Hoover FM216LI cordless vacuum refuses to charge

[flat-earther]

I have a Hoover FM216LI cordless vacuum which isn't charging its internal battery.

I opened up the unit, and it has 6x Samsung ICR18650-22P cells connected in series.
5 cells were at about 3.3-3.1V. one was at 2.4V, so I assumed this cell was bad that's why the vacuum wasn't charging so I replaced it with another 18650, but after that the problem still persists, the vacuum turns on a bit and then cuts out since the batteries are empty.

So I think the over discharged battery wasn't the culprit, it was a symptom of the unit not being able to charge its batteries.

The charger is an external AC to 26VDC charger with a barrel plug.

When I opened the vacuum I noticed that the PCB that controls everything had a lot of dust buildup on it, (I already cleaned it and didn't take photos of the dust).

This is the PCB after cleaning:


When I was doing some probing on the board, one of the 0603 68K resistors near the power switch (this area had a lot of dust buildup) didn't give any reading and then I noticed that the solder joints of that resistor crumbled from my sharp probe tips, what does that mean?:

After repeated probing of the resistor I managed once to get a 68kOhm reading measuring right on the resistor, but there was infinite resistance between the resistor and the transistor marked S3 even though they should be connected through a trace. (the other side of the resistor connected through a trace to the nearby 01C resistor had continuity)
So I tried to reflow one side of the resistor with a soldering iron but that didn't help and after probing it some more the resistor just detached from the PCB completely, and after that I saw that the problematic pad on the PCB was in rough shape, but I managed to somewhat revive it with solder and wick:


Anyway, now I fried that resistor so I need a new one.
Do you guys think that resistor was the culprit of not charging? I wonder whether the charging will work again if I replace that resistor.

Additional photos:

PCB markings:
AS-8206BPL2 21.6V
LA-PABS-8206-A2
This vacuum is designed such that the handheld unit detaches from the main body used for floor cleaning.
I noticed there is a vacuum from another brand that looks physically similar, Rowenta Dual Force RH6751WO

[CapLeaker]

I doubt the resistor is the cause. Resistor marking 683 = 68k ohm.
Did you verify the external power supply actually delivering 26VDC to the BMS board? Just to rule out any chance as a PSU issue. Look at that resistor I circled, it looks like it has a hole in it.

[flat-earther]

I doubt the resistor is the cause. Resistor marking 683 = 68k ohm.
Did you verify the external power supply actually delivering 26VDC to the BMS board? Just to rule out any chance as a PSU issue. Look at that resistor I circled, it looks like it has a hole in it.

Thanks.
My bad, that white is residue from when I cleaned the board with alcohol and didn't dry it well. I wiped it clean now, here's some new photos.

That resistor you circled measures 0.0 Ohms so it's okay.

I have labeled the wires connected to the board so it's clear what each is for:


Measuring the charger plug output without connected to the vacuum, it is 27.50V, a bit higher than labeled, I assume that's fine and since it's not under load.

[CapLeaker]

Check QC1 and D13. QC1 = AO4459 P-channel (so it is negative) mosfet with a body diode. Look at the gate drive. If there is no gate drive, the mosfet wont do anything. It is configured as a on/ off switch.
Plug the charger in and measure the voltage coming in and going out of this mosfet.

[flat-earther]

With the charger plugged in, black probe on ground, measuring voltage at input of the charging mosfet, I see it's constantly jumping between 27.5 and up to even 27.88V:

At the mosfet's output the voltage is constantly jumping between 7.6 and 8V:


At the mosfet's gate it's 27.2-27.6V:
Is such high voltage at the gate normal?

[CapLeaker]

Carefully get rid of that white beigeish goo on the inductor and suroundings. All of it. You may use heat and 99% IPA to pick / peel it off...

[flat-earther]

BTW I have manually charged the cells with an external charger so they wouldn't be empty. Right now the battery is at 24V.

How do you want me to check D13?

[flat-earther]

Just to experiment, I have set an external power supply to 26V because that way I can see current draw on the power supply.
When I plug that power supply to the vacuum I see a very brief about 1A current draw and then it instantly drops to zero.

[flat-earther]

Well removing that glue was extremely difficult. I used a 25W iron with sharp tip to pick it, then tweezers, and I destroyed the inductor by accidentally pulling the copper wire along with the glue with tweezers and the wire snapped.
I will need to try and resolder that wire back but it's so short I'm not confident in it, if that doesn't work I may unwind it one rotation to make it longer. I hope I didn't short the winding by scratching it with the tweezers.

From one side of the inductor to the end of the broken wire it measures 0.3 Ohms, is that still okay?

[CapLeaker]

The inductor should measure a short. Disconnect everything from the board and check D13 in diode mode. Should have something like 0.3V drop one way and an open the other way.. hmm dunno… maybe the HT66F018 went south. That would suck, as it is a programmable chip.

[flat-earther]

Disconnected all power from board, D13 in diode mode is open line one way and 0.217V the other way.

BTW while the inductor is essentially disconnected is there anything I should test in this state before reconnecting it?

[CapLeaker]

Check that gate drive circuit again with that z5 etc in diode mode, no power.
that zener diode is a 15V zener. So it should be clamping down any excess voltage to 15V.

[flat-earther]

All power disconnected, measuring Z5 in diode mode, OL one way, 0.725V the other way.

[CapLeaker]

That's o.k.

[CapLeaker]

I'd say replace that resistor that's broken make sure the traces around there are going where supposed to be going, check across all the MLCC that there is no short, then try charging again.

[flat-earther]

I tested all ceramic caps, didn't find any shorted. Ceramic cap C9 near motor wires is about 1 Ohm but the motor wires are soldered there so I think that's just the motor resistance.

All power disconnected, I measured in diode mode diode D1, 0.663V one way, 0.571V the other way. Is that okay?

[CapLeaker]

Take it out or lift one leg and check that diode again. Diodes should read to 99% one way but not the other, unless it is a bidirectional TVS diode.

[flat-earther]

I removed the diode D1, in diode mode it measures 0.575V one way, OL other way.
In resistance mode 68kOhm one way, OL other.

[flat-earther]

D1 diode soldered back.

Inductor repaired. I bent the copper wire down and soldered it to the pad.

[CapLeaker]

Now replace that 68k resistor and make sure the traces are good around that area.