ASUS ux410u - No power no charge

[alexvan59]

Hi,

I have an ASUS ux410u that is no boot no power. I'm trying to find out the issue and I'm seeking for help. I'm not a specialist, I have a background in electronics from engineering schools but nothing more.

The good news is that I have another ux410u that is working perfectly, so, I could compare if needed.

I've disassembled it and isolated the mainboard of the defective one and made some checks:
* the power port is ok, I get 19.4v on the board
* After the two first mosfets, I still have 19.4v
* Without doing any voltage injection (just the official power adapter), a component is getting super hot: I can feel it with my finger. With alcohool and microscope, it's pretty clear that it is this component.
* The hot component is 2N=4a... which seems to be Richtek RT8249C: https://www.richtek.com/assets/produ...8249ABC-02.pdf
* For sanity, I tried to find a short around the hot component but no "success".

I've made some measurements and have some questions:
* my understanding is that RT8249C should do linear output 3.3V on pin 3 and linear output 5V on pin13. My measure are not even close to that. For that chip, I get 19.3V for vin (pin12), ~0.8v for pin 3 (vs expected 3.3v) and roughly 0V everywhere else. Pin 13 (5V) is giving me ~0V but is not grounded.
==> I guess that getting 0.8v rather 3.3 can let me think that there's a short on the powerline. Is that a correct hypothesis?
==> How could 5V becomes 0v?
==> Is it common to have this chip defective? or is it more just a consequence of another short on the mainboard?
* During my sanity checks, I tried to measure most mosfet that could be connected to the 19v powerline. One of them (which seems ok) has one pin to 25v. Is it normal? I would expect to have nothing more 19v. (maybe a newbie question )
* Most mosfet are 4N VUC. My research let me find that it could be PE528BA. However, I'm not able to find out what VUC stands for? Are 4N VUB and 4N VUC the same components? https://datasheetspdf.com/pdf/1094106/UNIKC/PE528BA/1

I'm looking for any guidance, any advice. I'm really doing that only for few weeks so, I'm clearly not an expert. I can provide pictures if it can help.

Thanks
Alex

[techtiger]

For the Fet on the +19V line it needs the gate pass to switch on more than +22 to +28V ( N-Channel Fets ) is need to gate by pass the Voltage, for the Richtek RT8249C pin 6 & 20 voltage ? what is the resistence to GND on the Colis of 3/5 ALWS voltage , find out the Bordview or close on 1

[mcplslg123]

Yes, your +3V LDO is shorted somewhat. Measure resistance on PIN3 as well as Pin 13. Then remove the chip from motherboard(2N). Now measure resistance on both pins again. Compare Values. If the resistance goes significantly up, most likely the chip itself is bad. Else, you need to inject some voltage on those pins to find the shorted component. There is bright chance of EC being bad.

[alexvan59]

Hi techtiger and mcplslg123,
First, I want to thank both of you for your reply, it's appreciated .

I have attached some pictures took with my phone (sorry for the quality).

@techtiger:

• I didn't find the schematic for boardview. I believe they are not available for this board

• For voltage, I have: 0.8 for pin 3, 0 for 13, 0.175 for pin6&20.

• I don't have neither schematic nor voltage on coils. I'm not sure to know how to identify the coils you mention. What does ALWS stand for? (sorry, I'm still pretty "new" in this field). I would guess that these are the coils around but I'm not sure.

@mcplslg123:

• I've measured the resistance between these 4 pins and ground. My current results are: 30kohms on pin3, 99.2kohms on pin 6, 505kohms on pin 13 and 9.86kohms on pin20.

• I will remove the component this week-end and compare with these values. My understanding is that, if I see huge changes on these values, it would be a strong indicator that this chip is bad. If I need to do short injection, my plan would be to remove the component first, and then inject 1V (much less than 3V and 5V) on the pin that doesn't change so much, is that correct?

• You say: "There is bright chance of EC being bad." What does EC stand for?

Quick question: if I do voltage injection, should I put the cooler back on the cpu? just in case?

[alexvan59]

Answering to myself at least on one question: EC stands for Embedded Controler.

[techtiger]

Originally Posted by alexvan59

Hi techtiger and mcplslg123,
First, I want to thank both of you for your reply, it's appreciated .

I have attached some pictures took with my phone (sorry for the quality).

@techtiger:

• I didn't find the schematic for boardview. I believe they are not available for this board

• For voltage, I have: 0.8 for pin 3, 0 for 13, 0.175 for pin6&20.

• I don't have neither schematic nor voltage on coils. I'm not sure to know how to identify the coils you mention. What does ALWS stand for? (sorry, I'm still pretty "new" in this field). I would guess that these are the coils around but I'm not sure.

There are different states of MBD Power in SLP_S5# state ALWS (all ways) Voltage will be active depending on the platform in your case RT8249C is the Regulator which output 3/5 ALWS + LDO 3/5

EC embedded controller SIO (super input/output) KBC
microcontroller in computers that handles various system tasks

[alexvan59]

Thanks for the feedback. I knew what super io was but didn't know EC could be the same component.

So, I did some more tests and I believe I'm making progresses. The facts I have right now are:

• my main power rail is OK (19.6V)
• my RT8249C is getting hot. Its pins doesn't get correct voltage. 0.8V on pin 3 and 0v on pin 13. I don't know what would be the expectations on the pin 6&20 but 0.175v doesn't seem normal.
• My EC SIO is ITE IT8995VG-128. Under "normal" operation(charger plug in), I don't notice any hot on it.
• I've finally found at least 3 capacitors that seems shorted around my EC SIO. The impedance is around 1ohms on these capacitors. These capacitors are not in the exact same space but around my EC SIO.

These facts let me think that my EC can be dead (as mcplslg123 suggested). Before going further (attempting a replacement), I'm looking for a proof of that.

I'm considering doing that:

1. keep all chips on board
2. do voltage injection on pin3 of RT8249C. Voltage for injection would be ~1V and I would start with 1A and increase the amps if I see high consumption until I get something that is getting hot.
3. if previous step doesn't give anything, try to do the same but with pin13

Is it a good approach or not? I mean, I know it's what was suggested. What is unclear to me is whether I should do this voltage injection with or without the RT8249C on board.

I've made some researches on my EC and I know for sure that changing it is not so easy for an hobbyist like me (soldering and reprogram). I want to make sure I have the right diagnostic.

[Spider1211]

do voltage injection on pin3 of RT8249C.

There is no point injecting voltage on pin 3 since you measured its resistance to be 30Kohm and therefore is not shorted.

You mentioned measuring 1ohm across 3 capacitors. This is an indication of a short. I would instead inject voltage (starting with voltages and current you mentioned) directly across the capacitor and see what gets hot. Just make sure to identify which side is positive and ground.

If nothing gets hot right away, try increasing only the current first. Then increase the voltage if needed while making sure not to go above 2.5 - 3V.

Examine the caps you said had low resistance physically. Usually shorted caps have slightly different color than the rest. Although I suspect since they are close to the EC, it could well be a dead EC as mcplslg123 suggested.

[alexvan59]

Ok, will try that tomorrow.

Currently, on the + side of these 3 caps, I get ~0.18V and 1 ohm to GND.

[alexvan59]

Adding a picture to show the shorted caps (in case somebody else get a similar issue in the future).

I have another caps behind the board that seems totally shorted (less than 1ohm on both sides of it).

[alexvan59]

Hi,

Made the voltage injection this morning. Here are the results:

• I had nothing with 1V and not so much amps were consumed (less than 1A). Decided to increase the voltage.
• No better results with 1.5V.
• I started to feel some heat with my finger around 2V. Still less than 1A. Decided not go further.
• Made the videos of alcohol on the EC => injected two times: https://youtu.be/e4m7ZoyPm3Q
• Came to the conclusion this chip is faulty

The chip ITE IT8995VG-128 is cheap (on AliExpress). The programmer is actually expensive (~100€ on ebay). Still worth it if it enables me to save this laptop. Since I don't know whether the other parts of the laptop are ok, I would like to delay the purchase of the programmer as much as I can.

My plan would be the following:

1. Remove the chip from the board (since I don't need to extract anything of it with the programmer, right?) and validate I don't have short anymore (not plugin in, just checking with multimeter the caps are no more shorted)
2. Order new chips (~5 parts in case I fail to replace or I get faulty ones)
3. Replace the chip on board and validate I don't have short and, on 19v, the RT8249C has correct behavior (no hot, correct LDOs)
4. Power on the system. I expect a black screen
5. If everything is ok, order the programmer and make the actual programing

Is it a good plan? Is it risky to replace the chip without programming it and powering the system? If yes, what can I do to delay the purchase of the programmer to maximum?

Sorry for my long messages, I'm trying to give as much information as I can to enable decision making.

PS: I have rework station, flux, etc. Never changed BGA on a computer mainboard on but I have changed some other chips like "tristar" on iphone.

[Spider1211]

You can actually buy the chip pre-programmed. This seller did come recommended in a few places (https://www.ebay.com/usr/pitstopcomputers?ul_noapp=true). And they do reply to messages fairly quickly.

Try contacting them with the chip you need and your laptop model.

My plan would be the following:

1. Yes, correct

3,4. You could do that, just keep in mind that some of the chips labelled as "new" on aliexpress are parts that have been pulled from other devices and may already have a program in it. Not sure to what extent that could be harmful (if at all) when soldered on a different device.

[alexvan59]

Originally Posted by Spider1211

You can actually buy the chip pre-programmed. This seller did come recommended in a few places (https://www.ebay.com/usr/pitstopcomputers?ul_noapp=true). And they do reply to messages fairly quickly.

Try contacting them with the chip you need and your laptop model.


1. Yes, correct

3,4. You could do that, just keep in mind that some of the chips labelled as "new" on aliexpress are parts that have been pulled from other devices and may already have a program in it. Not sure to what extent that could be harmful (if at all) when soldered on a different device.

Great, I will do step one. I've contacted the seller too. Will keep this thread updated.

[mcplslg123]

Programming this chip is not easy either-so the best bet is to buy pre-programmed chip from a reliable store.

[alexvan59]

Ok, will prefer the option of buying preprogrammed if I can.

So, I did 1 and so, removed the chip:
* the three caps around the ec are no more shorted
* one cap behind still totally shorted (see picture)

I removed the remaing cap to see if it could be it. However, the short remain. Since I really don't know which side is + (0 ohms and was 0v), I'm a little bit lost on how to find out what could generate this short.

The bad news is that, while removing the chip, 2 pads are gone => picture. This is where I start to have doubts I have the right skills/equipment to repair this board. I mean, I know that it's not game over to have pads gone, but BGA + missing pads, maybe I'm too newbie yet

[Spider1211]

What is the reference of the IC in close to the cap? Look for its datasheet and use its pinout as reference to find which pad of the cap is ground. Looking at the picture, I would guess the top pad of the cap is ground.

For the BGA maybe you pulled it before it was fully melted, but the damage does not look bad. Its easily fixable with jumper wires. Just make sure to properly secure the wire once soldered (ideally you should use UV curable soldermask paint).

Do you have schematic and/or boardview for your motherboard?

[alexvan59]

Originally Posted by Spider1211

What is the reference of the IC in close to the cap? Look for its datasheet and use its pinout as reference to find which pad of the cap is ground. Looking at the picture, I would guess the top pad of the cap is ground.

For the BGA maybe you pulled it before it was fully melted, but the damage does not look bad. Its easily fixable with jumper wires. Just make sure to properly secure the wire once soldered (ideally you should use UV curable soldermask paint).

Do you have schematic and/or boardview for your motherboard?

• Nice approach. Seems obvious when you say it but I didn't think about it
• Maybe it was not fully melted. First time it happened to me . I try to never put too much heat (I've trained my skills on many "dead" boards before). I will look at how to repair them once I'm sure it's worth to try. For the moment, as long as I have a short, it seems a waste of time to fix the chip.
• can't find the schematic/boardview, seems unavailable

[Spider1211]

Different boards have different thermal mass which affects how long it takes for the solder to melt. What I usually do (especially with bga) is to nudge the chip first, if it moves and goes back to its initial position you know all the balls have melted and is safe to remove the chip.

Mistake happens, that how we learn and improve our skills Up till now, we know for sure the chip is dead. So you will definitely need to be buying a replacement anyway. Might as well have it on order while you continue your troubleshooting.

Try to look on the forum or post a request in the schematic section.

[alexvan59]

The chip close to the cap has "548 TI 711 ATCK". Didn't find the datasheet yet.
Just to be sure: you believe top pad is ground because we have a huge track or is there any other reasoning?

[alexvan59]

Originally Posted by Spider1211

Different boards have different thermal mass which affects how long it takes for the solder to melt. What I usually do (especially with bga) is to nudge the chip first, if it moves and goes back to its initial position you know all the balls have melted and is safe to remove the chip.

Mistake happens, that how we learn and improve our skills Up till now, we know for sure the chip is dead. So you will definitely need to be buying a replacement anyway. Might as well have it on order while you continue your troubleshooting.

Try to look on the forum or post a request in the schematic section.

Actually, I bought this computer 10€ to a company that wanted to trash it : I didn't test the screen but I strongly believe it's good => so whatever happens to the mainboard, I'm good. I bought it only to learn and to see if I can avoid ewaste .

Thanks for all your tips, it really helps me.

I don't give up until I'm sure it's total game over