HOW TO RESET YOUR LAPTOP LITHIUM BATTERY PACK
X-Acto knife, plastic separator tools, screwdriver, etc. depending on battery pack construction.
Voltmeter - can do without it but it will give you a level of confidence
internet connection - to get info
magnifying glass - for us old folks
This will not be successful on all battery packs. Some will actually have bad batteries. The symptom is that the battery runs down VERY rapidly and also charges to full charge VERY rapidly. Well, full charge as indicated by the computer. The problem is that it is not really fully charged, but the "smart" controller in the pack thinks it is, shuts down the charging, and tells the computer that it is fully charged. If it runs down fast but takes a long time to charge or never attains full charge then resetting may not accomplish anything except give you practice for the next time.
I have done this successfully on at least three different battery packs. The first one was from a Micron laptop. It was VERY easy to do as lithium battery packs were just arriving and the manufacturer was smart enough to include the reset function on an I/O pin so disassembly was not required. It was a matter of shorting the pin to ground for a few seconds.
The second one was a Dell 8200 with 2 batteries. One of the batteries began exhibiting the symptoms. Unfortunately, at this point manufacturers did not bring the reset pin to the connector as it was on the Micron battery. This is a BIG mistake. MANY battery packs can be brought back to life simply by resetting them saving people lots of money and recycling costs.
Here is the deal. Battery packs have a built in “smart” controlling system. This is required because of the disastrous consequences of over-charging or over-discharging the lithium batteries. The smarts is a special purpose microprocessor. It has memory and monitors each battery or set of parallel units. The memory’s purpose is to build a database so that it can determine when it should tell the computer to shut down to prevent the pack from over-discharging. It also insures that the batteries fully charge to equal voltages.
Sometimes it goes dumb and the data tells it to stop charging the batteries too soon. I may also tell the computer to shut down too early. Again the symptom of this is a short time to recharge and a short discharge life. To my knowledge, ALL of these microcontrollers have a reset pin, at least the two I have looked up so far have. The reset will clear out this bad data and allow it to start fresh.
So, back to the Dell. I decided to disassemble the pack to figure out if I could reset it. I carefully separated the clamshell plastic cover; pulled out the PCB (not disconnected) that has the charge monitor circuit on it, located the microcontroller, looked up the specs to find the reset pin, and shorted it to ground. I then replaced the PCB and put the battery back into the computer to charge.
This time it took a long time to charge – as it used to. When I ran it on battery, it lasted almost as long as it did when it was new. SUCCESS!
OK, I still have the Dell, but the batteries are way overdue to replace. But now I have a 5 year old Averatec and I have replaced the battery once already. It is now seldom used – only when I need to run XP. The battery started exhibiting the same symptoms as above. It also would go dead in a short time just sitting in the off condition (not sleep). Time to try again.
I carefully separated the glued clam shell plastic and opened the case. The trick to separating them is to figure out how they went together. In this case, I carefully inspected the obvious separation gap to determine which half went INTO the other. Other battery packs may be screwed together or just snapped together. This is the hardest part, figuring it out and getting it open. Sometimes, multiple methods are used and screws are often covered with labels. You can find hidden screws by running your nail or other dull object, like the back end of a ball point pen, over the plastic. If you hit a soft spot it MAY be a screw hole. It could also be a mold mark in the plastic. You will have to dig to find out.
Sometimes none of this is “obvious”. Once I determined this I slid an Xacto knife into the split.
It went in pretty easily. If I had chosen the wrong side it would have been more difficult as I would be cutting plastic and not glue. I then ran the knife all around the case till all sections were loose. Be sure to insert knife at a steep angle so as not to damage anything inside. (Do be careful, those knives can do bodily harm.) Off it came. There were no screws.
This unit has 8x 18650 lithium cells arranged 2x4. The capacity is 2x whatever one cell is. That is difficult to determine from print on the cell since there usually is none regarding AHr and if there is it is usually an overstated lie.
I then measured the voltage on each parallel bank of cells. They were all pretty much the same. Shutdown voltage was 3.7v or about 14.8V total. Fully charged it was about 15.8V. This pack was pretty easy, some are not because of how they are arranged or the lack of convenient points to place a voltmeter. It may be difficult to determine how they are connected. I had one pack that I had to remove the batteries to figure it out.
As you can see, this pack is 2x2 in line with a wire going to the PCB from each junction. Note that there are three protection devices – a thermal fuse (silver cylinder), a thermistor (covered in white adhesive), and a multifuse.
The monitor point wires are Black (ground), Green (1st cells), Yellow (2nd cells), Red (3rd cells), and main + (4th Cells). The Black wire is the main negative line. The Red wire coming from the end of the thermal fuse is the main positive lead and the 4th cells monitor line.
If the cell voltages are pretty much even, then there is a good chance the reset is going to work. If any cell(s) are very different from the rest (one cell being 2.4v and the others being 3.7v for example), odds are against a reset fixing things. Makita drill packs are famous for failing this way. However, I have revived one of those, but that is for another Instructable.
Now you must determine which IC is the microcontrollers. It is either the largest one or the one with the most leads.
Here you see a large IC with 20 pins and a smaller one with 32 pins, 8 on each side. My first choice is the smaller one. It is an Amtel Mega 168 (as read from the IC). I googled this and found a datasheet. It is indeed a microcontroller.
The data sheet info that I need is:
1. Which pin is reset? – in this case it is pin 29 or the 4th from the corner CCW with the index mark on it. I followed the circuit trace on the PCB to a resistor/capacitor junction (typical for a reset pin circuit) and chose that spot to use for resetting (the IC pins are just too close for me). As long as there is no direct Vcc pins nearby you can probably get away with an accidental contact to adjacent pins. Here the Vcc pins are on adjacent sides so I am well clear if I chose that.
2. Which level is reset? – in most cases it is ground. It is usually indicated by labeling "reset" with a line over the top – that depends on company policy. If there is a line, it is ground to reset. If there is not a line further reading may be advisable. Since I can't easily create an overscore I will use a "#" --> #reset.
I’m ready. I connected one end of a test lead (wire) to the black wire on the battery and touched the R/C junction slowly a few times for good measure. I plugged the battery back into the computer and plugged in the charger. This time it took over an hour to charge. A GOOD SIGN!
BETTER THAN NEW. But then it was a “NEW” fleeBay battery. Each set of cells is now 4.16V for a total of 16.64V.
When booted up after fully charging, the time to shutdown was 2.34 hours, up from 22 minutes before the reset. When the battery was "new" it was less than 2 hours. I think that was a successful endeavor! What do you think?
You may not want to glue the halves together too tightly. If you are lucky it will snap back together and hold without glue. If you are real daring, and you have an unattached pin on your I/O connector, you might consider connecting it to the reset point. (You will definitely need a volt/ohm meter to figure this out. Be sure to measure from both ground and V+ to the pin. Don't rely on following traces on the PCB.)
Then instead of a disassemble, you just need to short that pin to the ground pin for future resets. Just don’t forget which one is #reset and which one is ground. Unfortunately, mine are all spoken for, but the lid has snapped back in place without glue.