Battery Selection, Safety, and Monitoring in Mobile Applications

1

Battery Selection, Safety, and Monitoring in Mobile Applications

Yevgen Barsukov

Texas Instruments

22006 TI Portable Power Seminar

Agenda

IntroductionBattery Charging / Discharging characteristicsSelf- discharge / AgingCharging algorithm/Charge TerminationSpecial requirementsMain applicationsAdvantages and LimitationsBattery Treatment

2


Principles of Chemical Power Source

Vdis

Vch

Cathode Anode

Vdis

Vch

Cathode is aggressive oxidizing agent (like oxygen, MnO2, NiO(OH)). It really likes to rip electrons from anything near it

Anode is a willing donor of electrons, a reducing agent like hydrogen, Zn, Cd, Pb. It has excess electrons to give.

When both are electrically connected, spontaneous electron flow from anode to cathode occurs. Ions move in electrolyte to keep electric neutrality

e-

C+

Electrolyte

-+

R

Ion+


Battery Electrical Equivalent Circuit

I.RBAT

CapacitorCapacitor + resistorBattery

Constant current discharge,transient effect

ONOFF

Cathode AnodeElectrolyte

Separator

+ -

Positiveelectrode

Negativeelectrode

Battery AC modelCathode

+ CBAT RBATAnode

-

Battery DC model

3


Battery Chemical Capacity

EDV=3.1V/cell

Qmax

• Battery capacity (Qmax) is defined as amount of charge which canbe extracted from the fully charged cell until its voltage drops below end of discharge voltage (EDV).

• EDV is minimal voltage acceptable for the application or for battery chemistry, whichever is higher.

Load current: 0.2C

CBAT RBAT

Battery Capacity: 1CDischarge rate 1C:Current to completely discharge a battery in one hourExample:2200mAh battery, 1C discharge rate: 2200mA

3.6V (Battery rated voltage)

1 2 3 4 5 60

3.0

3.5

4.0

4.5

Capacity, Ah

Voltage, V


Useable Capacity

If high current is flowing through the battery, EDV will be reached earlier because of I*RBAT voltage drop.External battery voltage can be roughly modeled as V=V0CV-I*RBAT where RBAT is internal resistance of batteryUseable capacity Quse of battery is capacity at given load I. Quse is less then Qmax

EDV

QmaxQuse

I*RBAT

0 1 2 3 4 6

3.0

3.5

4.0

4.5

Capacity, Ah

Voltage, V CBAT RBAT

VOCV

+ - I

+ -V=V0CV - I*RBAT

4


Battery Impedance and Cycle Life

During repetitive charging/discharging usable capacity is decreasing

It is common to define cycle life of battery as number of cycles when useable capacity decreases below 70% of original.

Depending on chemistry, different factors are influencing degradation. Common factors for all batteries are high charge rate and high temperature.

Battery AC model


Safety and Thermal StabilityIncreased temperature can not only accelerate degradation, but it can cause thermal run-away and battery explosion

This is a specific concern with Li-ion battery, but other batteries are also capable to explosive self-heating

Rapid temperature increase can happen if battery is overcharged at high current or shorted

During overcharge of Li-ion battery very active metallic lithium is deposited on anode. This material highly increases the danger of explosion

Material developments are ongoing to increase thermal runaway temperature.

Thermal runaway

5


NiCd Rechargeable Battery


NiCd Reactions

Discharge: Minimum 0.7VThe positive electrode is converted from nickel oxyhydroxide to nickel hydroxideThe negative electrode is converted from Cadmium to Cadmium hydroxide

Charge: Maximum 1.6V, average operating voltage: 1.2VThe reverse reaction occurs during chargeToward the end of charge the reaction efficiency decreasesOvercharge produces oxygen which then produces heat as it exceeds the electrode recombination rateCharge termination is -dV~10mV, dT/dt~1°C/min. At low rate no –dV/dt will occur, therefore termination by timer, temperature of capacity count is needed.

6


NiCd Charging


NiCd Discharging Characteristics

*http://www.buchmann.ca

CBAT RBAT

VOCV

+ - I

+ -V=V0CV - I*RBAT

Good characteristics for high discharge rate due to low internal impedance

7


NiCd Self-Discharge

• Aged battery has higher self-discharge rate


Advantages and Limitations: NiCd Battery

Advantages:

• Fast and simple charge: constant current charge

• High number of charge/discharge cycles >1000

• Good low temperature discharge capacity: 80%@-200C

• Low internal impedance

• Simple transportation

• CheapDisadvantages• Low energy density: 70Wh/kg (90Wh/kg: NiMH; 200Wh/kg: LI-Ion)

• Memory effect: Need periodically deep discharge to recover capacity

• Relatively high self-discharge at high temperature

• Environmentally unfriendly: Contains toxic metals

8


Treatment of NiCd battery

Healthy habits:Most stable in discharged state. However , due to fast self-discharge, to prevent over discharge should be placed for storage in charged state.If used in application needing prolonged trickle-charge (stand-by), should be periodically fully discharged

Degradation mechanisms: Voltage depression. Cathode material changes into inactive crystalline form when stored in charged state. Prevention –discharge regularly. Repair: Charge/discharge several times to 0.7V. For cases of severe capacity loss, keep at 0.7V for prolonged time (24 hrs)


Nickel-Metal-Hydride (NiMH) Battery

9


NiMH Reactions

Discharge: Minimal 0.7VThe positive electrode is converted from nickel oxyhydroxide to nickel hydroxideThe negative electrode is converted from a Metal Hydride to Metal plus Water

Charge: Maximal 1.6VThe reverse reaction occurs during chargeWhen the Metal is converted to a Metal Hydride, heat is produced. Towards the end of charge the reaction efficiency decreasesOvercharge produces oxygen which then produces heat as it exceeds the electrode recombination rateCharge termination: PVD<-5mV, dT/dt~1°C/min. Might not occur at low charge rate

Average operating voltage: 1.2V


NiMH Battery Charge Characteristics

At the End of Charge • Cell voltage drop, but not significant• Cell temperature Increase Charge termination

1 C

1 C

0.5 C

0.5 C 0.1 C

0.1 C

10


NiMH Discharging Characteristics

• End of Discharge Voltage=EDV=1.0V/cell• Lower useable capacity at high discharge rate due to IR drop

VBAT=VOCV – I RBAT

EDV=1.0V

CBAT RBAT

VOCV

+ - I

+ -V=V0CV - I*RBAT

1 C

0.2 C

3 C


NiMH Discharge under Different Temperatures

•No discharge capacity at -200C

11


NiMH Self-Discharge Characteristics

• Self-Discharge 2.5%/day at 25C• Higher T, Higher Self-discharge

1200

1000

800

600

400

200

0

Capacity mAh

0 10 20 30 40 50 60 70Time: Days

150 C

250C

350C

450C


Advantages and Limitations: NiMH Battery

Advantages:• Fast and simple charge: constant current charge

• 50-100% more capacity/energy than NiCd

• Environmentally friendly: Contains only mild toxic metals

• No transportation regulation

• Cheaper than Li-Ion, but 20% more expensive than NiCd

Disadvantages

• Limited service life cycles: 400-500

• High self-discharge at high temperature (30% per month)

• High maintenance: regular full discharge to prevent voltage depression

• Generate more heat during charge than NiCd

12


Treatment of NiMH Battery

Healthy habits:Most stable in discharged state. If used in application needing prolonged trickle-charge (stand-by), should be periodically fully dischargedAvoid high temperature as corrosion accelerates and charge efficiency drops

Degradation mechanisms: Voltage depression. Cathode material changes into inactive crystalline form when stored in charged state. Less expressed compared to NiCd because of newer technologies used in cathode material and electrolyte preparation. Prevention – discharge regularly. Repair: Charge/discharge several times to 0.7V. For cases of severe capacity loss, keep at 0.7V for prolonged time (24 hrs)


Li-Ion Rechargeable Battery

13


Li-Ion Battery Construction

Separator: Melt at 1500 CPositive Electrode: LiCoO2 (200um)Negative Electrode: Carbon (100-200um)Current Collector: Cu (Anode)-20um

Al (Cathode)-20umCasting: Steel


Li-Ion Intercalation

Cathode: LiCoO2 Anode: Carbon

14


Li-Ion Chemistry

Discharge: Minimum 2.7VLi ions leave carbon matrix and are adsorbed by LiCoO2 matrix

Charge: Maximum 4.1(Coke Anode) or 4.2V(Graphite)Li ions are extracted from LiCoO2 and are intercalated into carbon matrix. Li-carbon intercalation compound is extremely active, therefore only organic electrolytes can be used. Contact of interiors with air or water destroys the battery.

Specified Cell voltage: 3.6V

Cathode +

Anode +


Li-Ion 18650 Charging: CC/CV

CC Mode

CV Mode

15


Charge VoltageThe voltage between the charging terminals should be no more than 4.20 V (Set this at 4.20 V (max) after taking into account fluctuations in power supply voltages, temperature deviations, etc.). The charge voltage varies according to model cell chemistry. Cells with coke based anodes use 4.1V. Check manufacturer’s specifications for charging.

Charge CurrentCurrent should be limited to 1C rate to prevent overheating and resulting accelerated degradation

Li-Ion Charging: Special Requirements


Ambient Temperature of the Battery Pack During Charge0°C to 45°C. Charging at higher temperature results in accelerated aging

Low-Voltage Battery Pack ChargeWhen the voltage per cell is 2.9 V or less, charge using a charge current of 0.1 C rate or less. This is to provide recovery of passivating layer which might be dissolved after prolonged storage in discharged state. It also prevents overheating at 1C charge when partial Cu-deposition on anode shorted cells on over discharge.

Termination of ChargingThe system will determine that the battery is full by detecting the charge current.Stop charging once the current has reached 1/10-1/20C rate. This value can vary between different manufacturers.

Charge TimerA total charge timer and a charge completion timer should be included. Usuallycharging time can be limited to 3-5 hrs.

Li-Ion Charging: Special Requirements (Cont.)

16


• Over charging shortens battery cycle life

Charge Voltage Affects Battery Service Life

4.35V

4.3V4.25V

4.2V


Charge CurrentCurrent should be limited to 1C rate to prevent overheating and resulting accelerated degradation.

“Factors that affect cycle-life and possible degradation mechanisms of a Li-ion cell based on LiCoO2”, Journal of Power Sources 111 (2002) 130-136

Charge Current vs Battery Degradation

1.0C1.1C1.3C1.5C

2.0C

17


C10uF

AC

USB

STAT1

STAT2

USBPG

ACPG

ISET2

ISET1

TMR

VSS

OUT

OUT

OUT

CE

BAT

BAT

TS

LDO

PSEL

DPPM

103AT

RT2

BQ2403x

D1 D2

D4

D3

D6

D5

High: 500mALow: 100mA R1

R2

High: ACLow: USB

3.3V/20mA

HighEnable

System LoadAC Adapter

USB

RT1

R3

• Battery temperature monitoring• Pre-charge, 10% fast charge current, safety timer• Battery Management features maximize battery capacity, cycle life and safety

Linear Battery Charger Example


Li-Ion 18650 Discharge

1/5C1C

2C

Battery AC model

18


Li-Ion 18650 Temperature Discharge Profile

Using organic electrolyte make internal resistance of Li-ion battery more temperature dependent than other batteries.

Self-heat Effect (Internal impedance decrease when T increase)


Li-Ion Storage Characteristics

• 2-3% Self-discharge per month at 200C• Self-discharge rate doubles for every 100C

600C

400C200C

00C100

80

60

40

20

0

Capacity %

0 2 4 6 8 10 12Storage Time: Month

19


Li-ion Battery Pack Protection Function

Short Circuit Protection (SCP)

Overcurrent Protection (OCP)

Over Discharge Protection (ODP)

Over Temperature Protection (OTP)

Cell Over voltage Protection (OVP):1st level protection: > 4.25V, Turn off the protection MOSFETs2nd level protection: > 4.45V, blow the fuse


TempSensor

IntegratingADC

CoulombCounting

SMBus

Supply Voltage

Pack +

Pack -

ADCInstantaneous

Voltage

OV, UV, and Safety

HardwareCommunication

32KHzTime Base -Ext/Int Option

Gas GaugeProtection IC

OVP, OCPODP, OTP UVP, SCPbq29330

bq29312AI2C

Li-IonCell

Chemical Fuse

DischargeMOSFET

ChargeMOSFET

SenseResistor10mΩ

Smart Battery Pack Electronics

TI Gas Gauge Chip-set: bq20z80-bq29330; bq2084-bq29312A

bq20z80bq2084

20


Advantages and Limitations

Advantages:

• High energy volumetric and Gravimetric density

• One Li-Ion cell voltage =3 × NiMH or NiCd Cell Volatge

• Low self-discharge: 2-3%/month, double every 100C

• Low maintenance- no periodic discharge is needed, no memory

Limitations:

• Requires Protection Circuit: OVP, OCP, ODP, OTP, UVP, OSP.

• Subject to aging, storing in a cool place, and 40-50% SOC.

• Expensive due to rare Cobalt material and manufacture, 40% higher

in cost than NiCd.

• Subject to transportation regulations


Treatment of Li-Ion Battery

Healthy habits:Most stable in 50% charged state. High voltages accelerate corrosion and electrolyte decomposing. Charging should be limited to maximal voltage specified by manufacturer (4.2 or 4.4 V)Short deep discharge is not detrimental, but long storage in discharge state results in dissolution of protective layer and resulting capacity loss.High temperature is main killer. Take out battery out of notebook to prevent exposure of the pack to high temperatures. Use battery soon after manufacturing. Discharge capacity degrades even if not usedStorage at low temperatures increases shelf lifeIf used in stand-by application, charger should terminate charging and not resume until state of charge drops below 95%. Trickle charging is not recommended.

Degradation Mechanisms: Reaction of Li-carbon compound with electrolyte. Despite protective layer, this reaction is always ongoing and is accelerated by high voltage and high temperature.

21


Overview of Different Chemistries: Energy Density

2-10C3651354.2V3.6VLi-Ion

1-2C246691.6V1.2VNiMH

10-20C200461.6V1.2VNiCd

1-10C103362.7V2VLead-Acid

Dis. RateWh/lWh/kgMax. voltageAvg. VoltageChemistry


Summary

• Review: NiCd, NiMH and Li-Ion battery charge/discharge characteristics

• Battery aging, self-discharging and better treatment practice

• Li-Ion battery charge requirements vs cycle life

• Battery charger, gas gauge and protection ICs should be designed

based on the battery chemistry characteristics

22

Battery Cell Balancing: How to Balance

Yevgen Barsukov

Portable Power Management


Manifestation of the Problem

If two or more cells are connected in series, voltages of each cell diverge during battery use

Voltage difference can eventually bring one of the cell either to unsafe high voltages (explosion danger or early charge termination) or to unsafe low voltages (possible cell degradation).

Voltage differences between the cells can increase when current is applied to the battery, especially near the end of discharge.

Traditional solution was “to fight the symptoms” e.g. trying to reduce voltage difference when it is most noticeable

23


Reasons for Voltage Divergence

State of charge (SOC) differences between the cells

caused by increased discharge of some cell

• internal micro-shorts • external short-term discharge

caused by differences in total cell capacity. Same amount of passed charge causes differences in the SOC

The impedance differences between the cells has an effect on voltage when load is applied


Effect of SOC Unbalance at Battery Voltage

% SOC unbalance remains constant during entire discharge

Voltage differences between the cells vary with state of charge because dV/dSOCvaries with SOC (Fig. a)

Fastest voltage change with SOC is observed near the end of discharge, therefore highest voltage differences at the same SOC unbalance will be observed there

Fig. 1 (a) OCV dependence on SOC (b) OCV differences at different states of charge between

two cells with SOC unbalance of 1%

0 50 1000

500

1000

SOC, %

Dev

iatio

n, m

V

(a)

(b)

0 50 1002

3

4

5

SOC, %

Vol

tage

, V

24


Increased Variation Because of Sharp Impedance Change

Cell impedance increases rapidly near end of discharge

Small differences in SOC cause large differences in impedance near the end of discharge

Because cell voltage is V=OCV + I*R, larger R(SOC) results in much lower voltages for cells with lower SOC

Fig. 2 Voltage differences under C/2 load at different states of chare between cells with 1% of SOC unbalance. Solid line shows differences for OCV case for comparison.

0 50 1000

2000

4000

6000

SOC, %

Dev

iati

on, m

V


Total Cell Capacity Differences

SOC = SOC0 - Qpassed/Qmax

If Qmax1 for cell 1 is larger than Qmax2 for cell 2, the same passed charge Qpassed will result in different SOC1 and SOC2

Total cell capacity can vary due to manufacturing variations

Larger differences develop during battery pack life because of faster degradation of cells exposed to higher temperatures.

25


Impedance Differences from Cell to Cell

Impedance spectra differences between 50 cells in 1 batch for manufacturer (a) and manufacturer (b). Data is shown from 1kHz (left) to 10mHz (right)

Z’(Ohm)

• Impedance at low frequency has main iffect on voltage under load• +-15% deviation of low frequency impedance are typical within a batch of batteries

0 0.062 0.084 0.11 0.13 0.15R(Z) - Ω

-Im

(Z) -

Ω

0.05

0.025

0

10 mHz

1 kHz

Manufacturer (a)

10 mHz

1 kHz

0 0.042 0.064 0.086 0.11 0.13R(Z) - Ω

-Im

(Z) -

Ω

0.05

0.025

0

Manufacturer (b)


Effect of Cell-to-Cell Impedance Variation

Fig. 4 Voltage differences between 2 cells with 15% impedance unbalance at C/2 discharge rates, solid line. Doted line shows difference between the cells with 1% SOC unbalance for comparison.

0 20 40 60 80 1000

50

100

deviation from impedance variationdeviation from 1% disbalance

SOC

Dev

iatio

n, m

V

For new cells impedance can differ by +/-15%.

When current is applied, this will cause either positive or negative voltage shift by I*R

Even at moderate rate of C/2, effect of this difference on voltage exceeds that of the SOC effect

With battery aging, impedance differences from cell to cell increase

Deviation from Impedance variation with C/2Deviation from 1% unbalance

26


Effect of Unbalance on Battery Performance

Increase degradation of a cell with higher SOC during charging due to voltage above recommended 4.2V value.

Increased degradation reduces Qmax, those increasing SOC even more, creating a cycle of self-destruction

Early termination of whole pack charging because of single “high” cell, pack is never charged to full

Discharge capacity is therefore reduced

Early discharge termination if a low SOC cell reaches cell undervoltage threshold before other cells

Useable discharge capacity is reduced

Fig. 5 Individual cell voltage in 4 serially connected cells stack at the end of charge vs. SOC deviation of high cell.

0 5 10 15 20 25 304

4.1

4.2

4.3

4.4

4.5

4.6Low cellNormal cells

Capacity deficiency, %

Cel

l vol

tage

, V


Hardware Implementation of Cell Balancing

Capacitive current redistribution – more complex implementation, no energy loss. Low balancing rate. Moderate size requirements.

Inductive current redistribution – most complex implementation, no energy loss. High balancing rate can be supported. Highest size requirements

Current bypass – simplest method, but some energy loss. Both high and low rates can be implemented. Easy to integrate, small size and cost.

27


Capacitive Charge Redistribution

Charge redistribution eliminates energy loss to overall system

Capacitor charging efficiency of 50% only loses part of the “unbalance energy”.

Energy loss in the switches still has to be considered for thermal management

Difficult to integrate

Small voltage differences at most of SOCs causes low balancing rate

Used mostly in automotive systems due to cost/size

(a)

(b)

Fig. 7 (a) simple capacitor-based shuttle cell balancing circuit (b) shuttle circuit with remote cells connection capability. Reprinted from S. W. Moore and P. J. Schneider, Delfhi application note 2001-01-0959


Inductive Energy Redistribution

Pack energy is transferred into a single cell by directing pack current through a transformer which is switched to one of the cells that needs additional charge

100% coulometricefficiency eliminates energy loss to overall system

Eliminates low balancing rate due to small voltage differences between the cells.

Large size due to use of inductors

Mostly used in automotive applications due to cost/size

Fig. 8 Inductive converter cell balancing circuit. Reprinted from S. W. Moore and P. J. Schneider, Delfhiapplication note 2001-01-0959

28


Current Bypass

Simple implementation of cell-balancing includes a FET placed in parallel with each cell Easy to integrate, for example bq29312Aprotection/balancing IC.Sufficient to counter-act continuous unbalancing processes. Max self-discharge unbalance rate in 3s2p pack is 100uA, while typical 700 Ohm by-pass provides for 2.2mA balancing.External FETs can be used if rapid removal of large unbalance is needed Energy loss occurs during balancing, however it is negligible at the unbalance levels found in portable devices

Balancing controller

R1 R2 R3 R4 R5

Bypassing Current:

FET

CELLbypass R2R1R

VI

++=

Ibypass

RFET


Current Bypass Cell Balance Example: bq29312A

29


Algorithmic Implementation of Cell Balancing

Cell voltage based – turn ON balancing whenever voltage difference is detected

SOC based – measure SOC, assume equal total capacity and turn ON bypass until calculated amount of unbalanced charge is passed.

SOC and total capacity based – measure SOC and total capacity, bypass until calculated amount of unbalanced charge is passed.


Test Results for Cell Voltage Based Balance

Simple implementation

Impedance differences between cells can mislead algorithm

Improved version can only balance near end of charge to reduce effect of I*R drop/rise

More efficient implementation not only compare neighboring cells, but makes decision based on all cell voltages.

Example of implementation: bq2084

Despite the improvements, the need of relying on voltages limits balancing rate to “high difference”SOC regions

Fig. 9 Cell open circuit voltages of 4 cells pack at the end of charge during balancing. Initial unbalance 10%

Cell Voltage at the end of Charge (mV)

4120

4140

4160

4180

4200

1 6 11 16 21 26

Cycle Number

Cell 1

Cell 2

Cell 3

Cell 4

30


SOC and Total Capacity Based

Implementation requires means for measuring SOC for each cell and total cell capacity

Can take advantage of existing gas-gauging solution such as “impedance track” that already has this information

Example implementation: bq20z80

Can balance continuously during charge, discharge or even during inactivity period, regardless of SOC effect on voltage

Independence on voltage differences increases balancing rate about twice

Balancing not distorted by cell impedance differences

0 5 10 15 20 25 304140

4150

4160

4170

4180

cell 0cell 1cell 2

Cycle Number

Cel

l Vol

tage

, mV

Fig. 10 Evolution of cell voltages during SOC/Qmax balancing, starting from initial 2% down (cell 1) and 2% up (cell 2) unbalance


Summary

Cell unbalance can cause significant degradation to battery pack performance

Current by-pass implementation is sufficient for most of cases, but charge redistribution can have advantage in application where energy loss is critical

Voltage difference based algorithms are simply to implement, however are affected by impedance deviations and have lower balancing rate

SOC and total capacity based method achieve best balancing accuracy and rate, but have to relay on already existing

31

Portable Power Battery Management

Texas Instruments

Battery Authentication Architecture andImplementation for Portable Devices


Counterfeit Battery

– Problem: Cheap Replacement Batteries• Functionality Removed• Safety Circuits Removed• Protection Circuit Removed

– Result:• Loss of public confidence in technology• Loss of corporate goodwill• Loss of potential revenue

PACK+

PACK-

Battery Pack

TS

RT: 103AT

Protector

GasGauge

HDQ

32


In General:

A simple and cost effective method to identify and validate identity.

Specific to Peripherals:

A simple and cost effective method to ensure peripherals come from authorized sources.

What is authentication?


• Form factorsStrength: Economies of scaleWeakness: Hard to revise

• LabelingStrength: CheapWeakness: Easily copied & moved around

• User InterventionStrength: Informed consentWeakness: Requires user motivation, difficult to enforce

Current Options

33


Battery Electrical Identification (ID) Scheme

• Fixed Challenge, Fixed Response (ID)• bq2022/23 is an example• Counterfeit by replicating ID through oscilloscope• Adding cost to the system

PACK+

PACK-

SDQ

VSS

VSS

bq2022

SDQ

Battery PackProtector

TSRT: 103AT

CHALLENGER

CommandExpected Answer

RESPONDER

ID Response

Match?

Host System

Battery orPeripheral

AuthenticatedFail

NO YES


Is An ID Approach the Right Choice?

Strength: Requires dedicated electronics• May be cost prohibitive to unauthorized parties

Weakness: Not challenging to reproduce with generic hardware• Static challenge• ID may change from device to device, but only one is needed

to clone.

34


CHALLENGER

ChallengeAuthenticationTransform

Expected Answer

RESPONDER


Calculated Answer

Match?

Yes, thenAuthenticated

No, thenFail

Secret

Secret

Host System


• Challenge is randomnumber that changeseach time it is transmitted

• Secret can beshared or transmittedsecurely from one side to the other

• More securethan simple ID

• Need extra memory, algorithm development

Random Challenge-Response Authentication Scheme


Is A Challenge/Response Approach The Right Choice?

Strength:• Random from request to request• Introduces a secret, which is known only to the valid parties• The transform may be easily replicated, but the security lies in the

secret.

Weakness:• Requires dedicated secure memory on the host side.• Requires a small amount of additional code development and space,

dependent on the algorithm used.

35


Not a cryptographic functionSimple solution, small footprint, adds some burden to crack

CRC Approach(bq26150)

DisadvantagesAdvantagesAuthentication Transform

Proven cryptographicallyMore complicated algorithm that significantly increases the burden to crack

SHA-1/HMAC(bq26100)

Fixed inputs and fixed outputsSimplest solution, very small footprint, cheapID

(bq2022)

Summary of Authentication Schemes

No ID Simple IDDevice

bq2022

CRC basedAuthentication

bq26150

More SecureLess Secure SHA-1/HMACAuthentication

bq26100


Typical Application Circuit & System Diagram

Gas gauge

Q1 Q2

PACK+

PACK-

HDQ

VCC

VSS

SRP

SRN

BAT

HDQPHDQ

Protector IC

GNDGND

PWR

bq26150HDQHost Process

TI OMAP

Power Management

GPIO

TI Firmware or Customer Firmware

C10.1uF

R15kVSUPPLY

bq27000

Battery Pack

R2100k

36


What If Host Does Not Have Secure Memory?

CHALLENGER Challenge

RESPONDER


Peripheral Answer

Match?

Yes, thenAuthenticated

No, thenFail

Host Answer

Secret

Host System


Peripheral Answer

RESPONDER


Peripheral Answer

Secret


System dependent• Battery Packs:

– Allow discharge only

• Chargers:– Reduced charging rate

• Other Peripherals– Reduced functionality

May choose to simply log that an unauthorized peripheral was used for warranty information.

What if not authenticated?

37


Optimization of the Charger to System Interaction Enables Increased Handheld

Equipment Functionality

Battery ManagementTexas Instruments


1. Charger and system interactions that affect end equipment performance in commonly used topologies:

- Battery and charger connected to main system power rail

- Power path battery charger

2. Techniques that may be used to improve system performance

- Charger power management- Charger thermal management

3. Summary

Overview

38


Li-Ion Charge CC-CV Profile

IPRECHARGE

3.0V/Cell

Pre-charge

ICHARGE

4.2V/CellFast-charge

ITERMINATION

Taper Current

Battery Voltage

Constant Current Constant Voltage

Pre-chargeTimer Safety Timer


Charging with an Active System Load

ISYS: System load current

System

ISYS+

Adapteror USB

ICHG

IBAT

Charger

ICHG: Charger Output CurrentIBAT: Current going into the battery

(Effective charge current)

If charger is ON ,

with no system load: ICHG = IBAT

with a system load, ICHG = IBAT + ISYS

39


Charger output current is shared: ICHG = IBAT + ISYS

Charging with an Active System Load: Potential Issues

System

ISYS+

Adapteror USB

ICHG

IBAT

Charger

Potential Issues:

Timer faultNo terminationSystem lockup during startup


Issue 1: Pre-Charge Timer Fault

When in Pre-Charge Mode:

1. Charger output current is limited to pre-charge current2. System current “steals” charger output current3. Battery is charged at a very low rate !

Pre-charge timer may expire, turning off the charger

Solution: Keep the system off or in a low-power mode during pre-charge phase !

100mA80mA+

Adapteror USB

System

ICHG ISYS

IBATCharger 20mA

40


Issue 2: Safety Timer Fault

System

ISYS+

Adapteror USB

ICHG

IBAT

Charger

When in Fast Charge Mode:

1. Charger output current is limited to fast charge current2. System current “steals” charger output current3. Battery is charged at a lower rate !

Charge safety timer may expire, turning off the charger

Solution:• Increase the safety timer timeout value• Increase the fast charge current value

700mA

50mA

750mA


Issue 3: System Lockup During Initial Power-Up

System

ISYS+

Adapteror USB

ICHG

Charger

With a battery deeply depleted or no battery connected:

1. System voltage is initially at zero volts (pack open or no battery)2. Charger circuit detects a battery “short” and limits charge current3. System current needed for power-up exceeds available charge current4. System voltage stuck at very low value …..

System never starts

Solution: Reduce system start-up current

30mA10mA

41


Issue 4: Charge Termination NOT Detected

0

0.2

0.4

0.6

0.8

1

20 40 60 80Time

Cu

rren

t (A

)

IBAT

ICHG

ITERM

ISYS

+Adapteror USB

System

ICHG ISYS

IBATCharger

With the charger in regulation mode :

1. Charge current will taper down2. System current exceeds termination threshold3. Charger regulates system rail at charge voltage, ICHG > ITERM

Termination is not detectedCharger safety timer fault

Solution:Disable safety timer or supply additional current to the system


Supplying Additional Current to System

• A new circuit supplies the system load current directly from the input source during voltage regulation ; ICHG ≈ IBAT

• When charging is terminated, the STAT1 pin goes high and disconnects the supplement circuit.

• Design Example: VIN = 5V, VOUT = 4.2V, ISYS = 0.2A, Voltage drop across Rsupplement = 5 - 4.2 = 0.8V, Rsupplement = 0.8 / 0.2 = 4 Ω

QRSupplement

C1uF

IN

VCC

STAT1

STAT2

VSS

OUT

BAT

CE

TTE

ISET

BQ24013

D2

D1

AC Adapter

RSET

SystemVIN

ISYSIsupplement

42


0

1

2

3

4

2.4 2.7 3 3.3 3.6 3.9 4.2

1.24A(1800mAh Li-Ion)

VBAT (V)

700mA(1000mAh battery)

PLOSS (Watt)

Thermal Analysis of the Charger Stage

( ) CHGBATINLOSS I VVP ⋅−=

VIN= 5V+Adapteror USB

System

ICHG ISYS

IBATCharger

Under worst case conditions (high VIN, low BAT) :

1. Charger IC junction temperature is excessive:

Tj = 1300C @700mA, Tj = 192C @1.24A, 500C ambient temperature

Potential oscillatory behavior if charger has thermal shutdownCharger IC damage if charger has no thermal shutdown

Solution: Add thermal management to charger IC !


Time

ICHG

ICHG_THRM

Tj = 1250C

VIN

Thermal Management

Thermal management functions:• Regulate IC junction temperature by reducing

charge current , AND

• Turn off the charger when IC junction temperature

• is excessive

• Slows down the safety timers when the charge current is reduced by the thermal loop, avoiding a false safety timer fault

Common implementations:

• The IC junction temperature is regulated to a value just below the maximum operating junction temperature, 1250C typical

• The charger is turned off when the Charger IC

• junction temperature is excessive, 1500C typical

43


Battery Charger with Thermal Management

• Charger on/off control• Programmable timer – can be disabled by leaving TMR pin floating • Programmable charge current rate• Charger status: pre-charge, fast charge and fault detected• Integrated thermal loop• Charge Safety timer adjusted dynamically when thermal loop is active• Input over-voltage protection

C21uF

IN

TMR

STAT1

STAT2

VSS

OUT

BAT

CE

PG

ISET

bq24061/66

D2

D1

AC Adapter

RSET

System

VINC11uF


Adding Power Path Management to a System

• System power supplied from adapter through Q1• Charge current controlled by Q2• Ideal topology when powering system and charging battery

simultaneously is a requirement• Separates charge current path from system current path• No interaction between charge current and system current

C1

+

-

Adapter

System

Q1 Q2

Powering System

Charging Battery

Controller

44


Power Path Management: Potential Issues

Time

ICHG

VADP

IADP

Adapter current limit

ISYS System Current

Adapter voltage collapses

IADP

C1

+

-

Adapter

System

Q1 Q2

System Current

Charging Battery

ControllerIBAT

ISYS

VOUT

Input current : IADP = IBAT + ISYS

Issues:

• Input voltage collapses


Dynamic Power Path Management (DPPM)

C1

+

-

Adapteror USB

System

Q2CurrentControl

OutputControl

ICHG

IADP ISYS

Time

ICHG

System VoltageVDPPM

IADP AC adapter current limit

ISYS

DPPMMode

System voltage drops if (ISYS + ICHG) > IAC_LIMIT

Large voltage drops due to system load pulses generate undesired ripple at system power rail, causing reset or degrading system performance

DPPM function :

Reduces the charge current when the system voltage is below the user-defined Voltage threshold VDPPM

“Finds” maximum adapter power !!!

Battery Supplement Mode

VBAT

45


• AC adapter or USB port can power the system and charge the battery simultaneously• Dynamically reduces the charge rate to maximize adapter to system current• Battery Supplement enables use of “weak” adapters• Selectable charge current limits : USB 100/500mA, AC adapter up to 1.5A • Battery Management features maximize battery capacity, cycle life and safety

Dynamic Power-Path Management Charger

C10uF

AC

USB

STAT1

STAT2

USBPG

ACPG

ISET2

ISET1

TMR

VSS

OUT

OUT

OUT

CE

BAT

BAT

TS

LDO

PSEL

DPPM

103AT

RT2

bq2403x

D1 D2

D4

D3

D6

D5

High: 500mALow: 100mA R1

R2

High: ACLow: USB

3.3V/20mA

HighEnable

System LoadAC Adapter

USB

RT1

R3

Q1

Q2Q3


bq24032A Test Results

VAC: 1V/div

VOUT: 1V/div

0V

0A

ICHG: 1A/div

IOUT: 1A/div

VDPPM = 4.26V

DPPMMode

DPPMMode

Battery Supplement Mode

AC Adapter Voltage

System Voltage

Battery voltage

46


Summary

• Basic Configuration: system is directly connected to the battery

System rail interacts with battery charger

• Power Path: powers the system directly from the input

Eliminate charger-system rail interaction issues

• Dynamic Power Path Management and battery supplement mode

Prevent system issues during large system current load transients

Prevent battery MOSFET Body-diode conduction during large

system current load transients

Enable use of low cost wall adapters

• Thermal management


47


Circuit Design and Power Loss Analysis of a Synchronous Switching Charger with

Integrated MOSFETs for Li-Ion Batteries

Texas Instruments


• Battery Operated System

• Linear Battery Charger Challenges

• Why Switching Charger?

• EMI Analysis

• Loss Calculation and Analysis

• Switching Charger Example and Measurement

Results

• Summary

Outline

48


+

High Efficiency

DC/DCConverters

CPU Core

LED or LCD

I/O

Memory

BatteryCharger

?

Li-Ion

AC Adapter

Typical Battery Operated System

Digital Circuits


+

AdapterICHG

VBAT

Linear Charger

VIN

Linear Battery Charger

• Simple and low cost.

• Highest power dissipation from pre-charge to fast charge mode transition.

• Ideal for low charge current < 800mA.

• Thermal issue for ≥ 800mA charge current.

• For high charge current applications (Portable DVD, PMP)

( ) CHGBATINLOSS IVVP •−= 0

1

2

3

4

2.4 2.7 3 3.3 3.6 3.9 4.2

1.24A(1800mAh Li-Ion)

VBAT (V)

700mA(1000mAh battery)

PLOSS (Watt)

What’s the solution?

VIN=5.5V

49


Regulate the junction temperature

Thermal fold-back at 1250C by reducing charge current

Drawback: Increase the charge time, may expire the safety timer

Solution 1: Thermal Regulation

ICHG

TLIM

Tj

1250C

ICHG_TLIM

Time


Solution 2: High Efficiency Switching Charger

• Switching Charger: Higher than 90% Efficiency

For VIN=5.5V, VBAT=3.0V, ICHG=1.24A (1800mAh, 0.7C)

• Power Dissipation:

• Ideal for ≥1000mA charge current, high input and output voltage difference

• Typical Applications: Portable DVD Players, MP3 Players,….

18.40C1420CJunction Temp Rise

0.4-Watt3.1-WattPower Dissipation

Switching ChargerLinear Charger

50


L

C

RSNSAdapter

RT

PWM Controller

P-MOSFET

N-M

OS

FE

T

Battery

• Higher Efficiency but Larger Size and Higher Cost Compared to Linear Charger• Small Passive Component Size Needed: Increased Switching Frequency(Consideration #1)• High Switching Frequency: EMI issues (Consideration #2)

Synchronous Switching Charger Design Considerations

• Integrated Power MOSFETs: Smaller Size, Higher Die Temperature• Loss and Thermal Analysis (Consideration #3)


Inductor Size and Selection

• Low inductance at high switching frequency

CHGripplesIN

BAT

ripple

BATIN I 30%I ,f1

VV

IVV

L =−

= ∆∆

ISATURATION = (1.1-1.5) IPeak, max

,I21

II RippleeargChPeak +=

Design Tradeoff

Higher switching frequency, the smaller the inductor

Higher switching frequency, the higher the switching losses

51


Total Active Area Size Comparison

50 kHz vs 1.1 MHz Area Comparison

50 KHz Solution1.1 MHz Solution

Bq2410x EVM


EMI Considerations

EMI DefinitionThe interference of one piece of electronic equipment on the operation of

another by means of electromagnetic energy transfer.

• A generator of electromagnetic energy: (a source)

• Transmission of that energy between equipments: (a coupling means)

• A receptor circuit whose operation is negatively impacted by the

transmitted energy: (a victim circuit)

dtdv

Ci dtdi

Me ==EMI Principles:

Inductive Coupling Capacitive Coupling

52


Solutions to Reduce EMI

Adding filter (cost! may not work for radiated EMI)

Adding snubber (cost!)

Excellent layout (design difficulty)

Slower switching speed

- No extra cost

- Reduce both conducted and radiated EMI

- May impact efficiency


Major EMI Source of a Converter

T

VIN

VPH Waveform

High dv/dt and di/dtswitching on the phase node is the major EMI source in a Buck converter

IQ1

TD·T

D·T

IQ1 WaveformIOUT

VPH

VIN

IQ1 IOUT

53


Why Slower Switching Helps

1us

trise=5ns

1V

1us

trise=100ns

1V

Mag

nitu

de (d

B)M

agni

tude

(dB

)

Mag

nitu

de (d

B)

Frequency

5 6 7 8 9

5 6 7 8 9

Fourier Analysis

Fourier Analysis

100MHz

-175dB

100MHz

-200dB

About 25dB Less!

-100

-150

-200

-250100kHz 1MHz 10MHz 1GHz

-150

-200

-250

-100

100kHz 1MHz 10MHz 1GHz


CISPR (Europe)FCC

60

0

30

5040

2010

-10100M50M 200M 500M 1G

31dB

Switching Waveforms and EMI Spectrum @ Vin=9V Ichrg=1A

≈7ns

VPH:

≈20ns

VPH:

(dBuV)

(Hz)

After increasing the turn-on time by 3 times (bq24120/123)

Phase node voltage

Phase node voltage

CISPR (Europe)FCC

60

0

30

5040

2010

-10

(dBuV)

100M50M 200M 500M 1G(Hz)

46dB

• 15dB reduction around 150MHz

54


Loss Analysis: Synchronous Switching Charger

L

Cout

Q1

Q2

• MOSFETs

• Inductor

• RSNS

• Others

Conduction lossesSwitching LossesBody Diode Conduction Losses

Gate Drive Losses

Winding LossesCore Losses

Cap ESR Losses, IC losses …PCB Trace Conduction Losses

Driver and Controller

Cin

+

Resistive losses

RSNS


MOSFET Conduction Losses

L

C

IQ1 IOUT

Q1

Q2RDSON is a function of temperature, normally increases by 0.39%/°C

+= 22

1 12

1LOUTQ_RMS IIDI ∆

IQ1

IOUT

TD·T

∆IL

DSON2RMSCOND RIP ⋅=

IQ2

IOUT

∆IL

T(1-D)T

+−= 22

2 12

11 LOUTQ_RMS II)D(I ∆

IQ2

MOSFET Q1 MOSFET Q2

55


Upper MOSFET – Switching Losses

1ON_ISWt∆ 1ON_VSWt∆ 1OFF_ISWt∆

IDS

VDS

Turn-on Q1 Losses Turn-off

SWON_VSWON_ISWL

OUTINQ_swon f)tt()I

I(V.P ⋅+⋅−⋅⋅= 111 250 ∆∆∆

SWOFF_ISWOFF_VSWL

OUTINQ_swoff f)tt()I

I(V.P ⋅+⋅+⋅⋅= 111 250 ∆∆∆

1OFF_VSWt∆

2L

OUT

II

∆− 2L

OUT

II

∆+

INV

111 Q_swoffQ_swonQ_sw PPP +=The total switching losses of Q1:


Lower MOSFET – Body Diode Conduction & Reverse Recovery Losses

L

C

Q1

Q2

Ibd_Q2

Idrain_Q2

• Relates to the short periods of conduction before and after the on-time of Q2.

• Dead time -- 20–60ns typically

Vbd_Q2Ibd_Q2

Vbd_Q2

SWDTOUTBDQ_BD ftIVP ⋅⋅⋅⋅= 22

SWINRRQRR fVQP ⋅⋅=

QRR

56


Gate Drive Losses

L

C

Q1

Q2Gate Drive

2Q_DRV1Q_DRVDRV PPP +=

SW1Q_DRV1Q_g1Q_DRV fVQP ⋅⋅=

SW2Q_DRV2Q_g2Q_DRV fVQP ⋅⋅=


Loss Breakdown @ VIN=12V, VBAT=8.4V, ICHG=1.2A

C o n v e r te r Lo ss B re a k d o w n

0

0 .1

0 .2

0 .3

0 .4

0 .5

0 .6

0 .7

0 .8

1 2 3 4

Lo

ss

es

(W

Ta =2 5 °C

Ta =5 5 °C

FETs

InductorRsns

Capacitor

F ETs Lo ss B re ak d o w n

0

0 .0 5

0 .1

0 .1 5

0 .2

0 .2 5

0 .3

0 .3 5

1 2 3 4 5

Lo

ss

es

(W

Ta =2 5 °C

Ta =5 5 °C

Conduction

SwitchingDriver

Body diode

Qrr

87.4°C208.6°CIC Junction Temperature

32.4°C153.6°CTemperature rise

91.72%57.1%Efficiency

0.911W4.32WTotal Losses

SwitchingTam=55°C

LinearTam=55°C

57


Typical Applications Circuit

L10µH

RSNS0.1ΩADAPTER (9V or 12V)

RT103AT

PWM Controller

COUT10µF

VBAT

VIN

OUT

20

1

15

14

OUT

SNS

BAT

ISET1

ISET2

8

9

TS 12

VTSB

RT19.31k

1%

RT2442k

1%

RSET1: 7.5k

RSET2: 7.5k

VTSBC2: 0.1µF

C10.1µF

PACK+

PACK-

11

3IN

4

6

CIN10µF

C30.1µF

VSS

VCC

IN

10

2

19

5

7

STAT1

STAT2

PG

TTC

CE16

17

18

PGND

PGND

CELL

D3: Charge

R2: 1.5kΩ D2: Done

D1: Adapter

bq24103

CTTC: 0.1µF

R1:1.5kΩ

R3: 1.5kΩ

Q1

Q2

13

VIN=9V, Battery: 2-cell in series, ICHG=1.25A, Safety timer: 5 hours, IPRECHG = 125mA, Temperature range: 0-45C


Efficiency - Calculation vs Measurement

VIN=12V, VBAT=8.4V

Iout (mA)

1000 20000 500 1500

Effic

ienc

y (%

)

Calculation Results

100

80

60

40

20

0

Iout (mA)

Effi

cien

cy (

%)

Measurement Results

100

80

60

40

20

01000 20000 500 1500

0 500 1000 1500 2000

58


Summary

• Linear Charger: Low cost and charge current <0.8A

• Switching Charger: Ideal Solution for higher charge current

• Expensive and complicated compared with Linear Charger

• Reducing the voltage slew rate dv/dt can reduce EMI

• Loss analysis helps thermal design

• Main Applications: Portable DVD Players, PMP……


Documents

Battery Selection, Safety, and Monitoring in Mobile Applications