Laptop died - Purism librem15

[handle]

Hello,
I trying troubleshooting my laptop, a purism librem15, which suddenly yesterday morning died, it doesn't charge the battery, and the power led is not lit.
I never tried to repair a laptop, but looking at the PCB, I was persuaded that the problem was just a burned out fuse. (https://pasteboard.co/JaVHnFq.jpg)
I started my troubleshooting convinced that that fuse could be the problem, but then I realized that despite a black mark it had on its surface, it was conducting.
Then I tried to verify a few things I learned reading here and there.
The MOSFETs after the power jack are conducting both, and I was able to read 19v on their drain.
The power rail coming from battery either is conducting, and I read 12v on all the MOSFETs involved with battery.
At the power button, I read 3v. (2.9 on my multimeter)
I read that I should have read 3.3v at the power button (not pressed) since the power rail 3VALW, feeding the SIO is such, and because the button is pulled up. In my case, I read only 3v, little less actually.
I was not able to identify SIO btw, I think it is embedded in the I7 companion chip, but I can't be sure.
I read somewhere the BIOS flash, in my case gd25q128c, is also fed by the 3VALW. The same source, for this reason, stated that I should have read on VCC 3.3v, but in my case, it was 0.
Can anybody give me a few hints on how to proceed with the troubleshooting? I also appreciate if someone could confirm my suppositions, or confute them.
Thanks for help

[Techtiger]

Re: Laptop died

you must change that fuse from a scrape Board what is your Bord ID to find out if the Schematic is available ? test when you press power button if the Voltage goes 0V to 3V?
test in bios chip pin 1 & 8 for 3V ?

[handle]

Re: Laptop died

Hello
Thank you for answering me.
Now is a bit late in my time zone, and I place the board to the chassis with all the screws to avoid losing them. Tomorrow I will take it apart again and test the button. I didn't try to push the button and then check the voltage on it again, but I'm sure the voltage was 2.99V, both using power from the charger than using the battery. At a certain moment, I also removed the RTC battery to see if the 3v came from there. As a side note, the RTC battery is critical low, and I read just 0.43v.
Anyway, shouldn't I read 3.3v on the power button instead of 3.0v?
You told me to replace the fuse, and I will do as soon as I recover a new one.
Do you know how to identify it?
There seems no schematics available for this laptop (purism librem 15v3), so I do not know which one I should buy.

The fuse is before the first input MOSFET, and I verified the two MOSFETs placed after the fuse to be active. I read 19v on the drain of both. In such a situation, shouldn't be safe to assume the fuse is ok? The black sign made me guess it is broken, but then reading a tension on the MOSFETs, I got confused.

I read the datasheet about the Bios chip (gd25q128c), and then I directly tested the pin 8, which is labeled as VCC; I read 0v.
Then after your suggestion, I also checked the other pins. I can confirm To have read no tension on any pin, both using the battery and using power from the charger(19v).
Don't know if this adds anything.
Tomorrow I will test what happen when I push the button, and I'll report it in a new message.
I will also look for any board id on the other surface of the board.
This is my board, (https://pasteboard.co/JaYaVsB.jpg) please look if there's any other test you think it makes sense.

[mcplslg123]

Re: Laptop died

You should test the voltage on gate(pin4) of dc-in mosfet. As far as your question of missing VCC on pin8 of main bios-that may be normal as newer mb has 3VALW in suspend logic, ie, it will apear after turning on.Do observe whether power button is going 3V to 0V and back to 3V on pressing it. You need to find out a schematic which has similar kbc and troubleshoot further as this specific schematic may not be avlb. Do replace the cmos cell with 3V.

[handle]

Re: Laptop died

Hello,
Thank you both for having answered my help request.

You should test the voltage on gate(pin4) of dc-in mosfet.

By no means I intend to contradict your answer. My question is just because I want to understand the logic behind that.
Several MOSFETs compose the input 19v power rail, I tested the voltage at each source to be at 19v, and at the drain to be 19v. I did not test the gates, because I assumed that if MOSFET is conducting, the relative gate should be high. Where my assumption is failing?
I will check the gate of each MOSFET, to make sure that gate up corresponds to MOSFET conducting.

Do observe whether power button is going 3V to 0V and back to 3V on pressing it.

test when you press power button if the Voltage goes 0V to 3V?

It took me some time, but I did it this morning. I can confirm to have 3v on idle (button not pressed), that it goes to 0v when the button is pressed, and then again, it returns at 3v when button is released. I tested this condition twice in a row, to be sure. Is this the expected behavior?

You need to find out a schematic which has similar kbc and troubleshoot further as this specific schematic may not be avlb.

For what is my knowledge, this laptop has the mainboard derived by Clevo U953. Not sure if this is easier to find, but I did not have any luck on both sides.
The motherboard is equipped with ENE KB3930, don't know how much this component is common in other mainboards, some hints on how to search schematics would be also appreciated.
Suppose would need to replace it, should I put any particular precaution when it comes to buy it?
My readings on the argument points to the fact that this component is firmware based; are this components commonly shipped with the firmware?
Or is it needed to put the firmware on new components?
If it's so, should I expect this component to have a proprietary firmware?
All info on the argument would be also appreciated.
If it can be used in any way, https://pasteboard.co/Jb2DJXp.jpg is a picture of the mainboard bottom side.

[mcplslg123]

Re: Laptop died - Purism librem15

You can use HP G6 quanta R13 mb schematic as it uses KB3930QF-A1 kbc. Its not programmable, so if need arises you can replace it from any other mb.

As far as gate voltage for DC-IN mosfets are concerned-I didnt suggested to measure all mosfets gate rather only DC-In mosfets gate. Here voltage at the gate is important as it gives an idea about the charging IC (As gate of dc-in mosfet is connected to charging ic). Well, consider the case -your dc in mosfet is shorted-you will still get 19V at drain. My suspicion at dc-in mosfet and or charging circuit is 'coz you mentioned you dont have any charging light lit up.

[handle]

Re: Laptop died - Purism librem15

I hope not to abuse your patience.
I'm not used to your naming convention, but I guess when you said dc-in MOSFETs, you were referring to the first twos right after the power brick connector.
To make myself clear, I included this schema I made. (https://pasteboard.co/Jbf9fxF.png)
Tensions at gates seem me to be fine; I read 24.4v at MOSFETs gates.
The battery charger IC is ISL95520, and it seems OK too. Voltage readings are coherent with my understanding, and the package condition does not have any evident problem.
Again, I want to understand things, so I question your suggestions:
Before you suggested me to check the gates tensions, I had checked one by one the mosfets source-drain continuity; if there was a short, I would have found, don't me?

the current conditions are the followings:

• 19v Power rail ok,
• 3V at the power button. It goes to 0v if pressed, and back to 3v when released.
• Battery charge led is off.
• power led is off too. (no led is lit on the board).
• Bios flash nor pin 8 is 0v

Now I'm blocked.
Could you please suggest me what I should check next?

[mcplslg123]

Re: Laptop died - Purism librem15

I might have overlooked your drain-source continuity point but still a mosfet can have short between source-gate but may read ok source-drain. So better to test a mosfet between source-gate/source-drain & gate-drain. I stronly suggest to check basic requirement of your kbc Ene3930 like VCC,AVCC,AD_AIR,MBCLK,MBDATA,EC_PWROK,LID EC. Btw, which charger IC is being used here??

[handle]

Re: Laptop died - Purism librem15

Hello,
maybe I do not put the necessary enfasis on this point.
The battery charger IC is ISL95520, does this component have any impact on the issue.
I looked for its datasheet, but I just find a sort version which contains just a reference design an a description. It does not even list pin functions. What kind of issues may caused by this component failure?

[SMDFlea]

Re: Laptop died - Purism librem15

The compal la-c482p has the same ISL95520 ,see page 37 of the schematic.With charger attached check voltage on pins 1,2,18,31,32.

Post #3 https://www.badcaps.net/forum/showthread.php?t=57294

.

[handle]

Re: Laptop died - Purism librem15

Thank you for your precious suggestions. I'll do all the tests as soon as possible.

[handle]

Re: Laptop died - Purism librem15

Hello,
I did manage to do some tests.
Not completely sure of their results, but I'll try to summarize them in the following:
ISL95520 - the battery charger:

• pin 1 (ACIN) - 3v
• pin 2 (ACOK) - 0v
• pin 18 (DCIN) -19v
• pin 31 (CSIN) - 19.2v
• pin 32 (CSIP) - 19.2v

ENE 3930
I also tested VCC in various places:
pin 96, 33, 9, and it was 0v
I understand that this might be the problem I was looking for, but if it's so, why do I have 3v on the power button if it is supposed to be pulled up by the same power rail which feeds the ENE 3930?
I do not have the exact schematics for the laptop I'm trying to repair, so I can't be sure this laptop (HP G6 Quanta R13)equipped with the same kb3930 can be accurate on this power button element.

If this my continuous requests do not bother you that much, I will do really appreciate any other suggestion.

[handle]

Re: Laptop died - Purism librem15

Hello,
Further investigation I made this evening shown that at least ENE 3930 pins 96, 33 (VCC) have a continuity test ok with GND.
This is suggesting to me that my problem could be a short somewhere in this power rail.
My question is,
Can you suggest a reliable way to find short circuits in such situations?
I do not know where to begin.

[handle]

Re: Laptop died - Purism librem15

Hello,
During the day, I overviewed a few pieces of literature on the argument, as you already know I'm not keen on laptop mainboards troubleshooting.
I read on a post that an initial step for fine-tuning a mainboard short circuit, is to measure the resistance between the various coils on the board and the GND.
This should indicate that the power rail where the coil is in is defective (shorted). It seemed reasonable, so I took some measurement and which I summarized on this picture (https://pasteboard.co/JbZMMy7.jpg).
The post also said that to assume defects resistance between coil and GND should be less than 5 ohms. Assuming that statement to be accurate, it seems like that I have at least two rails shorted. Is this possible?
Am I on the right way to find my defect, or should I take a different approach?

[mcplslg123]

Re: Laptop died - Purism librem15

Going by your pic, can you confirm what the coil PL2 and PL3 stands for. To me it seems like 3V and 5V coils respectively. If my assumption is right, then you have a short on PL2 coil. You need to remove coil PL2 and measure resistance on both the pads of coil. Do measure kbc vcc pins for short after removing the coil.

PS: If you have a dc bench power supply, it will take just a minute or so to find the faulty component.

[mcplslg123]

Re: Laptop died - Purism librem15

Hello,
I did manage to do some tests.
Not completely sure of their results, but I'll try to summarize them in the following:
ISL95520 - the battery charger:

• pin 1 (ACIN) - 3v
• pin 2 (ACOK) - 0v
• pin 18 (DCIN) -19v
• pin 31 (CSIN) - 19.2v
• pin 32 (CSIP) - 19.2v

ENE 3930
I also tested VCC in various places:
pin 96, 33, 9, and it was 0v
I understand that this might be the problem I was looking for, but if it's so, why do I have 3v on the power button if it is supposed to be pulled up by the same power rail which feeds the ENE 3930?
I do not have the exact schematics for the laptop I'm trying to repair, so I can't be sure this laptop (HP G6 Quanta R13)equipped with the same kb3930 can be accurate on this power button element.

If this my continuous requests do not bother you that much, I will do really appreciate any other suggestion.

PIN2-ACOK has to be 3V for charging circuit to work.

[handle]

Re: Laptop died - Purism librem15

Thank you for the kind answer.

I will investigate your suggestion during the day, hopefully.
In the meantime, can you please tell me which is the reason you didn't consider the low resistance on the coil near to the power button?
(In the picture upper right)

PS: If you have a dc bench power supply, it will take just a minute or so to find the faulty component.

I heard somewhere the technique you're mentioning.
But I'm not quite sure to know all the details needed to use it proficiently
Could you please suggest me some sources where I can learn more on this?

[mcplslg123]

Re: Laptop died - Purism librem15

It is very difficult to understand what signal relates to a particular coil from images and without schematics. Moreover, you have to fix things one by one. As far dc bench power supply is concerned, you need to have that kind of psu with 30V and 5A specification. there are many make and models available online on so many sites. You can put power to the shorted side of a coil(after removing the coil obviously) and see what hots up. Thats called something like "short circuit test". You need to be careful not to give high volatge/ampere to start with. 1V/1A to begin with and then graduially increase the ampere(not voltage). If you maintain these safeguards, it takes just a minute to find the shorted component.

[handle]

Re: Laptop died - Purism librem15

Thank you for the precious suggestions.
I think I will need time, probably days, before I could test my board in the way you suggested me.
Before I will be able to do that, I must purchase myself the equipment currently do not have.

Before the greetings, I want to ask you again about the coil.
I understand that just a picture can't be enough to figuring out a complex circuit defect.
But perhaps you considered something I do not get when you decided that the upper left coil low resistance was not an alarm whereas the other was.
I would be really grateful if you would share these considerations with me.
Regards