MAX17059EVKIT# Maxim Integrated, MAX17059EVKIT# Datasheet - Page 7

• Manufacturer Part Number: MAX17059EVKIT#
• Description: Power Management IC Development Tools MAX17059 Eval Kit
• Manufacturer: Maxim Integrated
• Series: MAX17058, MAX17059r

Specifications of MAX17059EVKIT#

• Part # Aliases: 90-BCA1K#E00

The MAX17058/MAX17059 ICs simulate the internal, non-
linear dynamics of a Li+ battery to determine its state of
charge (SOC). The sophisticated battery model consid-
ers impedance and the slow rate of chemical reactions in
the battery
The ModelGauge algorithm performs best with a custom
model, obtained by characterizing the battery at multiple
discharge currents and temperatures to precisely model
it. Contact Maxim if you need a custom model. At power-
on reset (POR), the ICs have a preloaded ROM model
that performs well for some batteries.
In coulomb counter-based fuel gauges, SOC drifts
because offset error in the current-sense ADC measure-
ment accumulates over time. Instantaneous error can be
very small, but never precisely zero. Error accumulates
over time in such systems (typically 0.5%–2% per day)
and requires periodic corrections. Some algorithms cor-
rect drift using occasional events, and until such an event
occurs the algorithm’s error is boundless:
• Reaching predefined SOC levels near full or empty
• Measuring the relaxed battery voltage after a long
• Completing a full charge/discharge cycle
Figure 2. Block Diagram
CELL
GND
period of inactivity
V
DD
(Figure
ADC (V
REFERENCE
ModelGauge Theory of Operation
VOLTAGE
BIAS
CELL
GROUND
2).
)
IC
Detailed Description
MAX17058
MAX17059
Fuel-Gauge Performance
INTERFACE
TIME BASE
MACHINE
(32kHz)
STATE
(SOC)
I
2
C
1-Cell /2-Cell Li+ ModelGauge ICs
QSTRT
ALRT
SDA
SCL
The ModelGauge algorithm requires no correction events
because it uses only voltage, which is stable over time. As
the SOC accuracy without full/empty/relax shows the
algorithm remains accurate despite the absence of any
of the above events; it neither drifts nor accumulates
error over time.
To correctly measure performance of a fuel gauge as
experienced by end-users, exercise the battery dynami-
cally; accuracy cannot be fully determined from only
simple cycles.
The open-circuit voltage (OCV) of a Li+ battery uniquely
determines its SOC; one SOC can have only one value of
OCV. In contrast, a given V
ferent values of OCV because V
OCV, load, temperature, age, and impedance, etc.; one
value of OCV can have many values of V
one SOC can have many values of V
not uniquely determine SOC.
Figure 3
and 72% SOC.
Even the use of sophisticated tables to consider both
voltage and load results in significant error due to the
load transients typically experienced in a system. During
charging or discharging, and for approximately 30min
after, V
been affected by the preceding hours of battery activity.
ModelGauge uses voltage comprehensively by using
voltage measured over a long period of time.
Figure 3. Instantaneous Voltage Does Not Translate Directly to
SOC
MAX17058/MAX17059
CELL
100%
shows that V
80%
60%
40%
20%
0%
and OCV differ substantially, and V
0
3.81V = 2%
1
Battery Voltage and State-of-Charge
V
2
CELL
CELL
3
3.81V
TIME (Hr)
4
CELL
= 3.81V occurs at 2%, 50%,
3.81V = 72%
5
CELL
6
can occur at many dif-
SOC
3.81V = 50%
7
CELL
is a function of time,
8
CELL
, so V
4.2V
4.0V
3.8V
3.6V
3.4V
3.2V
. Therefore,
CELL
CELL
can-
has
7