Dell G3 Gaming Laptop, Charger Cuts Out - LFC
Vložit
- čas přidán 19. 05. 2022
- A "shorted" main-rail makes this Dell laptop look like an easy fix, but 2.6Ohms is not a short.
Check out Adamant IT 2 for the Pod Cast and More: / adamantit2
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 - Věda a technologie
You jinxed yourself when you called it easy in the beginning. 😀
A tip for locating bad power stage when you have several in parallel, is the measure resistance from gate ( both high and low) to the drain or source, the bad one will stick out like a sore thumb. It's likely the cpu is fine, about 2 ohms is within the good range.
this is the easy way, and you need to take it out always to test it, because the short on the gate to drain or source it will not let you turn on the laptop, and the chargers on dell, need to have the correct watts if not the laptop will not pass the 19v power rail or turn on any other power rails on
Im sorry to hear that you have been unwell graham,i hope you are starting to feel beter,please look after yourself.as allways ,a superb find,
with the lack of content, you should just upload videos about your repairs no matter how simple they are, at least for me im not a sub because the technical and unique repairs, i like your presentation and atmosphere, kinda what they say on some other yt channels i follow: "you can talk while showing paint drying and you still be interesting"
Very smart troubleshooting processes,
thank you
14:28 "Is it door number one?" Of course not, you never get that lucky.
In Telecoms a Low resistance is a heavy wire to wire , not bad enough to be a short/Loop yet . Hardest fault for us to detect as most adsl /pstn test equipment liked a short or open circuit one leg to be a better revealing test. Adding extra voltage while the premises end was off ( you don't damage their phones modems or alarms) helped break a low resistance further and made easier to detect- as a open or Short. That was all Dc Ohms testing btw
Hey adamant. Your channel finally got recommended back to me after such a long time xD. Glad to be back haha.
finally ended up throwing an ssd in my system, about 1 month back. Enjoying ever since. But yeah, after disabling memory compression for my ram, i understood its getting bottlenecked. Soo yeah i will need to upgrade it to 16gb. Lets see.
Dang, i wished that it never edp throttled. But yeah , laptop, nothing i can do about it. Sadly. Anyways Great Vid, and Have a nice day!
Lately, the manufacturers are prone to use these "power stages" (with all power built in...are mainly of 2 types, from ONsemi to Signetics -SICxxxx). And are exactly the same ISx gaming VGA boards are using. So be careful, and do not order one or two; a minimum of 10 are recommended, Adam.
That shine on the trackpad made a mirror out of it! :)
well I'd like to say that cpus are indeed a complicated thing especially when it come to their psu and it's Power management drivers... one thing we should know is the fact that cpus require more then one low voltage (0.8v to 1v) with higher / lower side MOSFETs . I'd say try hitting that coil with a 0.8v or 1v and set the amps as low as 2A or 3A auto and try powering it on again before deeming it a CPU failure. again, thank you for sharing and good luck Adam.
I don't understand the process you are referring to. You mean injecting 1V directly to the output of that coil? And what do you get with that? How do you tell if the CPU is good from there?
Thank you.
Great. Question: Could you get to this conclusion even without those 2 components in place? Awesome content you put out mate
Your videos always inspires me. My Asus laptop is dead now. I am gonna try fix it by myself.
Ah, right. It's never the CPU... unless it's a gaming laptop. That I can work with. ;)
It's never the CPU... unless it took 19.5v to the face because a power stage failed... in which case it's probably the CPU 😅
@@Adamant_IT well I'd like to say that cpus are indeed a complicated thing especially when it come to their psu and it's Power management drivers... one thing we should know is the fact that cpus require more then one low voltage (0.8v to 1v) with higher / lower side MOSFET ( you had a different thing you has probably a dual mosfet case in this "gaming" lol laptop ). I'd say try hitting that coil with a 0.8v or 1v and set the amps as low as 2A or 3A auto and try powering it on again before deeming it a CPU failure. again, thank you for sharing and good luck Adam.
your content is the best
I think the CPU is internally shorted based on how it behaved: the charger went off as soon as you plugged it in the first time, WITH the VRM still in place, because the VRM allowed B+ to go straight through the (shorted) CPU to GND. When you removed the VRM, B+ was no longer shorting out right away, because the VRM was now gone, so the charger stayed on, but as soon as you hit the power button and Vcore tried to go up, it was immediately shunted to GND through the CPU again, even though the other 3 VRMs are fine. Too bad....
VRMs - at least "fancy" ones as these - have over current limits. So even if the CPU would be shorted (which I don't think it is, since 2.5 Ohms is a fully normal reading for VCore), the VRM would prevent excessive current.
We live in a throw away society. So that's probably how manufactures build these systems, to get the user to buy another.
Hoping I can apply these principles to desktop motherboards, got a stack of dead ones would like to experiment with!
solder the negative wire on the on shorted power rail coil, then inject voltage so it won't go to the cpu
Great video
Had one of the the small smd caps near the cpu power fases blow out on a msi laptop. I tought for sure the cpu was dead but it luckily survived because it was in front of the fases which protected the cpu
The slight decrease in resistance seen at 14:55 was probably due to the board heating up from the iron.
isnt heat supposed to make the resistance go up?
@@dichoseadepaso That's true for metals, but semiconductors do the reverse.
@@man_eating_monkey oh i didnt know that! thanks
What a legendary job you just did! Hats off to you
Just thought id say thanks ! not really for this video, but one you did on getting wifi cards to work with a broken switch. id had to use a wifi dongle on my laptop for a couple of years as even changing the card didnt help, but a thin strip of tape as per your video worked ! differnt pin as different card, but now have working wifi again !............ thanks bud !...... winrar !!! lol :)
this NCP302045 is just an Integrated Driver and MOSFET in one package the real phase controller is NCP81215P and it's 4 + 2 + 1 phase three output buck if we are sure that this one is not shorted at least on that output it's not the first time a mosfet took a controller out or it might be that the chip checks phases before starting and sees 1-2-4 if it's not that then it's CPU and thanks for the great video as always
Thank you for those fantastic videos you make, can you please make a video to explain more on how did you conclude it one of the VRM ? and how you decide if you will inject voltage or not?
The CPU has an internal resistance of about 2-4 ohms - this is why all of the vCore components and capacitors will beep in continuity mode and look like they're shorted to ground.
So if you see the main power rail at somewhere between 2 and 4 ohms, that should immediately make you suspicious that this might be CPU resistance, meaning you have a shorted VRM. If this is the case, you can't inject power because that power will go directly into the CPU.
The easiest way to check if you have a VRM short is to measure the resistance to ground at a vCore inductor, and see if it's the same as your main-rail resistance. This is what we saw in the video, which means there's a very high chance of a VRM short.
Because there's usually only about 4 VRMs at most, it doesn't take long to manually desolder the inductors and see if the short disappears. In the video, it disappeared on the 3rd VRM, so now we know where the fault is without having to inject power. Sadly, if the CPU has seen 19v from the charger, the chances of it still working are extremely small.
GPUs are tough and will probably tank a zap, but the CPU is fragile and won't.
@@Adamant_IT Thank you very much for the detailed answer.
@@Adamant_IT is it true when measuring the resistance in the second mosfet or current sense resistor of inrush limiter protection faze if
• resistance < 1Ω short is in the main power rail and it's somewhat safe to injection
• 2Ω < resistance < 4 Ω: clasic signe of shorted VRM (your case ) or cpu failure
• resistance=0.9 Ω : gpu Vram or gpu failure of corse if there is a discrete /dedicated one in the laptop
I would check that chip if it has a pwr good or power ok pin . Also I would check if there power on the other 3 coils.
Couldn't you have injected a replacement voltage after removing the bad power stage, just to see if it turned on?
I've been thinking about this and couldn't you put the negative of the voltage injection at the vrm coil instead of GND to avoid injecting into the chip?
I might be wrong but i think that would inject power just through the vrm mosfets.
The thing is, if the short to ground is somewhere else then the injected power could come from ground through the chip back to the other end of the circuit.
Love the videos my friend! Would you be able to tell me what electric screwdriver your using?
Thanks in advanced ;)
It's a Sequre SQ-ES126 👌
@@Adamant_IT Thank you very much ;)
could you inject 1volt 2 amps into the main power rail and look for hot spots on the board with thermal camera
I had an HP elite book that was very picky about its power. It would only take power from a few chargers. The charger it came with wouldn't work after a while. Don't know what that meant.
"Is it Door No 3.." lol
6:27
"Spicy pillow" ahah
ghost at 22:01
Adamant IT
I've got my fixed last week by replacing one of the dead cap by power injection method (1.2V) to one of the MOS Fet around the power connector, same model G3 Gaming Laptop, it went succeed, u may try it again.
Your videos is verry good
Did you mount the RAMs back in when you took the final tests? I ask this because the keyboard backlight turned on, which means there was some POST but, without RAM, the POST will not continue.
Cheers!
I think cpu/gpu still working.
gfx and cpu power rails phase monitored by pwm ic, if pwm detected voltage/ampere above tolerance i.e faulty mosfet it will shut mosfet drivers to prevent damage to cpu/gpu.
The output rails usually have capacitors for filtering (solid/tantalum) this caps also act as extra protection.
pwm also has one pin dedicated to power good, it tell southbridge the vrm are ok then the SMC on southbride enable cpu/gpu to start.
you can tell if cpu/gpu exposed to high voltage, the output vrm filter caps will explode because voltage goes above maximum rating it can handle (max rate usually 2.5 - 3.3v)
Graham maybe you could connect your power supply to the cpu inductor after removing the shorted mosfet-driver and directly supply the required cpu voltage 1,8v (or whatever is the case to that cpu) to the missing phase and then try to power the laptop.
Note quite that simple... vCore is low voltage, _high amps_. Your bench power supply would need to be able to to 15-20amps, which is a lot.
@@Adamant_IT I'd try hitting that coil with a 0.8v -- 1v right after seting the amps as low as 2A or 3A auto ( it's a gaming laptop it can take 3A) and I'll try & power it on again. in a nutshell, tricking the cpu Vcore power rail . Vcore power is supplied, CPU is bamboozled thus pc is on. 1100010100110, Praise the Omnissiah.
@@drmoose7233is there any video in youtube someone does that?
At 12:09 the 4th chip(pu4802) at the second leg there is a damage? Or just the video tricks me?
You really need an silicon mat. That hard drive housing your using is robbing from your board temps.
Have you ever used a blacklight to check for odd spots on board? Sometimes it will show things that normally couldn't be seen
hello from a used laptop lcd? or what kind of blacklight
Good way to teach troubleshooting!
Question to all experts: are there other reasons, why the power rail becomes low resistance instead of being shorted?
The short is not directly to ground, but to CPU Vcore, which has a very low resistance to ground through the CPU. You can always have a short between different power rails, not just to ground.
Do you know if Ferrite beads are easily sourced I have been looking for a 500 Ohm 3Amp bead with no success. Are there others I can easily replace it with? Only asking because I have zero experience with chip ferrite beads, maybe they are like capacitors where I can replace them with the same uF and higher voltage.
I'm not knowledgeable at all when it comes to electronics and laptops. I just enjoy watching your channel and have been doing so for a while now. So my question might be an ignorant one. Is it not possible to replace a dead laptop cpu? I would think that would be a lot cheaper than to shelve, or bin, an expensive gaming laptop.
Hello. Which BIOS programmer do you use or Which programmer can you recommend to flash computers BIOS chip.
Hi .. power good is required from that power stage. Else it will not power on
I never repair gaming laptops, they run so much on the thermal limits, I cannot warranty the repair (can be destroyed again by user slightly blocking the airflow).
My PC Specialist/Clevo lappy does that to the charger, when I press the power button, but my IT friend doesn't know why. She thinks it could be the voltage regulator, but doesn't have the tools to investigate that possibility...
I always enjoy watching your work! Could I send you a stick pc for repair of the power connector on it?
If you're UK based, I can certainly try, postal address is on the website 👌
@@Adamant_IT Sadly not in the UK, I'm in Southeast Texas USA and can't seem to find a skilled pair of hands locally! Love your videos!
@@bigal1863 NorthridgeFix is in the US. Haven't used them but their videos show good work.
@@bigal1863 Try to contact Louis Rossmann in New York. He is expert in MacBook repairs but he might help you with others as well. Give it a try.😊 czcams.com/users/rossmanngroup
@@markusstrobl1067 I appreciate the referral Markus!
Ck with a dc bench how many amps its using and thermal cam
do they post without ram or wifi card?
my old all in one wouldn't post without the wifi card connected but when i put it back it would post but later crash due to it being broken and shorting out after a while
Hi, kept in count that you probably remounted the ram before the final tests, i think one pass lacked before giving a cross on to the cpu, and it is that the laptop probably stayed long time unused before came to you, the cmos battery was a bit low so a cmos reset might been useful, hi again, good channel..
I did actually pop a different BIOS battery in there while testing, so was being checked with a known-good cell in.
I had one RAM module in for testing, but it should've turned on with none in anyway.
@@Adamant_IT but what about cmos battery reset (s.c. terminals)?
I have this same issue with an optiplex 5050!!!
to be sure, you can put your termal camera and see if cpu hot ... or what is ... greeting from Portugal ;)
Nowdays laptops are using smart vrm's.
So only 1 or 2 vrm pushing the cpu, gpu, when idle and powering all up and giving power to each other. Like intel and nvidia said. The power sharing thingi.
Maybe it needs that stage. But yeah. 19v is still 19v. Fried cpu.
Question to all experts: how would the internal control (driver) and main power rail protective respond in each case when the silicon chip (CPU or GPU) has or hasn't experienced a 19V hit, indicating a short through silicon? Would a DC power cable be fast enough to react / protect the CPU, GPU, or PCB? what about the 3.3v power rail does play any role in all this. thank you
Just got a Dell E7450 that does this. Kills the adapter when connected to the Charging port. Got it with the the same charger you have there, but suspect its the wrong one.
Update: Found the shorted Mainboard capacitor and replaced it. The PC came back on and worked for a few minutes but went dead again after re-assembly >:( ...
Just Install a new vrm then you'll thank me later I've experienced this problem
Anyway you should avoid to power it up without desoldering the power stage. In this case you have the high side mosfet shorted ... if the internal control try to operating the two mosfet, when it open the low-side you have a dead short to ground of the main rail and it can explode damaging the pcb.
desoldering the power stage meaning what desoldering the high side mosfet or just the coil that diliver the power to silicon ? also wouldn't the internal control detect the absence of a power source for the CPU in such a scenario? And how would the internal control (driver) and main power rail respond in each case when the silicon chip (CPU or GPU) has or hasn't experienced a 19V hit, indicating a short through silicon? Would a DC power cable be fast enough to protect the CPU, GPU, or PCB
@@drmoose7233 That chip has the driver, the highside mos and the low side mos, all-in-one. If you have the low-side mos shorted (not this case), if you power it up you risk the drive open the high side mosfet and you have a dead short to ground. So, you should remove the entire power stage (single chip) to avoid the risk to do more damage.
Power supply for CPU and GPU are multiphases nowadays. So, it's possible that the system works even if you remove one phase. I have a gtx 970 at home who have 5 phases out of the original 6 (this because it was impossibile to rebuild the last one since the pcb has been seriously damaged) .... and it works fine (I just limited a bit the max watt available using power limit). Of course, it depends how the psu is designed ... sometimes it could require some modifications.
About the damage to the CPU: there is no a straight answer. You could have a shoprted cpu/gpu, you could have a cpu/gpu working badly, you could have ram problem (cpu/gpu has an integrated controller for the ram so, the voltage could it also this part of the chip ... and maybe you could think it's the ram but it's not). Could the protection be fast enough? Depend on the board and also luck. Anyway, usually is the main power controller who intervene in this case (the one who receive pgood signals from single power rails all over the board).
@@larrywildman4381 thank you for the valuable answers ...much obliged
Hi Graham how much amperage has your charger. DELL laptop are pique with low bios battery . I would change the bios battery too and do what more measurement before write it off.
I did test another BIOS battery in it, but I didn't make note of it because it didn't make any difference. It would've been worth pointing out though, because yea modern Dells do behave like this with a stuck-BIOS.
Recently i bought used gaming laptop for cheap but i start playing giving bsod "stop code critical process died" or unexpected store exception any solution for that?
Hi, my laptop 💻 does not work full the power led turn on but the fan not working.I open it and check this CPU coil and was all GND ..is it wrong sign ?
I thought if its shorted and u inject voltage whatever is shorted will take all voltage
I had a thought .when a mosfet it shorted and wich ever it's going through cpu/gpu and you want to inject power but can't.in this specific case cant we short the output side of the suspect mosfet to ground therefore bypassing the cpu/gpu then inject power.this will create a better short for the faulty component in this case .5 ohm and all the power injection will go though faulty mosfet to ground and not through cpu/gpu.never tried myself but seem valid
We've already figured out what the short is in this case - it's the VRM power stage. It's failed-short, so it's essentially behaving like a wire bridging the 19v rail directly into the CPU.
Do you wear any antistatic band or do you have any antistatic mat when you work on laptops? I look at your videos and never noticed anything like that.
Brain engaged. 50 points awarded
he wears a bracelet. Verge style
The power light is flashing on and off probably because the power adapter is not powerful enough? 65w brick on a 135w laptop?
It was a 90w brick, which should've been enough to at least turn it on. Most Dells can run on under-sized chargers in low-power mode.
Dell laptops will turn on and reduce the CPU power if a lower capable charger is connected,
Dear Sir ,
if it appears to be quite impossible for s.o. ordinary like myself , to obtain a scheme Mac Pro, components internal, I would promptly assume likewise ,in order to get a useful consequence ,when viewing Your esteemable collection of tut videos.
Got it ?
Best Regards
did you put ram back in for the power up test?
It shouldn't matter, the device should reach S0 (powered on) with no RAM - however, yes there was a single module in there.
My desktop cpu code out D0 or 00
And on display it says cpu not installed.
Mb Asus Maximus XIII Extreme. I know the 2 cpu I try to boot up w are good. I did get a little thermal Grizzly on socket.
What I did by mistake 2 times is plugging in the 24 pin to motherboard and as I was inserting seen the mb power up. I panic and turned off power supply. 2 times I did this. I not sure what distracted me. I usually diligent in power off and unplug as I build/repair.
Would this fry ruin the cpu socket.
Remember I tried 2 cpu 10900 11900. Both work and one in my other now. Cpu work but socket maybe in question.
Can 24 pin plug in as power on ruin cpu socket?
Could you not put the coil back on and perhaps bridge the area where the chip was?
You can't bridge in place of a voltage regulator...
what about studying the datasheet of that discrete so you can build one. I mean it's probably just a combination of two mosfets.
a bios battery should be 3volts so surely 2.98volts isn't that low
This is correct, but BIOS circuits are bad and can require 3v on-the-dot, so they're actually incredibly wasteful of a CR2032 which is 3.2 max, 3.0 nominal. 2.9 or over is normally good enough, but when you're diagnosing, assume anything below 3 is suspect.
lol i'd round of numbers. there 3 Volts. ; ) it's not noted as 3.00
So you won't be replacing that chip to see if that restores proper power to the CPU and thus revives the laptop ?
3 week wait for new chips, CPU's seen 19.5v and very likely dead... dunno. I'm taking a pass on it, but if I didn't have numbered time it might be worth a swing.
Hey.. probably no pwr goof from the vrm
Check the pin out of the vrm.. there is a 99% chance cpu ok... just now pwr good from that vrm
Like on a bq20 charger ic.
Im trying to figure out why the charger cuts out when powerbutton, gramham what you always say its never the cpu, except when it is! what im think is a component failed but didn't do a dead short
Dell laptops have a data line called OneWire going to the charger, which allows communication between the two - so the laptop can tell the charger to switch off because of a power fault.
Just a dumb thought, but could the battery be required for the controller to be satisfied? Battery is probably toast as you said, but the smart parts of it should still be working?
Is that multimeter reliable?? I have two of those. One is turning off as soon as I power it on. The other has battery contact problems and rotary switch issues.
Not had problems with mine, but it's a cheap meter, so I wouldn't take it into battle, per se.
I think that "turning off" thing is just a FLUKE :)
The ticking noise is coming from the cpu unfortunately. I recently uploaded a video with dead cpu with ticking. 😊👍
Did you reinstall the ram when trying to boot the laptop ?
Without ram it still turn on but don't boot
@@limwellz Not always. I've seen laptops not even wanting to turn on without RAM. Mostly ACER (i think, it's either ACER or ASUS)
@@HoboVibingToMusic it has power but won't post that's what i meant. the problem with this laptop is it doesn't want to get powered same thing with my mcbook pro
Hi Adam, I am a complete noob but im trying to fix a macbook air A1465 11inch (820-00164A). Orange light, no fan spin and wont respond to smc reset. Ive got 8.2v going on ppbusG3H. 3.10v out to SMC_reset L? Could this mean i need to replace the U5110 chip? Do you think its worth a try just taking it off the board?
8.2v on PPBUS_G3H means the SMC isn't running (needs to be 8.6) and 3.10 on SMC_reset_L is also a bit low, so yea I'm thinking the SMC is stuck in reset. I'd start by reflowing U5110 and seeing if it all comes good. Also check PP3v42 going into U5110 at pin 1 to make sure that's at 3.42v as well.
@@Adamant_IT Hi adam, I can confirm that there is a capacitor missing (C5127) which is connected to the U5100 via the "GND_SMC_AVSS" circuit? Remember im a noob so i cant be sure what im saying is electronically correct. Does this sound like I have found the root issue? Would you happen to know if a capacitor like this is an easy replacement? I do have a donor board but its from a Macbook Pro A1278?
Maybe a little early to call a dead CPU here. These laptops are so expensive a replacement part could be warranted, but I guess it could also be a waste of time.
If it had been the GPU, or if the power stage had shorted to ground, I'd say the odds are good that the chip were ok. But the power stage was bridged, and the CPU doesn't like getting zapped. So yea, unless you're happy to wait 3 weeks for new power stages for a hail-mary, it's not really worth continuing =/
hi where do you get the schematics from
Find the model number of the motherboard and google "model number schematic", and there's a good chance you'll find forum links. Quite a few repair discords servers for youtube channels like this where you can ask nicely as well, and people will check their collections.
@@Adamant_IT thanks
So, what happens now? Do you just tell the customer their laptop is toast? Would it be worth it to buy a replacement motherboard for this system?
Yea this one's being sent back. Replacement mobos are sometimes out there, but I generally don't go down that road with gaming laptops, because you never know the history of a second hand board, which is likely to cost a few hundred. Big roll of the dice for something you need to warranty.
@@Adamant_IT u get paid some? because ur wasted time i mean
Removing cpu and install new cpu.
👍
There's way too much paste on it ..take off the RAM cover and throw it away."
u should.try to fix it
But you don't need to inject 19v to figure out where is the short 😅
HOW can something Power Up when Part of Power Up Process is Missing? HOW does that mean CPU is Dead?
my non understanding idea: There is a VOID in the process, and power relies on completion of steps. ???
In the same way that a car engine can run with half the cylinders not firing. It's not going to run very well, but it can run.
@@Adamant_IT Oh, OK - Thanks. I've been watch videos & reading on 'electronics' electrical things for years. Truly Not my forte', lol but i find it Very interesting. Cheers!
shorted power rail... ouch... vrm back fire total dead new mobo..... it popcorns the traces and kills diods
flux....why are you so skimpy on flux!
Now people are out and about they're getting the ordinary flu. That's happening in Australia.
Is the ram soldered nope heheheh
Dan apa yang kamu katakan bisa sambil di tuliskan
Ist never the CPU, except....
Wrong video. I am looking for Q tips to clean
Those imbedded CPU's i hate - if the CPU goes bad you can't replace it with a good one - that's how the motherboard manufactures make more money
sei inglese?
Yes he is
@@gavinstirling7088 buddy
@@gavinstirling7088 I admire you ....collegue mate.....I do the same job here in italy ....i do servicing on tv and monitor and hifi equipment...fine business
listen mate have you never found any polimer cap failed on motherboards in your experience? which are the most common fails on motherboards you experienced ? which are the most common fails on pc's motherboards in your opinion and experience?