HP Victus 16 Hall Effect Sensor Problem

[mitchw]

I am repairing a HP Victus 16 with black screen, blinking screen, etc. Quick googling provided me the answer that the lid detect hall sensor is the issue. Confirmed with disconnecting the IR board (that contains also the hall effect sensor). Laptop itself is just over a year old, but out of warranty.

Apparently the sensor itself is a TCS40DLR, which fails after just months of use at elevated temperatures.
Now which one is the best to replace with?
Some suggested A1126LLHLT-T, but they had to modify the power supply to supply the 3.3V instead of 1.8V, as the sensor wasn't working. The A1126LLHLT-T datasheet specifies 3V as minimum working voltage.
Which is weird, as the original TCS40DLR requires at least 2.3V to operate. If they used 1.8V to power the chip that needs 2.3V, then it makes sense it will start to work erratically when the silicon degrades after some time, as it was barely working when new.

Which presents a question, perhaps the original TCS40DLR will work just fine (and for years to come) if one just powered it from 3.3V instead of 1.8V?
Or just get another type of hall sensor that is rated to work at 1.8V VCC?

The thing is, that the customer needs the laptop fixed soon, but the new hall sensor will take at least 10-14 days to get here.
Should I try this modification before I get the new hall sensor?

[mon2]

Can you link a schematic?

If the vcc of the component is 1v8 then they are mating likely with gpio pins that must also be 1v8 compliant. That is, if 3v3 for the vcc is permitted then the gpio voltage will escalate to 3v3 and possibly kill the target device using this output pin. Can review in the schematics.

[mitchw]

Will try to find a schematic, I hope it is available at all as the machine is almost new.
I was referring to this post: https://www.reddit.com/r/HPVictus/co...thread_laptop/

I would think that a device using 1.8VCC will also output 1.8V GPIO signal, and 3.3VCC will output 3.3V signal... makes sense.

[mon2]

Ok. Just skimmed through the linked webpage and the referenced datasheets. The hall effect sensor is open-drain output. This simply means the output will be ok with a logic low output but will float the line to permit an external pull-up voltage to soft-park the line to the desired logic '1' state. That is, consider this device like a sensor with a SPST switch. When closed, the sensor is driving the line to a logic low = '0'. When open, there must be an external pull-up resistor to park to a '1' state.

For this reason, the VCC of this part becomes a no-care. The Allegro part is a better choice since the operating temp is much wider and is automotive grade = ideally more robust. However, the VCC for the Allegro part is 3v or higher which may be a pain as a drop in.

Suggest instead to find an automotive grade sensor that is also SOT-23 (need to confirm this size or is it SC-70?). Pinout should match and VCC should be 1v8 BUT the temp should be again, improved over the original. Then you will be 100% drop in replacement ready.

One company we love and use for our many product designs is NVE. They are a leader in GMR designs and fairly sure they will have some great hall effect sensors for this case. Will review the options soon and update this thread.

[mon2]

Ok, the following *should* (my disclaimer) work as a drop in replacement:

TMAG5131-Q1 Automotive Low-Power, High-Precision, Hall-Effect Switch
* up to 125C (automotive Q1 suffix)
* works @ 1v8 for Vcc
* open-drain output like the original

https://www.ti.com/product/TMAG5131-Q1

Canadian Mouser link:

https://www.mouser.ca/c/?marcom=129467300

If you do try this as a replacement, do update this thread with your progress. It is always appreciated.

[mitchw]

I got the reply from the customer, it can wait till next week.
So I will be ordering a TMAG5131-Q1.

Which one should I select? They all differ in Operating Point Min/Max and Release Point Min/Max (Brp).
If I get one that needs a stronger magnetic field, then it might not work. But if I get one that is too sensitive, it might trigger on its own.



Anyways, I found the specs for the original TCS40DLR sensor, they have them listed on Mouser:
Operating Point Min/Max: 3.4 mT
Release Point Min/Max (Brp): 2 mT

So the closest will be this: TMAG5131C7D
Operating Point Min/Max: 2.4 mT to 3.8 mT
Release Point Min/Max (Brp): 0.8 mT to 1.9 mT
Output Type: Active-Low, Open Drain

Can you 100% confirm before I order?

[mitchw]

Anyways,
I think I found the real culprit. The Hall sensor turned out not to be the root cause.

Firstly, the sensor is powered by 3V, not by 1.8V. That was confirmed by measuring the pins on the JIR1 connector going to the IR/Hall sensor board. The output pin is also at 3V, that goes to 0V when hall sensor is triggered.

But when I measured pins on my laptop, I found one quirk. The VCC was 2.8V, then slowly going down to about 2.2V, and jumping all around. Something that should not be happening.
I then stumbled across a forum post that said, check the FU6 if experiencing problems with lid detect. That FU6 is a tiny 603 sized fuse near the JIR1 connector, and on one side I got stable 3.05V, but on other side I got ~2.3-2.4V. So something not right, fuse should not drop any noticeable voltage!

That FU6 is blown, but not 100%, as it still measured around 20K, which is apparently enough for the hall sensor to work sometimes.

As the value of the fuse is unknown (probably 1A?), I decided to slap across it a tiny 603 0-ohm resistor. It should blow up at around 2-3A, but my guess is that the FU6 actually got killed by surrounding heat, not by overcurrent, as the hall sensor draws few mA at most.

And what do you know, with new "fuse", the laptop now works with no issues, with the original hall effect sensor!


So, for now my theory is that the actual culprit was the FU6 that just opened, but not completely - it became a resistor of few K ohms. This dropped the voltage to the hall sensor below the operational voltage, which made it erratic - makes sense! A fuse that stops being a fuse, but not quite.
This is why the reddit poster had to wire up the different VCC to the new hall sensor as it still wasn't working (but it have should have worked!). And this is why he measured 1.8V. He was chasing the wrong suspect all along... he just had to replace the fuse.

[mon2]

Excellent work!! Very much appreciated.

[tomodachi]

Anyways,
I think I found the real culprit. The Hall sensor turned out not to be the root cause.

Firstly, the sensor is powered by 3V, not by 1.8V. That was confirmed by measuring the pins on the JIR1 connector going to the IR/Hall sensor board. The output pin is also at 3V, that goes to 0V when hall sensor is triggered.

But when I measured pins on my laptop, I found one quirk. The VCC was 2.8V, then slowly going down to about 2.2V, and jumping all around. Something that should not be happening.
I then stumbled across a forum post that said, check the FU6 if experiencing problems with lid detect. That FU6 is a tiny 603 sized fuse near the JIR1 connector, and on one side I got stable 3.05V, but on other side I got ~2.3-2.4V. So something not right, fuse should not drop any noticeable voltage!

That FU6 is blown, but not 100%, as it still measured around 20K, which is apparently enough for the hall sensor to work sometimes.

As the value of the fuse is unknown (probably 1A?), I decided to slap across it a tiny 603 0-ohm resistor. It should blow up at around 2-3A, but my guess is that the FU6 actually got killed by surrounding heat, not by overcurrent, as the hall sensor draws few mA at most.

And what do you know, with new "fuse", the laptop now works with no issues, with the original hall effect sensor!


So, for now my theory is that the actual culprit was the FU6 that just opened, but not completely - it became a resistor of few K ohms. This dropped the voltage to the hall sensor below the operational voltage, which made it erratic - makes sense! A fuse that stops being a fuse, but not quite.
This is why the reddit poster had to wire up the different VCC to the new hall sensor as it still wasn't working (but it have should have worked!). And this is why he measured 1.8V. He was chasing the wrong suspect all along... he just had to replace the fuse.

Great explanation, thanks !

[ragutom]

Anyways,
I think I found the real culprit. The Hall sensor turned out not to be the root cause.

Firstly, the sensor is powered by 3V, not by 1.8V. That was confirmed by measuring the pins on the JIR1 connector going to the IR/Hall sensor board. The output pin is also at 3V, that goes to 0V when hall sensor is triggered.

But when I measured pins on my laptop, I found one quirk. The VCC was 2.8V, then slowly going down to about 2.2V, and jumping all around. Something that should not be happening.
I then stumbled across a forum post that said, check the FU6 if experiencing problems with lid detect. That FU6 is a tiny 603 sized fuse near the JIR1 connector, and on one side I got stable 3.05V, but on other side I got ~2.3-2.4V. So something not right, fuse should not drop any noticeable voltage!

That FU6 is blown, but not 100%, as it still measured around 20K, which is apparently enough for the hall sensor to work sometimes.

As the value of the fuse is unknown (probably 1A?), I decided to slap across it a tiny 603 0-ohm resistor. It should blow up at around 2-3A, but my guess is that the FU6 actually got killed by surrounding heat, not by overcurrent, as the hall sensor draws few mA at most.

And what do you know, with new "fuse", the laptop now works with no issues, with the original hall effect sensor!


So, for now my theory is that the actual culprit was the FU6 that just opened, but not completely - it became a resistor of few K ohms. This dropped the voltage to the hall sensor below the operational voltage, which made it erratic - makes sense! A fuse that stops being a fuse, but not quite.
This is why the reddit poster had to wire up the different VCC to the new hall sensor as it still wasn't working (but it have should have worked!). And this is why he measured 1.8V. He was chasing the wrong suspect all along... he just had to replace the fuse.

Excellent work!! This info will help many people facing this issue. I am going to link this into my original post and strikeout the trace cutting. I was not a fan of the trace mods to begin with, but I figured all of the supporting circuitry would be on the opposite side of the motherboard, which I was trying to avoid fully pulling.

However, I am not convinced that the heat is damaging that fuse. I would expect the IR sensor (where I tapped the VCC from) to be powered through a similar fuse nearby... but it never gets damaged from the heat? That doesn't seem likely. The Toshiba part was not designed for this environment and it is stated in the datasheet that the part will do wonky things when exposed to environments that aren't even outside the suggested operating ranges. I believe the Toshiba sensor does something nasty in the presence of high temperatures and causes the fuse to blow. That fuse is likely only 20-50 mA at the most (the absolute maximum rating in the Toshiba datasheet is 10mA current draw so that fuse has to be ultra ultra low). The sensor itself survives, but the partial damage to the fuse is already done.

Based on this theory...a word of caution: On your customer's laptop, you should still replace the sensor with the TMAG or the Alegro and not leave the original in there, even though it works. Since you have put a 0 ohm resistor across that point, there is effectively no protection in place anymore for a very high probability failure. If this is caused by the sensor, your next repair on this laptop will likely be a larger, in-depth repair to the power supply or traces. I would not risk that possibility! Both of us are currently only theorizing about the cause. We don't have any facts to back up either way.

[mitchw]

The best way to solve the argument about the fuse rating, would be to check the schematics, which I doubt they exist for this model, at least for now.
If someone has them (or for some other Victus 16 / Omen 16 laptops - even with Intel CPUs), let me know, would love to check the FU6 rating.
However, I have never seen such a low current fuse you are describing on a laptop mainboard. They do exist, though.

I still do think that the hall sensor is fine, as the broken fuse was reading about 20k ohms, which caused for the voltage to drop to about 2.5V-ish, which means the hall sensor was drawing about 0.15mA of current (150nA!), which is extremely low. Also I don't think that the hall sensor gets to 85 degrees C, even under heavy loads. Perhaps I'm wrong, but the only way to be sure, would be to measure the temperature using a temperature probe placed on the hall / IR sensor PCB.

That IR sensor is powered from different voltage source that gets shut off when you turn off the laptop, but lid switch sensor is powered on all the time, so you can set the BIOS to power on the laptop automatically when you open the lid. Since you cut the trace, that functionality now will not work. As the lid sensor is energized even when the laptop is off, they most likely added some protection just for the sake of it (or if someone disconnects the JIR1 connector with the battery still connected, etc.).

That 0-ohm resistor will also act as a fuse if needed, but will take about 3A to blow.
There is almost no way this will damage anything, as all power supply circuits in laptop have current sensing/limiting and will just go into protection if any of them detects overcurrent.

I was thinking that even better way would have been to replace it with 100 ohm resistor, that would limit the fault current to 30mA if hall sensor shorts, which is a win-win situation. As the hall sensor is drawing just under 0.2mA when operational, the voltage drop will be negligible - remember the hall sensor was still kinda working with 20K "resistor"! I really don't know why they didn't just put a resistor instead of the fuse in the first place.

Also, I gave the customer 12 month warranty on repair, so if the problem happens again, I will be updating this post about it

[mitchw]

One more thing, if the fuse blew like you are saying, then it would open up - became an open circuit, and the hall sensor would just stopped working completely.
In this case, the fuse actually degraded / failed, as it was still allowing current through - resistance went up to 20K or so. Blown fuses don't do this. If the fuse was really blown, I would measure 0V on the other side, not the 2.8V. This looks much more like thermal degradation than a clean overcurrent event.

The fuse just fails because it gets cooked by the heat. It is never a good thing to put a fuse on part of the mainboard that gets scorching hot near the heat pipes. Those tiny 0402 fuses are only rated for about 100 degrees C, and the PCB gets almost that hot. Repeated thermal cycling can fatigue the fuse element over time. Instead of snapping cleanly open, it appears the fuse element partially degraded and turned into a high resistance path.

[JarJarBinks]

I wanted to add to this that I ran into this problem today. My fuse showed millions of ohms. But when I tested the voltage one side was 3.35v and the other side was bouncing around from 1.0v to 2v and sure enough replacing that fuse fixed the issue. Would of never figured that out without the help of this post...so thank you very much.

[thorcito]

I wanted to add to this that I ran into this problem today. My fuse showed millions of ohms. But when I tested the voltage one side was 3.35v and the other side was bouncing around from 1.0v to 2v and sure enough replacing that fuse fixed the issue. Would of never figured that out without the help of this post...so thank you very much.

Hi, I have de same problem. Can I short de fuse with soldering paste?
Thanks in advance.

Edited: shorted and it works.