maybe. 🤔🤔🤔
Having watched loads of these kind of vids: 1) You can’t tell if the mosfet is shorted on the board - It has to be taken off and the gates shorted by connecting the pins/gates with tweezers to discharge them.
2) It’s better to inject voltage at the short to see which component gets hot (Usually a capacitor) - Use a thermal camera or alcohol or fingers before injecting like 1 volt - Sometimes you may need to go a bit higher.
I suspect the chip in frame at 5:33, I believe it's a security chip.
I've seen a bunch of these power issues on Acer and Asus Chromebooks. A lot of them it's been a failed slimport, but if that doesn't solve it, it seems there's a data line that goes to the security chip.
I've reballed and replaced with a donor but without a way to restore the original data they boot without seeing and OS and refuse to reinstall.
*data line or shared power rail
Brave of you to Show a Fail, thanks
Good try and video!!
I do agree that the blown mosfet was prolly a symptom.. no the cause.. BUT.. it may need both mosfet (switches).. instead of ordering & waiting forever.. look thru ur parts boards for a similar voltage mosfet, flip it upside down, solder some thin enameled wires, its only 6 pins.. really only 4.. enable, in, out, gnd.. mount it upside down somewhere... run it...
at some point you have had a charging light and when you moved the charger it switched off witch made me all skeptical about the usb type c charging port, I hate those ones. other then that I was betting non the usb-c driver chip I'd have romouved that one all together and proceeded with taking measurement and plugin in the dc charger. again as always great video.
I thought I saw a tiny hole in the controller chip. Tiny holes means the chip has blown.
no hole on internet pictures, you might be right.
I haven't see you trying to look for shorts on that USB chip you were talking about. Maybe, just maybe you're right and that IC is faulty. Yes, it's too cheap to bother, but, for the learning experience would be worth :).
If you'd charge the battery, or use your bench power supply, will it turn on? If so, you can do the Frenkenstein with it ;). Only to try get rid of short.
It's similar for phones after they are 4 - 5 years old. 5 year old phones are about $100 2nd hand, and 2nd hand replacement screens are about $80, ie, not economical to repair.
But then all this flips if they are e-wasted to poor countries where labour and parts are cheap, ie, they might want these C-books.
I hate that so much, phone screens used to cost as little as $20-$50 now modern screens cost $99-$300, so even if you're willing to do the work yourself it's still just not economical to repair because a new used phone only costs $150.
@@vgamesx1 Yeah, I found that out recently. I guy sold me a supposedly working S8+ for $30, but after 5min the screen overheats and glitches white. In the end I had to get a 2nd hand screen with spots/faults for another $40, just to get close to 2nd hand working phones - working S8+ with faults usually go for $50-$60. lol -.
Good 2nd hand OLEDS are about $130+, and new TFT versions are about $80, which is more that working flagship phones after 5 years - ouch -.
SlimPort is usually made for USB-c to HDMI out.
When you say you have not achieved anything (around 17:00) the light on the USB cable does stay on. So you fixed that part at least.
Oh, I used to repair Chromebooks for a living for schools.
Nice one
Graham do you take doner parts, if so that could save you lots of money and time and then worth it to fix certain jobs?
Mostly only on Apple devices or common models. If you try and keep donors for everything, you end up with a room full of broken parts.
@@Adamant_IT Off topic: have you looked at these SDR USB devices, I've messed with them wondered if you have and what are your thoughts? It is a hobby connected to PC's but I found an online version of SDR and i use that now.
very funny and educative also :)
I thought a usb c cable worked either way around Graham?
So the capacitor you mentioned in the beginning of the video, measures 1.5ohms, so conclusion is must be shorted.
But I was watching the injection power video the other day and you were saying even if it beeps in continuity mode and has low resistance, doesn't mean it's a short, unless the reading is below 1ohm?
Apologies, just trying to understand the basics of investigating shorts
Yea, the short that was showing up here and there was about 1.2 - 1.5ohm, which is a touch too high for injection, and as you noticed, quite possibly low resistance through a chip like the CPU or similar. So yes, this is not technically a short, it's low resistance, but sometimes that means you're close to a short, or maybe bad probing means your meter can't quite get low enough to see less than an ohm.
The main stopping point was that the section I was looking at toward the end of the video, I'm honestly not sure what that circuit was doing. I could've spent another hour matching up components to the schematic to figure out where I was, but even then I still might've just found that I'm shorted into the CPU, and it was all a waste of time anyway.
Very good
Oh man I was just working on a shorted motherboard that was annoying like this. I couldn't find the short no matter what I did. Even injecting higher and higher current nothing got warm. I removed likely candidates and nothing fixed it. I suspect it was an internal power plane fault or something but it's so weird that something like that would just randomly happen. Without schematics it was really hard to even guess where the problem could be. If it was a trash board I would have kept going until I broke something but as it was it was practically new condition so I threw in the towel. I'm going to do some reading of research papers and do some experimentation to see if there's an easier way to locate really difficult faults like this somehow because this is not the first time I've had one of these. Hmmm, I've always wanted an x-ray machine, hehe.
hI, CAN I ASK WHAT OVERHEAD CAMERA YOU ARE USING? THANKYOU
Cameras are Insta360 Link, microscope is an Andonstar AD407
Hi.
Are those multimeter probes of any good? Can you give a link?
BST-050 JP Superfine, they're very good for the money. Don't have nice spring-contacts or anything, but they're very sharp. www.aliexpress.us/item/3256801370930419.html
Love bad caps!
Typical techy "I'm NOT done yet" LOL
I call this fail video a succes😉👍
Can you teach us some more hardcore complicated things?
I like your videos very much and you deverve all donations
That Phone Ringng..ROFL
Show me where you find a new Chromebook for £100 ...
What happens next, contact the customer explain the situation and just....pick up the next repair?
Pretty much yea. I tell them what the cost of a replacement mobo is if it's available, and also how much I would charge for a refurbished laptop or a brand new one. Then they can decide what they want to do next, whether that's to go with an option I presented, or just take the dead laptop and go.
First time seeing inside a chrome book. Most go straight to trash
you finally figured out that humans are not that important like was in industrial revolution, machine and I A....you are out of equation
You are sadly right that most cheap Chromebooks are E-Waste
However. Every so often you find a decent Spec one that dosen't have a locked down boot loader. Load up a 64-Bit or Arm compatible OS of your choice depending on the CPU and suddenly you have a slow but much more capable device.
I'm almost at the end and Graham hasn't informed us he is being honest. Not once. Disappointed. 🌝
To be honest, I don't even think about the phrases I repeat often
Wait...WHAT!? a chrome anything is a abortion from the start...let it RIP.
Stop! No!!!! 🤣🤣
More fail vid.,.. L'OL
saya hadir
pooooooooooooooooor
Would never buy anything made by Asus
Definitely not lately.
yeah i dont trust asua either
Their enthusiast lines (TUF/ROG) are basically a different world, but their consumer lines like these chromebooks and their common X-series tend to be awful in design, spec, and build quality. There's going to be better ones if you spend more, but certainly they churn out a lot of bottom of the barrel devices that get sold in big-box stores.
@@Adamant_IT Their gaming laptops are a pile of poo also..... Just search youtube on the countless repairs being done. SUS is in the name... so there is that.