Alienware 15 r3 dead, shot circuit when plugging in AC adapter

I left my 15R3 running over night for calculations, in the morning it was off and the blue light on the adapter plug was off. Re-plugging the wall adapter when leaving the laptop unplugged does bring life back into the adapter (blue led in plug goes on). But de blue led dies again when plugging the laptop in again. You can also hear a shorting sound and smell a bit of ozone afterwards.

Searched on the internet a bit, can find a lot of similar problems on dell laptops, and heard it was commen problem around my manufacturing time.

I opened up the laptop and unplugged everything I could from the mainboard (Battery and all other cables I could find) but still no life. Unpluging the AC cable toward the mainboard and testing this separately does not show any problems, so the problem must be on the main-board I guess.

I made a thermal image of the mainboard to hopefully find what is shortening and found a IC with the code : MDU1512 HS9501GP to heat up very rapidly during the short time the AC needs to detect an shorting before going into safety shutdown. There are 2 of those on the mainboard, and only 1 is heating up.

Here the thermal Imaging :


I hope Alienware can help me with this one. This is my first Alienware but have had a number of problems with it : (lid hinge was replaced, screen was replaced, and now mainboard needs to be replaced?). I hoped Alienware could be my High-End system supplier for rest of my live that stands behind there products. But with my current experience I'm unsure if I will stay with this brand and advice others.

Anyone any advice? The laptop was bought end 2017, so I guess there is no "official" warranty. New mainboards cost easily around €1000,-. But expecting some good will from Alienware if this is indeed a known problem because of bad component quality.

 

Contacted Alienware, they only give me the options to replace it for €1300 euro. Really?!?

Earlier experience with Dell/Alienware wasn't good for me either. In first 2 month paint started chipping of from the powder coated hinges. Even though I had bought premium on-side warranty support it still took me many mails en calls before they replaced it. I'm sooo carfull with this thing, but the paint just rubbed from going into a soft-shell or even when you rubbed it with your finger.

Dell kept saying it did not impact the functioning of the machine so no warranty was needed. I needed to make a comparison to a just bought car where the paint started falling off. After that they tryed to let met pay €157,50 for repair on-site, even though I had already paid for premium-on site support.

A year later the screen started failing. I got a new screen under warranty without much problems this time.


And now @2,5/3 year old ~2 evenings a week 3D modeling/gaming the €1300 mainboard just shorts and they say I need to buy a new one.They feel zero responsibility for designing it in a way that gives very high risks for laptops becoming total loss if just a single mainboard component fails. They also don't want to look into repairing it... customers need to pay the full prince (incl. profit for dell I guess). I have no problem spending some money a year to keep it alive. But when this is not possible because of there design choices I expect them to really trust what they are doing, stand behind there choices, and have a graceful policy for the very few cases that do fail unexpected from an design / expected life standpoint.

I can't deny all to positive reviews and customers of older alienware's. It convinced me to go for Alienware too while I made fun of Alienware in the past as an DIY builder. Now that I work I just wanted the best for my work and free time. And there isn't really a DIY laptop build option. But in my experience they still build awesome ridged devices, but Alienware/Dell lost there hearts for it and the image that it carried.


For me it is has become clear that Alienware builds awesome devices, but doesn't put the same effort and heart into making it function for the customer and damages there reputation. (unlike for example toyota who still does call-backs on my 11 year old, out of warranty, car and replaces/repairs parts for free. Same goes for Lenovo business. I thought/hoped Alienware was up there with these guy's. )

 

Only two options for me left now :

Throw it in the bucket.

or

Try and repair it myself:
I orders already a few MOSFETs (MDU1512) by 3 different chinees factory's for a total of only ~€10. Since only one of the 2 light up like a candle on my thermal camera. I'm hoping this bright one actually has decided to shorted out, following its dream to become the next hottest thing in my house (think he doesn't know about my acetylene welder) .







I need to wait ~4 weeks for the first chips to arrive and will attempt an hot-air transplantation and see if that fixed the problem. Will come back here to post my results.
If anyone else has advice your welcome of-course.

 

Today I got my new mosfet chips. I desoldered first the one warming up, but the new one did exactly the same. Next I started looking at the other 4 MDU1512 chips. I discovered one was shorting out. Desoldered it, double checked him now it was no longer on the pcb, and indeed it was shorting out. Replaced it and tried again. But by the first try the new one directly shorted out again. After some debugging but hard to find cause I decided to replace all 4 of them to be sure they are good. But still same problem, one directly shorts closed.

After that I did some more digging to see if I can find an other cause. Tried several things together with the help of the schematic I found. (Added it to my post). At some point I decided it was a lost case, but still kinda wanted to know what was the problem. I attached a external power supply directly on the B+ rail (after the MDU1512 mosfets), and checked with the thermal camera what components started to get hot. But the only thing warming up was the NVIDIA GPU itself. Since I have no experience with debugging laptop boards on this level I can only guess a few things. But my main guess it that the NVIDIA chip itself is shorting out, or the GPU voltage controllers are shorting out but have so much capacity that I don't see them warming as much. I also would not expect the GPU to directly start up and creating heat as I only apply voltage to the main voltage rail from the battery

I wanted to try a lot more things but the very high temperature solder combined with a really good heat dissipating PCB gave me the greatest difficulty getting components lose.

So i'm gonna leave it at this, a little bit upset that I did not get to at least find out what component(s) caused the shorting of MDU1512. But my guess is somewhere in the GPU area also causing it to directly heat up the gpu instead of waiting for the "ON" signal.


I hope this might help others in the future. I will sell parts of my laptops for spare parts. Hit me up if ya need something.
And I hope to have warned people about Alienware not being a premium company. Alienware just doesn't stand behind there product as I thought they would. Making dangerous design choices and leaving the customer with the consequences. Any defect will need a €1300 pcb replacement, even faulty components from factory. It has become just an other Dell product with factory faults that Dell is not actually interested in knowing about. You simply need to add an other ~1000 euro to the already overpriced hardware in order to get a guaranteed 5 years out of it by buying that many years on warranty. But you will also need to send it for repairs multiple times and you also need to fight for your bought warranty. It is just a real disappointment to me compare to other brands like the Lenovo ThinkPad's or brands like Toyota and Dyson.

 

These Alienware 15 R3 / Alienware 17 R4 motherboards must be the most poorly designed laptop boards to come from a major OEM, I am super surprised that Dell has not been sued yet over these defective boards.

http://forum.notebookreview.com/thr...e-17-r4-6820hk-1070m-dead-motherboard.832607/

 

Check the capacitors on the board, sometimes they short out

 

Alienware is not good anymore. Especially after 2015 when they go full BGA. BGA = disposable.

 

Same happened with my 13R3. Left it plugged in overnight on sleep,woke up the next day and saw no light on the adapter and when I pressed the power button it smelled like burnt electrics. For some reason the RAM would heat up despite the laptop not doing anything. This was a fairly new motherboard as well since the last one was swapped only a few months ago. They luckily replaced it outside of warranty for free but I'm really worried about it happening again. I've had so many replacements and repairs with my 13R3 it's not even funny anymore. Funny enough my old XPS 15 L502x is still chugging along just fine without any issues.

 

Yeah I'm really disappointment to by the amount of repairs I needed what the type of repairs (all design / manufacturing shortcomings) and specially the difficulty to explain the situation and actually receiving help.

Update :
Shortly after last post I decided to give it one more shot, cause I wanted to at least know where the problem was. After a night of sleep it bugged me that the GPU would directly try to start for the short few millisec the adapter takes to detect a short. The GPU should only power on after you press the "ON" button I thought. But since the mosfets are all fine now the problem must be in what ever is ordering the mosfets to directly start. After a little digging I came uppon the charging IC chip "bq24780s". After testing some pins I noticed it wasn't doing anything I could understand based on the papers. So I orderd a new one and replaced it.

After this I finally could plug in the adapter again without directly burning the (new) mosfets or making the adapter go into protection mode. But I could also not detect voltage present after the first mosfet.
After hooking up an oscilloscope I found that the IC chip (bq24780s) is actually detecting a proper adapter and voltages now and is giving out a AC-OK signal to the rest of the system for just a few milisec. It also closes the first mosfet set, and then the second mosfet set to start charging and stuff, but then suddenly restarts as it now properly detects a shorting on the mail voltage rail.

I tried to figure out where the short came from on the voltage rail, but I can't seem to find it, and I don't have the proper tools or knowledge to debug it further. Some repair guy now bought over my laptop and is trying to fix it himself. He seems really skilled at it and is really nice to share a bit of his development with me. I don't know exactly what he all did but he got the laptop to react with lights and stuff on the power button again. There was a short in the CPU and GPU part what confused me and made me detect shorts everywhere and I don't know how to debug that. He found a dead CPU Vcore mosfet, and found that the whole NVIDIA GPU itself was shorting out. He actually is able to desolder the whole GPU. He things the laptop is now able to start up again but without screen since there is no GPU at the moment. I guess he will try to use an other GPU if he stumbled across one.

So yeah. turns out the Alienware did a full cascading self destruct on itself. From one failure somewhere the following things where destroyed as we know now :
- 2 mosfets
- Charge controller IC
- CPU Vcore mosfet
- NVIDIA GPU
- ?? unknown if there is more damage, like on the GPU memmory chips?

That is a lot of damage in my opinion. I thought every sub system had there own protection against shorts and stuff. It is just a guess, but I would guess the Charge controller IC is the cause of the problem, as it was acting really weird, trying to close all the mosfets directly power detection without giving AC-OK signal. One might even speculate it was programmed to self destruct X month after warranty expires, but I might just go to far here . It is just guesses and speculations and no solid prove. But it all still leaves a real bitter taste with the brand Alienware.


Hope this might help other peoples in either repairing there Alienware, or thinking twice before buying one.

 

I think it's fairly well established that if you purchase a BGA laptop if any component on the board dies, its a full MB replacement. You're the exception to that, obviously, as most of us don't have the skills or resources to do a Louis Rossman style deep dive into board-level repair of a modern laptop.

It's an accepted risk, I suppose; BGA means lighter and thinner assemblies which is a tradeoff from the old LGA/MXM solutions. They were heavier and thicker but gave us the option to upgrade or repair components that would otherwise be inaccessible. Manufacturers build what gets bought. If we want a different product, we vote with our wallets. Right now, the signal being sent to Dell is that this direction is okay as the market has moved nearly exclusively to BGA (and worse) solutions. We're seeing RAM being soldered down in Alienware laptops, which makes no sense to me.

Also, I think your diagnosis may be missing something; you mentioned cascading failure but those downstream components that got fried may not have been blown on the first failure. Is it possible those failures happened after repeated component replacement and subsequent power on attempts? My thought is maybe the damage wasn't initially that bad but without finding the initial failure point, the repeated power cycling caused additional damage?

 

@AcE Krystal Have you tried asking @Khenglish ?