Announcement

Collapse
No announcement yet.

Laptop battery charging circuit

Collapse
This is a sticky topic.
X
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

    Laptop battery charging circuit

    1. Some theory
    1.1. Introduction
    Almost every laptop has a charging circuit of some sort, that does a few things:
    • Manages switching between AC adapter and battery power
    • Manages battery charging and generates power rail to charge battery


    There are 2 different designs to handle switching between AC adapter and battery power and feeding the system:
    • Hybrid power boost (HPB) and traditional circuits without power boost (see Warning in HPB section)
    • Narrow VDC (NVDC)

    Texas instrument has a nice overview of these technologies with some more details at https://e2e.ti.com/blogs_/archives/b...ing-topologies . The essential information is summarized below.

    There are also 2 different designs to handle battery charging rail generation:
    • Buck converter
    • Buck-Boost converter


    Almost all laptops without USB-C charging capabilities use a buck converter. Most of these use HPB, except Apple laptops which always use NVDC. A bunch of non-Apple ultrabooks use NVDC as well.
    Almost all laptops (incl. Apple) with USB-C charging capabilities use NVDC with a buck-boost converter.
    Some netbooks or entry-level laptops (especially ARM platforms and some Atom platforms) can have a circuit different to what is presented here, using a PMIC that handles almost every power-related features. Since these aren't common and highly depend on the board and the ICs used, they are not covered here.

    For MacBooks without USB-C you have some more details on the charger IC here: https://logi.wiki/index.php/ISL6258_...roubleshooting



    1.2. Hybrid power boost design
    From https://en-support.renesas.com/knowledgeBase/13886169

    Originally posted by Renesas
    In a Hybrid Power Boost configuration, the synchronous buck converter runs as a normal buck converter when the adapter provides power to the system and is charging the battery. When the adapter power is not sufficient, the synchronous buck converter runs in reverse to boost the battery voltage to around 20V. Thus the battery supplements the adapter whenever the adapter power is not sufficient. This requires no circuit changes from a traditional adapter.

    The change is required in the control circuit (the battery charger controller). The advantage of this system over a traditional charger is that the battery is able to assist the adapter during turbo workloads. This system has the disadvantage that the light load efficiency is pretty low as it is difficult to achieve high light load efficiency at high input voltage.

    In this design, the system is fed from the AC adapter, and a MOSFET (we will call it battery-to-system MOSFET) allows the current to flow from the battery to the system when on battery only, while prevent current from flowing directly from the AC adapter to the battery when AC adapter is plugged in.

    Therefore, when on AC adapter, the main power rail voltage will be the voltage provided by the AC adapter. In general this voltage is between 19V and 20V. A few machines use 12V or 16V.
    When on battery, the main power rail voltage will be the voltage provided by the battery. This depends on the number of cells in series in the battery and the state of charge, but it is lower than 19V. (see Battery voltage section)

    Warning: Hybrid power boost provides an additional feature where the battery can provide additional power to the system even when running on AC adapter. Older designs used the same circuit but the charger IC was not able to provide this feature. The presence or abscence of this feature is irrelevant to this article, so the older designs will fall under the hybrid power boost term as well.


    1.3. Narrow VDC design
    https://en-support.renesas.com/knowledgeBase/6680047

    Originally posted by Renesas
    This figure shows the Narrow VDC (NVDC) topology. Here, the system bus (Vsys) is not connected directly to the adapter. It is connected to the output of the buck converter. Hence, NVDC operates only as a buck converter, both when NVDC charges the battery and when the battery supplements the adapter and provides power to the system. NVDC implementation reduces the switch-over period between the charging mode and the hybrid power mode. NVDC implementation allows the system to minimize the period of overloading the input power source when CPU is in Turbo Boost mode.

    The advantage of using the NVDC system is that the overall system efficiency is better compared to the Hybrid Power Boost (HPB) charger. The system can be designed for a smaller voltage rating since the system has a lower Vin. The disadvantage is that the charger components’ size and power dissipation increases.

    In this design, the system is fed directly from the battery charging rail. The voltage on the main power rail is always close to the voltage on the battery. (see Battery voltage section)

    1.4. Buck converter
    This works the same as any single-phase synchronous buck converter. There's one high-side MOSFET and one low-side MOSFET turning on and off alternatively to step down the voltage from the input, for example from 19V to 12.6V.


    (from bq24715 datasheet)

    1.5. Buck-Boost converter

    A synchronous 4-switch buck-boost converter uses a first set of high-side and low-side MOSFETs before the inductor. They can act as a regular buck converter to step down the voltage coming from the AC adapter. There's a second set of MOSFETs after the inductor, they are used as a boost converter to step up the voltage coming from the AC adapter.
    That way, the machine can take 5V to 20V from an USB-C charger and still be able to charge the battery. Of course, with only 5V as the input, the power will be lower than with 20V, so the battery will charge slower and the laptop may not turn on without battery or when the battery is discharged.


    (From ISL9238 datasheet)

    1.6. Battery voltage
    Laptops battery packs are based around lithium cells.
    These are either Li-ion cells, round cylinder similar to your alkaline battery but in a 18650 package, or LiPo cells, flat rectangular cell like in a smartphone.

    Li-ion cells typically have a nominal voltage of 3.6V to 3.7V, LiPo cells typically have a nominal voltage of 3.7V to 3.85V. It varies depending on the brand and series of the cell.


    The battery voltage and by extension the voltage used to charge the battery depends directly on how many cells there are in series inside the battery pack.
    Battery packs can be described with a "xSyP" number, where y is the number of cells in a group connected in parallel, and x is the number of groups connected in series.
    All cells in parallel in a group will have the same voltage across them. Putting groups in series will sum their voltages.

    For example a 3S2P pack with 11.1V nominal voltage contains 6 cells, 3 groups in series of 2 cells in parallel, for a total voltage across the pack of 3x3.7V = 11.1V.
    The number of cells put in parallel in a group helps increasing the capacity, but it doesn't change the voltage. We are interested in the voltage here, so we will ignore the cells in parallel.

    The "nominal voltage" represents the voltage around which the cell is throughout most of its discharge, it should be what you are able to measure on the pack when it's charged to around 50%.
    When charged fully, the voltage will be higher, when completely discharged, the voltage will be lower.

    Charging voltage will be a bit higher than the fully charged voltage of the pack. For example, an 11.55V pack (3S of 3.85V nominal) can in general be charged at 13.1V (4.35V per cell). An 11.1V pack (3S of 3.7V nominal) can in general be charged at 12.6V (4.2V per cell). A 10.8V pack (3S of 3.6V nominal) can in general be charged at 12.3V (4.1V per cell).
    This of course highly depends on the exact type of cells used, and using a voltage that's too high for a given cell type can at best damage the cell, at worst be a safety threat.
    Attached Files
    Last edited by piernov; 08-04-2021, 02:30 PM.
    OpenBoardView — https://github.com/OpenBoardView/OpenBoardView

    #2
    Re: Laptop battery charging circuit

    2. Troubleshooting charging circuit

    2.1. No power
    First thing to check if the machine is not powering on is if the main power rail is present. If it is present, then in general the problem is not in the charging circuit. You can still check "Shorted DC-in MOSFET" and "Missing voltage on the charger IC" as they could still be a source of trouble in some cases.

    To be able to check the main power rail, you have to identify if you have an HPB or an NVDC configuration.
    • With an HPB configuration you have to check the voltage after the 2nd DC-in MOSFET and after the current sensing resistor for AC adapter power input. You are supposed to find the AC adapter voltage.

      (from Quanta X63 schematics)
    • With an NVDC configuration, you are supposed to find the battery voltage (see Battery voltage section):
      • If it's a buck converter you have to check the voltage after the buck converter inductor and after the fuse if it exists.

        (from Apple 820-00165 schematics)
      • If it's a buck-boost converter you have to check the voltage at the output of the 2nd set of MOSFETs and after the fuse if it exists.

        (from Apple 820-00239 schematics)


    If you confirmed your main power rail is missing, there are several possibilities

    2.1.1. Burnt fuse
    You found voltage before the fuse (after the buck converter output inductor or buck-boost 2nd set of MOSFETs), but not after it: fuse is blown. Replace it and go to the next section "Short to ground on main power rail". Don't apply power before resolving the short to ground.

    2.1.2. Short to ground on main power rail
    Most common issue in this case is a short to ground on main power rail caused by a capacitor, and IC, a MOSFET…
    In this case you have to measure *exact* resistance to ground on the main power rail (same point as previously). If it is lower than several kohms there's a short to ground.
    As indicated in the "Begginer's guidelines: what not to do" section, DO NOT inject 19V. Check high-side MOSFETs for short, easiest way is to place one probe on the main power rail and the other probe on each large inductor of the board. If one is lower than 1 ohm, there's a possibility of high-side MOSFET short. If not, you can start injecting voltage *at 1V*, and double check none of the large BGA is warming up. If 1V is not enough, you can start increasing voltage, after making sure none of the large BGA is warming up again.
    Once you find the bad component, remove it, confirm short to ground is gone, and replaced it. Don't apply power to the board if the component is removed without being sure that it'll not cause any harm.

    2.1.3. Shorted DC-in MOSFET
    This is the second most common issue. What is troubling for beginners here is that the MOSFET that is blocking the current is *not* the MOSFET that has failed.
    What happens here is that the MOSFET that shorts is the one with the body diode in forward (in the illustration below, the 2nd one, note that they can be exchanged), not the one with the body diode in reverse.
    Effectively, the shorted MOSFET will have drain/source/gate connected together. Both DC-in MOSFET gates are also connected together. And Both DC-in MOSFET sources are connected together. So with gate/source of DC-in MOSFET in forward bias shorted, it also causes gate/source of the DC-in MOSFET in reverse bias to be shorted. This MOSFET is therefore turned off permanently, not letting the current through.

    The other MOSFET could also be shorted, in the case of an HPB topology the main power rail will be present since current will flow through the shorted MOSFET and through the body diode of the other MOSFET, but some signal from the charger IC might be missing causing the platform not to turn on. In an NVDC topology the main power might be missing.

    Simply check resistance between drain/source/gate of both DC-in MOSFETs to confirm they are not shorted. It should show at least several kohms.


    (from Quanta X63 schematics)

    2.1.4. Burnt current sensing lines
    The charger IC also monitors the current flowing from the AC adapter, and from/to the battery to prevent overcurrent conditions. It can happen that the current sensing circuits get burnt, either the current sensing resistors themselves or the resistors that connect them to the current sensing amplifier inputs of the charger IC.
    Easiest way is to measure resistance directly across the pins of the current sensing amplifier inputs of the charger IC. The measured resistance should be the sum of the current sensing resistors, and the resistors that connect it to the charger IC. This also accounts for a possible break in the line that could be caused by liquid damage for example.


    (from Quanta X63 schematics)

    2.1.5. Shorted gate on charging rail buck converter MOSFET
    A gate to source on the buck converter MOSFET can cause the internal LDO output (REGN) to be pulled low, in turn causing the charger IC not to work.
    In this case it is recommended to replace both the MOSFET and the charger IC at the same time, as either of them can cause the other to fail.

    2.1.6. Missing voltage on the charger IC
    Charger IC requires several voltages to work.
    Exact names depend on the charger IC used, but there is at least:
    • DCIN/VCC: input voltage from the AC adapter, should be close to AC adapter voltage
    • ACIN/ACDET: voltage divider from the AC adapter to detect a correct voltage range. Exact required voltage depends on the charger IC but it often needs to be higher than 2.4V or 2.6V
    • REGN/VDD: internal LDO output in general between 5V and 6V, input taken from DCIN/VCC.
    • CMSRC (present only on some charger IC paired with N-channel MOSFETs): voltage reference/source taken in between the 2 DC-in MOSFETs to generate N-channel DC-in MOSFET gate drive voltage (ACDRV). Should be close to AC adapter voltage.


    These signals should come up once the previous signals are good and the previous problems have been excluded:
    • ACOK: asserts that the input voltage is good. For P-channel DC-in MOSFETs it often drives their gates directly and is active low. For N-channel DC-in MOSFETs it should be active high and may be pulled-up externally to a 3.3V rail.
    • ACDRV: only present with N-channel MOSFETs, drives their gate, should be 6V above input voltage so in general around 25V



    2.1.7. Damaged passive component
    This will be harder to spot, but depending on the circuit a missing/bad resistor or bad capacitor could be the culprit. Visual inspection is the key here. This is very rare though, except in case of liquid damage. If there is liquid damage, follow the visual hints and look for corroded traces, pads, pins and probe points as well.
    One possible issue would be the gate resistors for the DC-in MOSFET, as well as the resistor connected to CMSRC pin when it is present, so you can always check them if you're in doubt.

    2.1.8. Dead charger IC
    If you excluded all the previous possibilities, the most probable is that the charger IC itself is dead. Replace it and see if the behaviour changes.


    2.2. Battery not charging
    First thing is to try another battery. Then try another battery. Finally try another battery. 99% of battery problems are the battery itself, even "new", market is filled with garbage batteries. A "new" battery should come charged at around 50%, at worst between 20% and 80%, and the machine should turn on from it. Some batteries need to be "kickstarted" by plugging in the AC adapter for a short while to turn them on.

    Some laptops with non-removable battery (newer Acer) have a battery enable/disable switch. Back cover must be properly assembled otherwise the battery won't work.

    Some of the above points at 2.1 No Power can also apply to battery charging so you should check them as well (in particular 2.1.3. Shorted DC-in MOSFET).
    Once again it is important to know if you have an NVDC or an HPB topology.
    Battery has to be detected to start charging.

    2.2.1. AC adapter not detected, wrong AC adapter
    Laptops with AC adapter identification circuits such as HP and Dell won't charge the battery if the AC adapter is not properly recognized or of the incorrect power rating.
    For HP, the laptop must be able to turn on from AC adapter only. For Dell the AC adapter must be recognized properly in the BIOS setup with the correct power rating.

    2.2.2. Battery not detected
    Issue on charging circuit SMBUS.
    Check voltage and diode mode to ground on charging circuit SMBUS SDA and SCL lines


    (from Quanta X63 schematics)

    2.2.3. Low voltage on battery charging rail
    If it's an NVDC topology, the machine wouldn't turn on with a too low voltage on the battery charging rail. However with NVDC there is a charging MOSFET between the buck converter output and the battery power rail that could be the culprit or its gate resistor.

    If it's HPB topology, the battery is in general connected directly to the buck converter output. A bad or deep discharged battery will pull down the output. Let the battery charge overnight and see if it recovers, if not, replace the battery. Without battery, in HPB tology the output of the battery charging rail will often be low.
    DC-in MOSFET or current sensing could be the problem, see "No power section".

    Some charger IC have a "CELL" pin to control the default voltage to apply to a battery relative to its number of cells in series. Wrong setting can cause wrong voltage.

    2.2.4. High voltage on battery charging rail
    It could be a shorted high-side MOSFET on the buck converter for the battery charging rail.
    In the case of HPB toplogy, it can also be a shorted battery-to-system MOSFET.


    (from Quanta X63 schematics)

    Some charger IC have a "CELL" pin to control the default voltage to apply to a battery relative to its number of cells in series. Wrong setting can cause wrong voltage.
    Attached Files
    Last edited by piernov; 06-18-2023, 06:25 AM.
    OpenBoardView — https://github.com/OpenBoardView/OpenBoardView

    Comment


      #3
      Re: Laptop battery charging circuit

      Reserved
      OpenBoardView — https://github.com/OpenBoardView/OpenBoardView

      Comment


        #4
        Re: Laptop battery charging circuit

        Thanks for this wealth of information all in one place Piernov, it's very welcomed
        My YouTube channel - https://www.youtube.com/channel/UCiD...5d6OgizDqoi_SA

        Comment


          #5
          Re: Laptop battery charging circuit

          thanks for the useful info

          Comment


            #6
            Re: Laptop battery charging circuit

            Thank you..Very useful tutorial especially for beginners like me...

            Comment


              #7
              Re: Laptop battery charging circuit

              Thank you very much Piernov, for taking the time to post such useful material

              Comment


                #8
                Re: Laptop battery charging circuit

                Good job @piernov. Well researched post.

                Comment


                  #9
                  Re: Laptop battery charging circuit

                  Another ABC dignostic ..nice thread and very helpful..Chapeau bas pierre

                  Comment


                    #10
                    Re: Laptop battery charging circuit

                    Originally posted by piernov View Post
                    Reserved
                    Merci infiniment pour ta contribution oh combien précieuse, félicitation pour ta maîtrise du sujet, j'aurais une question a te poser si tu permets , quelle est la manière dont est détectée la charge maximale de la batterie qui fait basculer la led orange vers la led blanche ( pour HP par exemple) , en précisant le role des différents circuits qui interviennent dans le processus . MERCI en attente de ta réponse qui sera pour moi une lanterne dans une caverne obscure.

                    Thank you very much for your contribution oh how precious, congratulations on your mastery of the subject, I would have a question to ask you if you allow, what is the way in which the maximum charge of the battery is detected which switches the orange led to the led white (for HP for example), specifying the role of the different circuits involved in the process. THANK YOU waiting for your answer which will be for me a lantern in a dark cave.
                    Last edited by SMDFlea; 11-23-2020, 11:10 AM. Reason: google translate ,merci beaucoup

                    Comment


                      #11
                      Re: Laptop battery charging circuit

                      Lol some jealous person deleted my post.

                      Talk about freedom of speech 🤣


                      Anyhow just helping

                      You have forgotten cmsrc line resistor resistance check , beginners fail to understand and keep changing charging ic and KBC but unable to repair. Sometimes cmsrc resistor value increase to out of range due to short in somewhere and this cmsrc signal will not reach charging ic and it will not let current pass until cmsrc signal received and your laptop charging led will blink for a second and turn off . In this situation if you change KBC or Io it will not work until you pull out resistor and replace it.

                      Do not delete this as well ,I am helping people.

                      Comment


                        #12
                        Re: Laptop battery charging circuit

                        It is true that missing/damaged resistor going to CMSRC pin when it exists on the charger IC will prevent it from working properly (N-channel DC-in MOSFETs gate drive voltage won't be generated).
                        It's far from being a common problem though and not all charger IC have a CMSRC pin. It's also basically part of "2.1.6. Damaged passive component", others damaged/missing resistor or capacitors can cause an issue (like bootstrap cap/resistor for example) but without liquid damage they're very rarely the problem.
                        I still added a note regarding that pin in 2.1.5. and 2.1.6 as it might be useful to some people, thanks.
                        OpenBoardView — https://github.com/OpenBoardView/OpenBoardView

                        Comment


                          #13
                          Re: Laptop battery charging circuit

                          very important topic tanks

                          Comment


                            #14
                            Re: Laptop battery charging circuit

                            Thank you for the information, is very usefull.

                            Now, in case of battery not detected and not short in SMBUS SDA and SCL lines, what other thing can cause the fail?

                            Comment


                              #15
                              Re: Laptop battery charging circuit

                              Hello, good morning thanks for accepting me I have a question
                              It had a board in which there was no scheme but its input mosfets were channel N, the charger source was the Bq25a, therefore the charger activated the gates to the mosfets, however, in the gates of the mosfets there was no voltage, now the charge ic was powered it had the acdet = 2.41, the acin was 0v and the regn = 0v but the capacitor that goes in that Regn line was beeping on both sides, I took the capacitor to measure outside and it was fine to feed the plate without that capacitor and in the regn terminal had no voltage, my query is, could the ic charger be bad?
                              I also took out the second mosfet to measure it outside and it's ok, thanks
                              Last edited by piernov; 12-01-2020, 03:08 PM. Reason: google translate

                              Comment


                                #16
                                Re: Laptop battery charging circuit

                                thamks

                                Comment


                                  #17
                                  Re: Laptop battery charging circuit

                                  Originally posted by jessy View Post
                                  Hello, good morning thanks for accepting me I have a question
                                  It had a board in which there was no scheme but its input mosfets were channel N, the charger source was the Bq25a, therefore the charger activated the gates to the mosfets, however, in the gates of the mosfets there was no voltage, now the charge ic was powered it had the acdet = 2.41, the acin was 0v and the regn = 0v but the capacitor that goes in that Regn line was beeping on both sides, I took the capacitor to measure outside and it was fine to feed the plate without that capacitor and in the regn terminal had no voltage, my query is, could the ic charger be bad?
                                  I also took out the second mosfet to measure it outside and it's ok, thanks
                                  If charging ic is not generating REGN and charging ic vcc is ok and acdet is normal then yes you are right otherwise check logic and signal state Active on low or high

                                  If you read sheet there you will find comparing circuit read it for more information

                                  Comment


                                    #18
                                    Re: Laptop battery charging circuit

                                    new ideas noted keep sharing

                                    Comment


                                      #19
                                      Re: Laptop battery charging circuit

                                      I have a question for a problem with the two n-channel mosfets - which I already replaced to be on the safe side. The BQ24725 emits every 1.3 seconds a very short pulse with an amplitude of approx 11v on the gates of the two N channel mosfets. When does the charge controller go in such a mode ? On battery alone, the mainboard correctly powers up. On the DC jack alone, I get nothing unless I bridge the 1st mosfet. Then the power button starts the laptop. Having both battery and a lab power supply connected to the DC jack won't start the laptop unless I bridge again the 1st mosfet. But then the current limiter of the lab psu kicks in for good reasons. Any ideas would be appreciated.

                                      Comment


                                        #20
                                        Re: Laptop battery charging circuit

                                        Thank You for this informative thread

                                        Comment

                                        Working...
                                        X