M306N4FGTFP#U0 Renesas Electronics America, M306N4FGTFP#U0 Datasheet - Page 232

MCU 5V 256K T-TEMP PB-FREE 100-Q

M306N4FGTFP#U0

Manufacturer Part Number
M306N4FGTFP#U0
Description
MCU 5V 256K T-TEMP PB-FREE 100-Q
Manufacturer
Renesas Electronics America
Series
M16C™ M16C/6Nr
Datasheets

Specifications of M306N4FGTFP#U0

Core Processor
M16C/60
Core Size
16-Bit
Speed
20MHz
Connectivity
CAN, I²C, IEBus, SIO, UART/USART
Peripherals
DMA, WDT
Number Of I /o
85
Program Memory Size
256KB (256K x 8)
Program Memory Type
FLASH
Ram Size
10K x 8
Voltage - Supply (vcc/vdd)
4.2 V ~ 5.5 V
Data Converters
A/D 26x10b; D/A 2x8b
Oscillator Type
Internal
Operating Temperature
-40°C ~ 85°C
Package / Case
100-QFP
For Use With
R0K3306NKS001BE - KIT DEV RSK RSK-M16C/6NKR0K3306NKS000BE - KIT DEV RSK RSK-M16C/6NK
Lead Free Status / RoHS Status
Lead free / RoHS Compliant
Eeprom Size
-

Available stocks

Company
Part Number
Manufacturer
Quantity
Price
Company:
Part Number:
M306N4FGTFP#U0M306N4FGTFP
Manufacturer:
TI
Quantity:
3 001
Company:
Part Number:
M306N4FGTFP#U0M306N4FGTFP
Manufacturer:
RENESAS
Quantity:
36
Company:
Part Number:
M306N4FGTFP#U0M306N4FGTFP
Manufacturer:
Renesas Electronics America
Quantity:
10 000
Company:
Part Number:
M306N4FGTFP#U0M306N4FGTFP
Manufacturer:
RENESAS/瑞萨
Quantity:
20 000
Company:
Part Number:
M306N4FGTFP#U0M306N4FGTFP#UKJ
Manufacturer:
Renesas Electronics America
Quantity:
10 000
M16C/6N Group (M16C/6N4)
Rev.2.40
REJ09B0009-0240
16.2.5 Current Consumption Reducing Function
16.2.6 Output Impedance of Sensor under A/D Conversion
When not using the A/D converter, its resistor ladder and reference voltage input pin (VREF) can be
separated using the VCUT bit in the ADCON1 register. When separated, no current will flow from the
VREF pin into the resistor ladder, helping to reduce the power consumption of the chip.
To use the A/D converter, set the VCUT bit to 1 (VREF connected) and then set the ADST bit in the
ADCON0 register to 1 (A/D conversion start). The VCUT and ADST bits cannot be set to 1 at the same time.
Nor can the VCUT bit be set to 0 (VREF unconnected) during A/D conversion.
Note that this does not affect VREF for the D/A converter (irrelevant).
To carry out A/D conversion properly, charging the internal capacitor C shown in Figure 16.10 has to be
completed within a specified period of time. T (sampling time) as the specified time. Let output impedance
of sensor equivalent circuit be R0, internal resistance of MCU be R, precision (error) of the A/D converter
be X, and the resolution of A/D converter be Y (Y is 1024 in 10-bit mode, and 256 in 8-bit mode).
Figure 16.10 shows the Analog Input Pin and External Sensor Equivalent Circuit.
When the difference between VIN and VC becomes 0.1 LSB, we find impedance R0 when voltage
between pins VC changes from 0 to VIN-(0.1/1024) VIN in time T. (0.1/1024) means that A/D precision
drop due to insufficient capacitor charge is held to 0.1 LSB at time of A/D conversion in 10-bit mode.
Actual error however is the value of absolute precision added to 0.1 LSB.
When f(φAD) = 10 MHz, T = 0.3 µs in the A/D conversion mode with sample & hold. Output impedance R0
for sufficiently charging capacitor C within time T is determined as follows.
Thus, the allowable output impedance of the sensor equivalent circuit, making the precision (error) 0.1
LSB or less, is approximately 13.9 kΩ. maximum.
Apr 14, 2006
VC is generally VC = VIN {1 – e
And when t = T,
Hence, R0 = –
T = 0.3 µs, R = 7.8 kΩ, C = 1.5 pF, X = 0.1, and Y = 1024. Hence,
R0 = –
1.5 ✕ 10
page 208 of 376
0.3 ✕ 10
C • ln
–12
VC=VIN –
e
• ln
C (R0 + R)
T
-6
C (R0 + R)
X
Y
1024
1
0.1
1
– R
Y
X
T = ln
–7.8 ✕10
T
VIN = VIN(1 –
C (R0 + R)
=
1
Y
X
X
Y
3
t
= 13.9 ✕ 10
}
X
Y
)
3
16. A/D Converter

Related parts for M306N4FGTFP#U0