View detail for UHF Transmitter IC ATA5749

APPLICATION NOTE
UHF Transmitter IC ATA5749
ATA5749
Introduction
With the new UHF transmitter ATA5749 Atmel® offers an optimized solution for unidirectional access systems. The transmitter is primarily designed for automotive applications
such as Tire Pressure Monitoring Systems (TPMS), Remote Key-less Entry (RKE) and
Passive Entry Go (PEG). In addition the ATA5749 can also be used for any access and
remote control system in industrial applications. This application note provides technical
background information which helps to quickly understand and apply the device.
General Description
The ATA5749 is an integrated fractional-N-PLL transmitter that operates in the Industrial
Scientific and Medical (ISM) frequency band. The IC transmits signals with Amplitude Shift
Keying (ASK) and Frequency Shift Keying (FSK) modulation with a data rate of up to
40kBits/s in Manchester coding. Several features can be set by using the SPI interface, for
example, the operating frequency in the range of 300MHz to 450MHz and the transmit
power in the range of –0.5dBm to +12.5dBm. To ensure maximum battery lifetime the
transmitter features low current consumption, for example, just 7.3mA at 5.5dBm output
power (typical). Due to the small-sized TSSOP10 package only few external elements are
needed to realize an end application with minimum board space consumption (please refer
to Figure 1-1 on page 3).
9169C-RKE-05/15
Figure 1 shows the Pinning of the ATA5749, whereas the pin description is listed in Table 1.
Figure 1.
Table 1.
ATA5749 Pinning
1
10 EN
SDIN_TXDIN
2
9
GND
SCK
3
8
VS
ANT2
4
7
XTO1
ANT1
5
6
XTO2
ATA5749
ATA5749 Pin Description
Pin
2
CLK
Symbol
Function
1
CLK
CLK output
2
SDIN_TXDIN
Serial bus data input and TX data input
3
SCK
Serial bus clock input
4
ANT2
Antenna interface
5
ANT1
Antenna interface
6
XTO2
Crystal/CLOAD2 connection
7
XTO1
Crystal/CLOAD1 connection
8
VS
Supply Input
9
GND
Supply GND
10
EN
Enable input
ATA5749 [APPLICATION NOTE]
9169C–RKE–05/15
1.
Circuit Block Descriptions
This section describes the individual circuit blocks, providing more detailed information than in the ATA5749 datasheet.
Figure 1-1 illustrates the transmitter's system block diagram.
Figure 1-1. System Block Diagram
ATA5749
1
Clock
Driver
CLK
10
EN
Power
up/down
1
FREQ[0:14]
Digital
Control
and
Registers
FSEP[0:7]
CLK_OFF
GND
Frac.
Div.
PFD
8
VS
CP
EN_CLK
4
9
433_N315
SCK
FSK_mod
ASK_mod
3
PWR[0:3]
2
SDIN_TXDIN
DIV_CNTRL
4 or 8
ANT2
7
XTO1
LP
XTO
(FOX)
5
ANT1
PA
6
VCO
XTO2
PLL
1.1
Crystal Oscillator (XTO)
This block is based on the Pierce oscillator circuit with an amplitude regulation. The definition of the crystal oscillator’s
features and functionality is fixed. To start the crystal oscillator, pin EN must be set. The amplitude detector, which is
implemented in the crystal oscillator block, monitors the oscillation amplitude. If the defined amplitude is reached, an internal
signal bit XTO_RDY will be set (see Figure 1-1) and the clock output (pin CLK) will be switched on.
This method has two benefits. First it guarantees a stable PLL reference frequency and an optimum clock frequency
accuracy. Second, the parasitic coupling from the clock output signal to the crystal oscillator pins (XTO1 and XTO2) - which
may disturb the start-up process of the crystal oscillator - will not be critical since the clock output signal will be generated
only if the oscillation amplitude is high enough. Nevertheless, the board layout must be designed properly to minimize
parasitic coupling between the CMOS output of the pin CLK and the crystal oscillator.
The time period between the activation of the IC (EN=’HIGH’) and the output of the clock signal is defined as the crystal
oscillator's start-up time, which mainly depends on the crystal’s motional capacitance (Cm). The typical value of 200µs can
be obtained using a crystal with a motional capacitance (Cm) of 4fF and a shunt capacitance (C0) of 1.5pF. The higher the
motional capacitance (Cm) the faster the crystal oscillator's start-up.
ATA5749 [APPLICATION NOTE]
9169C–RKE–05/15
3
The ATA5749’s crystal oscillator features a high oscillation margin (negative resistance) of typically RXTO12_START > 1500.
This helps to reduce the failure rates caused by sleeping crystal phenomena. The external load capacitors must be selected
so that the XTO oscillates on the loaded resonance frequency of the specified crystal. To ensure the specified start-up
behavior and a low XTO current consumption it is recommended to choose a crystal with a load capacitance (CL) of 9pF.
1.2
Fractional-N PLL (Phase-locked Loop)
The Table 3-1 and Table 3-2 in Section 3. “ANNEX” on page 21 show the ATA5749’s register configuration. The
configuration of FREQ[0:14], FSEP[0:7] and S434_N315 influences the frequency generated by the PLL block (please refer
to Figure 1-1 on page 3). The ATA5749 provides two different VCO tuning ranges: 300MHz to 368MHz, and 367MHz to
450MHz. As can be seen in Table 1-1, the bit S434_N315 in the register sets the VCO’s RC oscillator and the fractional
divider of the PLL.
Table 1-1.
Bit S434_N315 Configuration
S434_N315
ATA5749 Frequency Range
LOW
300MHz to 368MHz
HIGH
367MHz to 450MHz
In case of FSK modulation the carrier frequency will be internally modulated by FSK_mod Signal, which is identical to the
available signal on the pin SDIN_TXDIN.
Table 1-2 shows the calculation of the RF output frequency based on the setting of FREQ[0:14], FSEP[0:7], S434_N315 and
the value of FSK_mod (SDIN_TXDIN).
Table 1-2.
RF Output Frequencies’ Calculation Based on the Register Setting
S434_N315 = ‘LOW’
S434_N315 = ‘HIGH’
fTX_FSK_mod='LOW'
 FREQ + 0.5 
24 + ----------------------------------  f XTO
16384
 FREQ + 0.5 
32.5 + ----------------------------------  f XTO
16384
fTX_FSK_mod='HIGH
 FREQ + FSEP + 0.5 
24 + ------------------------------------------------------  f XTO
16384
 FREQ + FSEP + 0.5 
32.5 + ------------------------------------------------------  f XTO
16384
FSEP-------------f
32768 XTO
fDEV__FSK
+ FSEP FREQ
---------------------------------+ 0.5


2
24 + -------------------------------------------------------  f XTO
16384
fRF (1)
Note:
1.
FSEP-------------f
32768 XTO
+ FSEP FREQ
---------------------------------+ 0.5


2
32.5 + -------------------------------------------------------  f XTO
16384
The ASK carrier is corrected by adding FSEP/2 to FREQ for ASK transmission. If FSEP is an odd number, the
FSEP/2 value is rounded to the lower next integer value, therefore, the ASK center frequency may differ to the
FSK center frequency by 396Hz.
Definition:
● In ASK modulation fRF is the center frequency of the transmitted signal.
●
In FSK modulation fRF is the virtual carrier that is defined using the Mark and Space Frequencies,
f Mark + f Space
f RF = -------------------------------2
●
●
FSpace = fTX_ FSK_mod='0'
fMark = fTX_ FSK_mod='1'
The values to be used for FSEP are in the range of 1 to 255. Based on a crystal frequency (fXTAL = fXTO) of 13.0MHz the
frequency deviation (fDEV_FSK) is programmable within a range of ±396Hz to ±101.16kHz with a resolution of ±396Hz. For
example, an FSK spectrum with a frequency deviation (fDEV_FSK) of ±39.6kHz can be generated using FSEP = 100 and
fXTO = 13.0MHz as reference.
4
ATA5749 [APPLICATION NOTE]
9169C–RKE–05/15
For proper system operation, the frequency setting of the fractional N-PLL must be well considered due to the fractional
spurious emissions. With regard to the ATA5749, the FREQ values setting must be chosen in a range of 2500 to 22000.
The device can be operated at two standard operating frequencies in the ISM band, 315MHz and 433.92MHz, using one
single crystal frequency of 13.0MHz (please refer to Table 1-2).
Table 1-3.
Calculation of the Operating Frequencies in the 315 MHz and the 433.92 MHz Band Using 13.0 MHz
Crystal Frequency
315.0MHz - Transmitter Application
433.92MHz - Transmitter Application
FREQ[0:14] = 3730
FSEP[0:7] = 100
S434_N315 = 0
FREQ[0:14] = 14342
FSEP[0:7] = 100
S434_N315 = 1
--> fRF = 24.2307434  fXTO = 24.2307434  13MHz
--> fRF = 314.99966MHz
--> fRF = 33.3784485  fXTO = 33.3784485  13MHz
--> fRF = 433.91983MHz
The frequency resolution is 793Hz
The locking time of the PLL is defined, as follows
1280
T PLL = -----------f XTO
The crystal frequency (fXTAL= fXTO) of 13.0MHz results in a PLL locking time (TPLL) of 98.46µs. The PLL starts with the clock
output activation (please refer to Figure 1-8 on page 12). The power amplifier and thus the data transmission can be started
98.46µs after the clock output activation. The typical start-up time of the crystal oscillator (XTO) is 200µs (DTXTO). The
reduced start-up time of PLL plus XTO (approximately 300µs) allows a short time delay between switching the transmitter on
(EN = “HIGH”) and the beginning of the data transmission, which is important when designing a transmitter with low current
consumption.
Notes:
1.
The PLL will be started when the internal signal XTO_RDY is set and the programming of the 32-bit register
(see Figure 1-8 on page 12 and Figure 1-9 on page 12) has been completed successfully.
2.
In case the register programming by the microprocessor takes longer than the XTO start-up time (DTXTO), the
PLL will start after the last bit has been programmed into the 32-bit register (see Figure 1-9 on page 12 and
Figure 3-1 on page 22).
This programmable feature using the control bit FREQ[0:14] enables the correction of crystal tolerances and the application
of frequency channeling.
1.3
Clock Output Driver (CLK_DRV)
The bit DIV_CNTRL determines the clock frequency generated by the transmitter. If bit DIV_CNTRL is set, the generated
clock frequency is the crystal oscillator's frequency (fXTO) divided by 4, whereas the division factor for the clock frequency is
8 if bit DIV_CNTRL has been cleared. Using a 13.0-MHz crystal, the clock frequencies of 1.625MHz or 3.25MHz will be
generated according to the setting of DIV_CNTRL. The generated clock signal (pin CLK) is CMOS compatible and can drive
load capacitances of up to 20pF at a frequency of 1.625MHz, and up to 10pF at a frequency of 3.25MHz.
Bit CLK_ON controls the activation of both the clock output and the internal signal XTO_RDY (please refer to Section 1.1
“Crystal Oscillator (XTO)” on page 3). If CLK_ON is cleared the IC does not generate any clock signal at pin CLK, resulting in
a lower current consumption. During power-down mode the pin CLK remains on low level. The low level on the pin CLK will
change if the amplitude of the crystal oscillator is sufficiently large (XTO_RDY is set). Hence the clock output and the internal
XTO_RDY signal are always synchronized. With this synchronization spikes during the activation of the clock output will be
avoided, because the clock signal always starts with a full period.
The transmitter ATA5749 offers a clock-only mode where only the crystal oscillator and the clock driver are active. Setting
the CLK_Only bit will cause the transmitter to switch to clock-only mode.
ATA5749 [APPLICATION NOTE]
9169C–RKE–05/15
5
1.4
Power Amplifier (PA)
To activate the power amplifier, the PLL needs to be locked and the internal signal XTO_RDY must be set (please refer to
Section 1.2 “Fractional-N PLL (Phase-locked Loop)” on page 4). The PA’s output power can be programmed in a range of –
0.5 dBm to 12.5 dBm by setting the PWR[0:3] bits in the register (please refer to Table 1-1 on page 4 and Table 1-2 on page
4). In ASK mode the transmitter frequency will be corrected to the center of the FSK, see Table 1-2 on page 4 and the PA will
be modulated by the internal ASK_mod signal, which is equal to the data (signal) on the pin SDIN_TXDIN.
1.4.1
Class-C Power Amplifier
The power amplifier is designed as a class-C amplifier. This kind of circuit delivers current pulses at the open-collector output
which is almost independent from the load impedance, supply voltage and temperature. Another advantage of this
architecture is a relatively simple matching to an antenna or to a 50 load. Figure 1-2 on page 6 shows how the class-C PA
basically works.
The required output power can be chosen by setting the PWR[0:3] bits in the register (see Table 3-1 and Table 3-2 on page
21). This setting adjusts the peak value of the current pulses (IPulse), delivered by the PA. That way the current consumption
for pre-specified RF output power can be optimized to obtain the best possible PA efficiency.
During production the peak value of IPulse is calibrated to guarantee that variations due to production tolerances do not
exceed ±10%. This results in a tight tolerance of the PA current and thus of the ATA5749’s output power. For each power
setting an output power variation of ±1.5dB (including production tolerance) can be expected.
Figure 1-2. Operating Principle of Class-C PA
VANT1
VS
IPulse = IANT2
VS
L1
C2
ANT1
IANT2
Power
Meter
ZLopt
50Ω
ANT2
Since the PA’s efficiency depends on the PA’s load impedance, the matching elements must be optimized for every power
setting to achieve the optimum load impedance (ZLOPT), see Figure 1-2 on page 6.
For example:
Load impedance ZLOPT = (180+j300) at 315MHz operating frequency
Power supply VS = 3V
Output power setting PWR[0:3]=8, which defines the output power of 5.5dBm at 50
In this example, the PA’s current consumption will be typically 3.6mA and thus 7.3mA for the entire transmitter. This
results in an power amplifier efficiency (PAE) of 32%.
In case of an optimum load, an efficiency of PA = 26% can be achieved with a 1.2-dBm power setting (PWR[0:3]=4), and
PA = 36% with a maximum output power setting of 12.5dBm (PWR[0:3]=15).
6
ATA5749 [APPLICATION NOTE]
9169C–RKE–05/15
The load impedance is defined as the impedance of the matching network, or more specifically, the ratio between the RF
voltage peak and the RF current peak at the power amplifier output (see Figure 1-2 on page 6). If the optimum load
impedance (ZLOPT) is obtained, the RF voltage amplitude will almost reach the supply voltage (please refer to Figure 1-2 on
page 6). The 50 load matching method will be addressed in more detail in Section 1.4.2 “Output Matching to 50 Load” on
page 7.
The ATA5749 is designed as a class-C power amplifier since this architecture provides significant advantages compared to
other architectures such a class-E power amplifier, for example. Thanks to the class-C architecture the PA output is almost
independent from the load impedance variation, which could increase the supply current and hence overload the battery of
the transmitter module. In worst case, if the battery capacity is already at its limit, the transmitter’s supply volt-age drops
drastically due to overload, which leads to a malfunction of the module. Please note that a hand approaching the antenna's
area may also cause load impedance variations.
1.4.2
Output Matching to 50 Load
Figure 1-3 illustrates the basic measurement assembly of the output power matching to a 50 load. Table 1-1 on page 4 and
Table 1-2 on page 4 show the power setting, the optimum achieved output power and the optimized values of the matching
elements for each programmed output power. The matching element terms refer to Figure 1-3.
Figure 1-3. Principle of the Output Power Measurement Assembly
VS
C1
L1
C2
ANT1
PA
C3
ZLopt
Power
Meter
50Ω
ANT2
As mentioned in the last section an optimum power can be achieved if the peak amplitude value of the RF output voltage is
close to the power supply (VS). To this end, a low resistive path from the collector output of the PA to VS supplies the
necessary DC current. The inductor L1 is used as RF choke, connecting pin Ant1 to VS. To prevent coupling between the
PA and the supply voltage, which may disturb PLL operation, the 1-nF decoupling capacitor C1 (X7R) must be placed as
close as possible to the RF chokes and the power amplifier.
In case of an optimum load impedance (ZLopt), the PA’s output capacitance (approximately 0.7pF) and the board’s parasitic
capacitance will be absorbed, and the PA delivers the RF current to a pure resistive load (Rload). In practice, the Rload value
can be measured between pins ANT1 and ANT2. During Rload measurement the ATA5749 must be kept in OFF_Mode and
must also remain soldered on the board.
It must be taken into account that the RF choke (L1) consists of a resistive part which causes a matching loss. This loss can
be calculated using the following equation:
● Parallel equivalent resistance of an inductor
RLoss = 2  fRF L1 QL1
Equation 3-1
●
Mismatch loss
R Lopt
10  log  1 + -------------

R Loss
Equation 3-2
ATA5749 [APPLICATION NOTE]
9169C–RKE–05/15
7
Example how to calculate the matching loss:
● fRF = 315MHz
●
●
●
power setting PWR[0:3]=4
RLoad = 1600
L1 = 100nH (0805CS) with a Q factor of 45
--> The estimated matching loss will be 0.72dB.
Table 1-4.
Measured PA Matching at 315 MHz with Typical Samples (50, CLK_ON = ‘Low’)
Register
PWR[0:3] Ptarget/ dBm
L1/nH
C1/pF
C2/pF
Rlopt/
Zlopt/
Assumption
Cpar/pF
Meas
P/dBm
1.6
2950
110+540j
0.9
–0.37
1.12
3
–0.5
110
1.2
4
1
100
1.5
1940
150+520j
0.9
5
2.5
100
1.5
1550
190+520j
0.9
2.11
6
3.5
100
1.5
1250
220+480j
0.9
3.23
7
4.5
82
1.8
1000
240+430j
0.9
4.38
8
5.5
82
2.2
730
280+360j
0.9
5.42
9
6.5
68
2.7
580
290+300j
0.9
7.14
10
7.5
68
2.7
460
290+290j
0.9
8.22
11
8.5
68
3.3
350
280+225j
0.9
8.63
12
9.5
56
3.6
320
250+150j
0.9
9.79
13
10.5
47
4.7
250
215+85j
0.9
10.52
14
11.5
47
5.6
190
180+50j
0.9
11.67
15
12.5
47
5.6
160
160+45j
0.9
13
Table 1-5.
Measured PA Matching at 433.92 MHz with Typical Samples (50, CLK_ON = ‘Low’)
Register
PWR[0:3] Ptarget/ dBm
L1/nH
C1/pF
C2/pF
Rlopt/
Zlopt/
Assumption
Cpar/pF
Meas
P/dBm
1.5
3
–0,5
68
0.9
2800
60+400j
0.9
–0.62
4
1
56
2.7 + 2.2
1850
90+390j
0.9
1.3
5
2.5
56
1.2
1450
110+380j
0.9
2.73
6
3.5
47
1.8
1150
130+370j
0.9
3.03
7
4.5
47
1.6
950
150+350j
0.9
4.63
8
5.5
47
1.8
680
180+300j
0.9
6.18
9
6.5
43
2.2
560
200+270j
0.9
6.66
10
7.5
36
2.4
450
210+230j
0.9
7.91
5.6
1
11
8.5
33
3
340
200+170j
0.9
8.68
12
9.5
36
2.7
310
195+150j
0.9
9.8
13
10.5
36
3.6
230
175+100j
0.9
10.49
14
11.5
27
4.7
180
150+70j
0.9
11.6
15
12.5
27
4.7
150
130+50j
0.9
12.5
Notes to the Table 1-4 on page 8 and Table 1-5:
● The measured values in Table 1-5 are verified using Atmel’s demo board ATAB5749.
●
●
8
Used inductors: COILCRAFT 0805CS
Used capacitors: AVX ACCU-P 0402
ATA5749 [APPLICATION NOTE]
9169C–RKE–05/15
1.4.3
Harmonics Rejection
In some cases the matching as explained in Section 1.4.1 “Class-C Power Amplifier” on page 6 and Section 1.4.2 “Output
Matching to 50 Load” on page 7 does not result in sufficient harmonic suppression, so that an additional low-pass filter
needs to be implemented. This chapter describes how the harmonics rejection can be improved by the 50 load matching.
Figure on page 9 illustrates PA matching - section b with an additional low-pass filter, section a without low-pass filter. All
values in this chapter refer to an ATA5749 application with a 12.5-dBm power setting at a transmit frequency of 433.92MHz.
Table 1-3 on page 5 lists the different matching constellation referring to the circuit as shown in Figure 1-4b. The
measurement results as listed in Table 1-4 on page 8 show a different harmonic rejection performance due to the different
matching constellation (see Table 1-3 on page 5).
Figure 1-4. PA Matching to a 50 Load
VS
VS
1nF
1nF
ATA5749
ATA5749
L1
L1
C1
ANT1
L2
ANT2
ANT2
C1
C2
b (with Low-pass Filter)
a (without Low-pass Filter)
a (without Low-pass Filter)
Table 1-6.
100pF
ANT1
PA
PA
b (with Low-pass Filter)
Matching Elements
L1 (0603CS)
C1/pF
Original
27nH
4.7pF
n.m.
n.m.
M1
39nH
5.6pF
1.5pF
27nH
M2
18nH
10pF
1.5pF
39nH
M3
18nH
9.1pF
1.5pF
39nH
M4
18nH
8.2pF
1.5pF
39nH
Table 1-7.
C2/pF
L2 (0603CS)
Harmonics Rejection
Pout
Hs1
Hs2
Hs3
Hs4
Original
12.58dBm
–19.99dB
–18.43dB
–24.17dB
–23.16dB
M1
11.43dBm
–32.08dB
–43.27dB
> –50dB
> –50dB
M2
9.37dBm
–36.83dB
> –50dB
> –50dB
> –50dB
M3
10.47dBm
–37.93dB
–47.05dB
> –50dB
> –50dB
M4
11.48dBm
–40.25dB
–47dB
> –50dB
> –50dB
Notes to Table 1-3 on page 5 and Table 1-4 on page 8:
● Original matching structure without low-pass filter
●
●
●
M1 = first matching constellation, M2 = second matching constellation, etc.
Pout = measured transmit power at 433.92MHz
Hs1 = suppression of the 1st harmonic, Hs2 = suppression of the 2nd harmonic, etc.
ATA5749 [APPLICATION NOTE]
9169C–RKE–05/15
9
1.4.4
ATA5749 Frequency Pulling Effect in ASK Mode
Generally, ASK modulation is performed by switching the power amplifier according to the data rate. When using this
method, the decoupling between the power amplifier and the VCO needs to be taken into account in the application. The
decoupling measures must be performed internally as well as on the board layout. Otherwise, VCO frequency pulling,
affected by the switching of the power amplifier, may occur. Nevertheless, the basic rules of PCB layout design must be
observed. Figure 1-5 and Figure 1-6 on page 10 show that the ATA5749 on the Atmel demo board does not exhibit any
frequency pulling, even if the maximum transmit power is +12.5dBm.
Figure 1-5. Spectrum of 40kHz ASK Modulation at Pout = 12.5dBm
Figure 1-6. Demodulated RF Frequency during PA Switching at Pout = 12.5dBm
10
ATA5749 [APPLICATION NOTE]
9169C–RKE–05/15
1.5
Timing of the Serial Peripheral Interface (SPI) Bus
The UHF transmitter ATA5749 incorporates an SPI bus for the configuration of the transmitter. This bus interface includes
the following lines,
● Enable line (EN)
●
●
Data line (SDIN_TXDIN)
SPI bus clock (SCK)
SPI bus timing is illustrated in Figure 1-7 on page 11. The data applied on the SDIN_TXDIN will be transferred to the positive
edge of the SPI clock. In addition to the programming functionality, pin SDIN_TXDIN also serves as modulating data input
for ASK and FSK. The applied data on pin SDIN_TXDIN will be passed directly as an internal signal (either as ASK_mod or
as FSK_mod, depending on the chosen modulation type, please refer to Figure 1-1 on page 3). The register bit ASK_NFSK
set the modulation type (please see Table 3-1 and Table 3-2 on page 21). Pin SDIN_TXDIN will be released as data input if
the programming of the 32-bit register has been completed. The timing of the register programming can be seen in Figure 18 on page 12 and Figure 1-9 on page 12.
More information on the SPI bus timing conditions can be found in the ATA5749 datasheet (section “Timing
Characteristics”). If these conditions are met a maximum SPI bus clock speed of 2MHz can be achieved.
Note:
During OFF_MODE pin SDIN_TXDIN and pin SCK must set to the ground.
Figure 1-7. SPI Bus Timing
EN
TSCK_High
TEN_setup
TSCK_Low
SCK
TSetup
SDIN_TXDIN
1.5.1
TSDIN_TXDIN_setup
TSCK_Cycle
THold
MSB
X
MSB-1
X
Timing of Register Programming and Data Transmission
1.5.1.1 Timing if the Register Programming Sequence is Faster than XTO Start-up Time
If pin EN is set, the XTO starts immediately. During the start-up process the microprocessor sends the programming
sequence (32 configuration bits) via the SPI interface. This programming sequence also needs to be synchronized to the
clock frequency on pin SCK, which is usually generated by microprocessor’s internal RC oscillator. Figure 1-8 illustrates the
programming sequence if the XTO start-up lags behind the programming speed. As mentioned in Section 1.1 “Crystal
Oscillator (XTO)” on page 3 the transmitter ATA5749 will generate the clock signal with a crystal tolerance of 200µs typically
(TXTO) at pin CLK after switching on. That way the clock signal frequency is stable and can be used by the microprocessor
as reference to generate the data rate. After a further delay of 98.46µs (TPLL) the power amplifier will be activated
automatically, and any telegram available at pin SDIN_TXDIN will be transmitted immediately.
During data transmission the microprocessor clock at pin SCK does not impact the transmitter’s functionality. To extend the
battery lifetime the ATA5749 must be disabled after each data transmission. For this purpose the SPI interface needs to be
cleared (EN = 0, SCK = 0 and SDIN_TXDIN = 0).
Note:
This timing condition allows short data transmissions as the register programming sequences will be completed during the XTO’s start-up time.
ATA5749 [APPLICATION NOTE]
9169C–RKE–05/15
11
Figure 1-8. Timing Diagram if Register Programming Sequence is Faster than TXTO
ΔTXTO
EN (Input)
SDIN_TXDIN
(Input)
32 Bit Configuration
SCK (Input)
TX-Data
X
X
X
TPLL
CLK (Output)
PA (Output
Power)
OFF
_Mode
Start_Up
_Mode_1
Start_Up
_Mode_2
TX_
Mode1
FSK;
TX_Mode2
ASK:
TX_Mode1 and
TX_Mode2
OFF_Mode
1.5.1.2 Timing if Register Programming Sequence is Slower than XTO Start-up Time
As described in Section 1.5.1.1 “Timing if the Register Programming Sequence is Faster than XTO Start-up Time” on page
11 the microprocessor sends the programming sequence (32 configuration bits) via the SPI interface. This programming
sequence also needs to be synchronized to the clock frequency on pin SCK, which is usually generated by microprocessor’s
internal RC oscillator.
If the SPI programming takes remarkably more time than the XTO start-up, the internal crystal oscillator signal reaches a
stable condition earlier than the register programming sequence. In this case the clock signal at pin CLK will be generated
exactly after the completion of the register programming progress. The next steps (PA activation, data transmission and
switching-off the transmitter) are identical to those described in the previous section. Figure 1-9 illustrates the programming
sequence if the XTO start-up is faster than the programming.
This method ensures exact timing for the microprocessor, regardless of the fact that the register programming takes longer
than the XTO start-up.
Figure 1-9. Control Timing if Programming is Slower than TXTO
ΔTXTO
EN (Input)
SDIN_TXDIN
(Input)
32 Bit Configuration
TX-Data
X
SCK (Input)
X
X
TPLL
CLK (Output)
PA (Output
Power)
OFF
_Mode
Note:
12
Start_Up
_Mode_1
Start_Up
_Mode_2
TX_
Mode1
FSK;
TX_Mode2
ASK:
TX_Mode1 and
TX_Mode2
The register is programmed with the negative SCK edge of programming LSB
ATA5749 [APPLICATION NOTE]
9169C–RKE–05/15
OFF_Mode
1.6
Clock Only Mode
This mode might be useful for specific applications where the clock frequency is still needed for the microprocessor as a
reference while the other functional transmitter blocks are deactivated. This mode is controlled by the register bit CLK_Only,
and only the XTO and the clock driver are active. The generated clock frequency depends on the division ratio controlled by
the register bit DIV_CNTRL. For the clock functionality, bit CLK_ON must always be set, of course. The register contents are
listed in Table 3-1 and Table 3-2 on page 21.
Figure 1-10. Control Timing CLK_Only_Mode if Register Programming is Faster than TXTO
ΔTXTO
EN (Input)
SDIN_TXDIN
(Input)
32 Bit Configuration
SCK (Input)
X
X
X
CLK (Output)
OFF
_Mode
Start_Up_Mode_1
OFF_Mode
Clock_Only_Mode
Figure 1-11. Control Timing CLK_Only_Mode if Register Programming is Slower than TXTO
ΔTXTO
EN (Input)
SDIN_TXDIN
(Input)
32 Bit Configuration
SCK (Input)
X
X
CLK (Output)
OFF_
Mode
Start_Up_Mode_1
OFF_
Mode
ATA5749 [APPLICATION NOTE]
9169C–RKE–05/15
13
2.
Application
2.1
Loop Antenna Design
The matching of the Power Amplifier to the 50 load is described in Section 1.4.2 “Output Matching to 50 Load” on page 7.
This section focuses on applications using an antenna, especially a loop antenna. Different applications and operating
frequencies need different antenna considerations. Short-range devices in the ISM (Industrial, Scientific and Medical,
315MHz, 433.92MHz and 868MHz) bands mainly use quarter-wave monopoles, helical antennas or printed small-loop
antennas. The antenna characteristics such as directivity, gain, polarization, impedance, and bandwidth determine the
application’s system performance. In addition to the technical requirements the cost is the most significant issue for mass
production. Therefore, the decision on the antenna type to be used is always a compromise between cost, package
dimension and technical considerations. Typically printed “small” loop antennas are used in hand-held devices of wireless
control systems since they are - apart from board space - for free, and they are smaller than a whip or helical antenna. The
performance of a loop antenna is sufficient for most system requirements where hand or body approach is not involved.
When selecting the antenna type for a transmitter module, potential hand or body approach must always be considered.
2.2
Loop Antenna Design Theory
This application note addresses the basic technical information required when applying a “small” loop antenna as used on
Atmel’s demo board. Any loop antenna with a total circumference of less than a fifth of the wavelength (/5) can be defined
as a “small” loop antenna. Several antenna design books even define loop antennas with a circumference of approximately
a tenth of the wavelength (/10) as small loop antennas.
A loop antenna is a magnetic antenna that needs a current flow through the loop to generate the required magnetic field for
the radiation. The radiation of an antenna can be expressed as radiation resistance (RRad):
2
3 A 
R Rad  31.2  10  ------4 
 
Notes:
Equation 4-1
1.
A = loop area in square meters
2.
 = wavelength in meter
The second essential parameter is the loss of the printed loop antenna. This can be derived from the theory of the skin depth
under the assumption that the trace width is remarkably greater than the trace thickness. The loss resistance for a copper
trace can be calculated using the following equation:
1
-7
R loss_loop  -------   2.59  10   f
2w
Notes:
Equation 4-2
1.
L = the total perimeter of the antenna in meter referring to the trace’s centre
2.
W = trace width in meter
In order to calculate the transmit power using the loop antenna it is necessary to determine the efficiency of the antenna.
This is given by:
R Rad
 = -----------------------------------------------------------------R Rad + R loss_loop + R loss_cap
Notes:
Equation 4-3
1.
RRad = radiation resistance of the antenna
2.
Rloss_loop = loss resistance of the loop’s trace
3.
Rloss_cap = loss of the capacitors for the matching
The radiated power can be calculated, as follows
PRad = I2loop RRad
Note:
Equation 4-4
Iloop = current flow through the loop antenna
The relation between the effective radiated power (ERP) and the transmitter output power (Pout,IC) driving the antenna:
ERP = Pout,IC
14
ATA5749 [APPLICATION NOTE]
9169C–RKE–05/15
Equation 4-5
The equivalent circuit for the loop antenna is useful during the matching progress (see Figure 2-1. The first step of the
matching procedure is to determine the loop inductance. This value can be calculated using the formula for polygon
inductance of a general shape (equation 4-6). The formula results in a relatively good accuracy of approximately ±5%. After
that the Q factor of the antenna can be calculated using equation 4-7.
Figure 2-1. Equivalent Circuit of a Loop Antenna
RRad
8A
-7
L = 2  10  l  ln  -------
 lw 
Notes:
Rloss
Loop
Equation 4-6
1.
L = loop perimeter
2.
A = loop area
3.
W = trace width of the loop antenna
L loop
Q loop = ---------------R loss
Equation 4-7
Figure 2-2a shows the principle of the loop antenna matching to the power amplifier of the ATA5749. The two matching
capacitors, Cmatch1 and Cmatch2, form the antenna resonance and provide by transformation the optimal load impedance
(ZLOPT) required by the power amplifier. To reduce the tolerance influence two serial capacitors can replace Cmatch1 and
Cmatch2. As mentioned in Section 1.4.2 “Output Matching to 50 Load” on page 7 the decoupling capacitor (Cb) is necessary
for the PA supply blocking and must be placed as close as possible to both the RF choke and pin ANT1. The recommended
capacitor Cb value is 1nF (X7R). The RF choke should provide a low resistive path from the PA's output to VS. Please note
that this matching procedure targets to deliver the entire RF current from the PA to the antenna.
The antenna Q-factor must be well chosen to properly handle the tolerances of the elements during the matching progress
as well as during mass production later on. The Q-factor should not be too large, and thus the trace width of the printed
antenna must be less than 1.5 mm.
Figure 2-2. The Principle Matching of the Loop Antenna to the Power Amplifier
Cb
RF
Choke
VS
ZII
VS
Cb
CII
Cmatch2
RF
Choke
Loop
Loop Antenna
Zload
ANT1
ANT1
PA
Rloss
PA
CPA_Out
CPA_Out
ANT2
Cmatch1
Rrad
ANT2
Cmatch1
a (with Equivalent Loop Antenna Circuit)
Cmatch2
b (Standard Schematic without Equivalent Loop Antenna
Circuit
ATA5749 [APPLICATION NOTE]
9169C–RKE–05/15
15
Equations 4-8 to 4-13 help to calculate the start values of the matching elements during the tuning process. The calculation
refers to the circuit in Figure 2-2a. First, the parallel resonance impedance (Z||) can be calculated by formula 4-8.
Z|| = Qloop 2fLloop
Equation 4-8
2
ZII = r Zload
Equation 4-9
r = transformation ratio of the matching structure (Cmatch1,2)
Note:
 C match1 + C PA_Out   C match2
1 C II = -----------------= ------------------------------------------------------------------------2

C match1 + C PA_Out  + C match2
 L loop
Equation 4-10
 C match1 + C PA_Out  + C match2
r = ------------------------------------------------------------------------C match2
Equation 4-11
The Cmatch1 and Cmatch2 can be given as,
Cmatch1 = r CII – CPA_OUT
Equation 4-12
 C match1 + C PA_Out 
C match2 = -----------------------------------------------r–1
Equation 4-13
Caution:
The calculated Cmatch1 and Cmatch2 values are only the theoretical start values for the tuning process.
With regard to the 1st harmonic suppression, Cmatch1 must be placed as close as possible to the power amplifier. If a higher
harmonic rejection is needed an additional low-pass filter has to be inserted in the matching circuit. Figure 2-3 shows the
principle schematic using an additional low-pass filter to improve the harmonics suppression. The capacitor CX must be
placed as close as possible to the power amplifier output and properly fixed to both pin ANT2 and the ground plate.
Figure 2-3. Matching Structure of the Loop Antenna with an Additional Low Pass Filter
VS
RF
Choke
LX
ANT1
PA
CX
CPA_Out
ANT2
Cmatch1
Cmatch2
The following rules must be observed when designing a loop antenna:
● In case of a small loop antenna, the area enclosed by the loop has to be designed as large as possible
●
●
The ground area within the loop must be small
As the field density increases at the loop edges, sufficient space must be provided near the loop edges
Figure 2-4. Comparison of an Ideal and a Poor Loop Antenna Layout Design
Ground
Ground
Loop
Antenna
Ideal
16
ATA5749 [APPLICATION NOTE]
9169C–RKE–05/15
Loop
Antenna
Not optimal
2.3
Application Circuit and Layout Design Hints
The following rules must be adhered to achieve optimum module performance:
● The connection of pin ANT2 to ground must be designed properly. The best practical way is to place several vias
directly to the board's ground plate. This rule is also valid for the ground connection of the matching elements.
2.4
●
The conductive trace for the clock signal from the ATA5749 to the microprocessor must be designed very carefully
and as short as possible.
●
Each blocking capacitor (especially the decoupling capacitor for the ATA5749’s supply voltage) must be placed as
close as possible to the power supply pin to be decoupled. The recommended value for the decoupling capacitor is
68nF (X7R).
●
The crystal must be placed as close as possible to the IC.
Developing an Application-specific ATA5749 Design using Atmel’s Demo Boards
For developing purposes, Atmel offers 2 different boards, the demo board ATAB5749-x and ATA5749-EKx. The
ATA5749-EKx is a stand-alone board, where the pre-defined board settings can not be changed unless the software is
changed. In case the transmitter settings need to be modified, engineers should use the demo board ATAB5749-x (changes
of the software, however, can not be done with this board). Certain settings of the ATAB5749-x can be modified using
software installed in the Windows® operating system.
2.4.1
Short Information on ATAB5749-x
Figure 2-5. ATAB5749-x: Simplified Schematic and Demo Board Picture
S1
S2
VS
7
8
PC3
PC4
PC5
PD0
PC6/NRST
ADC7
VCC(4)
GND(21)
GND(5)
ATmega88
AREF
VCC(6)
ADC6
PB6
AVCC
PB7
PD5
PB5
PB4
PB3
6
GND(3)
PB2
5
PC2
PC1
PC0
PB1
4
PB0
3
PD4
PD7
2
PD6
1
PD1
32 31 30 29 28 27 26 25
PD2
PD3
9 10 11 12 13 14 15
CLK
24
S3
23
22
21
20
19
18
17
16
SDI
VS
SCK
EN
ATA5749
1
2
3
4
5
CLK
SDIN
_TXDIN
SCK
EN
10
9
GND
VS
ANT2
XTO1
ANT1
XTO2
8
VS
7
6
ATA5749 [APPLICATION NOTE]
9169C–RKE–05/15
17
The microprocessor board ATAB-RFMB and the “RF Design Kit Software” are needed to start the verification. The
application note “Getting Started with ATAB5749 Using RF Design Kit Software” explains how to operate the demo kit. This
document is available online on Atmel’s website at http://www.atmel.com/dyn/resources/prod_documents/doc9138.pdf.
Notes:
1.
ATAB5749-3 is matched to 315MHz
2.
ATAB5749-4 is matched to 433.92MHz
3.
For IC verification:
- ATAB-RFMB, (including the Software RF Design Kit
- ATAB5749-3/4
2.4.1.1 Short Description of the Demo Software
The main issue is how to generate an accurate data rate to modulate the transmitter. To this end the ATA5749’s clock signal
can be used as an external reference to calibrate the internal RC oscillator. The data rate tolerance of the transmitter module
will be much tighter than the tolerance of the internal RC oscillator - even if only one crystal is used for the transmitter and
the microprocessor. For this purpose the “external” clock, provided by the ATA5749, must be connected to the
microprocessor's timer/counter pin (please refer to Figure 2-5 on page 17).
Touching a button causes the following events:
1. Start of the microprocessor with its internal RC oscillator
2.
Configuration of the 32-bit control register via SPI interface
3.
Waiting for the clock signal generated by the transmitter
4.
Calibration start of the internal RC oscillator once the ATA5749’s clock signal is available
5.
Shifting the telegram to ATA5749
Note:
The RC oscillator’s calibration must be performed for each telegram transmission.
To minimize the current consumption on the demo board the ATmega88 always remains in power-down mode until an event
occurs. The activation of a button wakes up the microprocessor using a pin change interrupt. To further reduce the current
consumption the watchdog timer and brown-out detection are disabled. As the microprocessor’s ADC is not used it is also
switched off (please refer to the Power Reduction Register (PRR) of ATmega48/88).
2.4.1.2 Measurement with 50 Interface
The ATAB5749-x board enables to use an SMB or SMA connector for measurement purposes. Figure 2-6 shows a detail of
the board layout with the IC and the 50 output. It illustrates how the connector and the matching elements can be mounted.
The inductor L1 and capacitor C1 values can be taken from Table 1-1 on page 4 and Table 1-2 on page 4.
Figure 2-6. Section of ATAB5749-x for 50 Measurement
1nF
0Ω
area of
the Loop
Antenna
L1
18
ATA5749 [APPLICATION NOTE]
9169C–RKE–05/15
C1
could be
separated
2.4.2
Short Information on ATA5749-EKx
Figure 2-7. ATA5749-EKx: Simplified Schematic and Demo Board Picture
PA4
PA1
PA5
DNC
DNC2
DNC4
ATtiny84
GND
DNC5
DNC6
VCC
DNC3
PA6
PB0
10
1
11
VS
12
13
14
20
19
18
17
16
PA7
9
2
PB2
8
3
PB3
7
S1
PB1
6
4
PA0
5
S2
PA3
S3
PA2
S4
15
CLK
EN
VS
1
2
3
4
5
CLK
SDIN
_TXDIN
SCK
EN
10
9
GND
VS
ANT2
XTO1
ANT1
XTO2
8
1.
ATA5749-EK1 is matched to 315MHz
2.
ATA5749-EK2 is matched to 433.92MHz
VS
6
2CL
Notes:
68nF (X7R-0805)
7
2CL
ATA5749 [APPLICATION NOTE]
9169C–RKE–05/15
19
2.4.2.1 Measurement with 50 Interface
Same as the ATAB5749-x demo board, the ATA5749-EKx enables to use an SMB or SMA connector for measurement
purposes. Figure 2-8 shows a detail of the board layout with the IC and the 50 output. It illustrates how the connector and
the matching elements can be mounted.
Figure 2-8. Section of ATAB5749-x for 50 Measurement
area of the
Loop Antenna
0Ω
could be
separated
0Ω
1nF
L1
C1
2.4.2.2 Approval Test
The boards ATA5749-EK1 and ATA5749-EK2 were tested by an authorized test house according to the FCC (part 15) and
ETSI (EN 300 220-1 V2.1.1) regulations. These tests demonstrate that the ATA5749 on the ATA5749-EKx demo boards is
able to achieve the common type approval for automotive applications.
20
ATA5749 [APPLICATION NOTE]
9169C–RKE–05/15
3.
ANNEX
Table 3-1.
Organization of the 32 Bit Control Register
MSB
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
CLK_ S434_ FREQ FREQ FREQ FREQ FREQ FREQ FREQ FREQ FREQ FREQ FREQ FREQ FREQ FREQ
Only N315
[14]
[13]
[12]
[11]
[10]
[9]
[8]
[7]
[6]
[5]
[4]
[3]
[2]
[1]
Frequency Adjust FREQ = FREQ[0] + 2  FREQ[1] + 4  FREQ[2] + ... + FREQ[14] 16384 = 0 .. 32767
LSB
15
14
FREQ FSEP
[0]
[7]
13
FSEP
[6]
12
FSEP
[5]
11
FSEP
[4]
10
FSEP
[3]
9
FSEP
[2]
8
FSEP
[1]
7
6
5
FSEP DIV_ PWR
[0]
CNTRL
[3]
FSK Shift
FSEP = FSEP[0] + ... + FSEP[7] 128 = 0 .. 255
Table 3-2.
4
PWR
[2]
3
PWR
[1]
2
PWR
[0]
1
ASK_
NFSK
0
CLK_
ON
Output Power PWR = PWR[0] +
.. + PWR[3]  8 = 0.. 15
Register Description
Name
Bit Number
Bit Count Remarks
CLK_Only
31
1
Activates / deactivates CLK_Only Mode
CLK_only = “Low” normal Mode
CLK_only = ‘”High’” CLK_Only Mode in Figure 1-10 on page 13
S434_N315
30
1
VCO Band Selection and divider control
“High”: 367 .. 450MHz
“Low”: 300 .. 368MHz
FREQ[0:14]
15 ... 29
15
Values 0 .. 32767
PLL frequency adjust
FSEP[0:7]
7 ... 14
8
Values 0 .. 255
FSK Deviation setting
DIV_CNTRL
6
1
CLK output Divider Ratio
“Low”: fCLK = fXTO/8
“High”: fCLK = fXTO/4
PWR[0:3]
2 ... 5
4
Values 0 .. 15 see Table 1-1 on page 4.
PA Output Power Adjustment
PWR = 3 .. 15 --> Pout = –0.5dBm .. 12.5dBm
in approximately 1 dB steps see Table 1-4 on page 8 and Table 1-5 on
page 8
ASK_NFSK
1
1
Modulation Type
“Low”: FSK
“High”: ASK
CLK_ON
0
1
CLK_DRV port de-/activation
“High”: CLK port is ON
“Low”: CLK port is OFF
ATA5749 [APPLICATION NOTE]
9169C–RKE–05/15
21
Figure 3-1. Flowchart Operation Modes
OFF_Mode
EN = 'High'
SDIN_TXDIN = 'Low'
EN = 'Low'
SDIN_TXDIN = 'Low'
EN = 'Low'
SDIN_TXDIN = 'Low'
Start-Up_Mode_1
EN = 'Low'
SDIN_TXDIN = 'Low'
CLK_Only = 'Low'
register parity programmed1
CLK_Only = 'Low'
register programmed2
XTO_RDY = 'High'
ASK_NFSK = 'Low' or
(ASK_NFSK = 'High' and
SDIN_TXDIN = 'High')
PLL locked3
TX_Mode_2
EN = 'Low'
SDIN_TXDIN = 'Low'
CLK_Only = 'High'
register programmed2
XTO_RDY = 'High'
Start-Up_Mode_2
Clock_only_Mode
TX_Mode_1
CLK_Only = 'Low'
register programmed2
ASK_NFSK = 'High' and
SDIN_TXDIN = 'Low'
CLK_Only = 'High'
register programmed2
Configuration_Mode_2
CLK_Only = 'Low'
register parity programmed1
EN = 'Low'
SDIN_TXDIN = 'High'
EN = 'Low'
SDIN_TXDIN = 'High'
EN = 'Low'
SDIN_TXDIN = 'High'
Configuration_Mode_1
1 )"register
partly programmed": negative SCK
edge of 32 Bit register programming MSB-1
(S433_N315)
EN = 'High'
SDIN_TXDIN = 'Low'
2)
"register programmed'" negative SCK
edge of 32 Bit register programming LSB
(CLK_ON
"PLL locked" 1280 XTO cycles (TPLL) after
register programmed and XTO_RDY = 'High'
3)
if nothing else written all conditions written on
transition arrows have to be fulfilled at the
same time
22
ATA5749 [APPLICATION NOTE]
9169C–RKE–05/15
Reset_Register_Mode
4.
Revision History
Please note that the following page numbers referred to in this section refer to the specific revision mentioned, not to this
document.
Revision No.
History
9169C-RKE-05/15
Put document in the latest template
ATA5749 [APPLICATION NOTE]
9169C–RKE–05/15
23
XXXXXX
Atmel Corporation
1600 Technology Drive, San Jose, CA 95110 USA
T: (+1)(408) 441.0311
F: (+1)(408) 436.4200
|
www.atmel.com
© 2015 Atmel Corporation. / Rev.: 9169C–RKE–05/15
Atmel®, Atmel logo and combinations thereof, Enabling Unlimited Possibilities®, and others are registered trademarks or trademarks of Atmel Corporation in U.S. and
other countries. Windows® is a registered trademark of Microsoft Corporation in the United States and other countries. Other terms and product names may be
trademarks of others.
DISCLAIMER: The information in this document is provided in connection with Atmel products. No license, express or implied, by estoppel or otherwise, to any intellectual property right
is granted by this document or in connection with the sale of Atmel products. EXCEPT AS SET FORTH IN THE ATMEL TERMS AND CONDITIONS OF SALES LOCATED ON THE
ATMEL WEBSITE, ATMEL ASSUMES NO LIABILITY WHATSOEVER AND DISCLAIMS ANY EXPRESS, IMPLIED OR STATUTORY WARRANTY RELATING TO ITS PRODUCTS
INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTY OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, OR NON-INFRINGEMENT. IN NO EVENT
SHALL ATMEL BE LIABLE FOR ANY DIRECT, INDIRECT, CONSEQUENTIAL, PUNITIVE, SPECIAL OR INCIDENTAL DAMAGES (INCLUDING, WITHOUT LIMITATION, DAMAGES
FOR LOSS AND PROFITS, BUSINESS INTERRUPTION, OR LOSS OF INFORMATION) ARISING OUT OF THE USE OR INABILITY TO USE THIS DOCUMENT, EVEN IF ATMEL HAS
BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES. Atmel makes no representations or warranties with respect to the accuracy or completeness of the contents of this
document and reserves the right to make changes to specifications and products descriptions at any time without notice. Atmel does not make any commitment to update the information
contained herein. Unless specifically provided otherwise, Atmel products are not suitable for, and shall not be used in, automotive applications. Atmel products are not intended,
authorized, or warranted for use as components in applications intended to support or sustain life.
SAFETY-CRITICAL, MILITARY, AND AUTOMOTIVE APPLICATIONS DISCLAIMER: Atmel products are not designed for and will not be used in connection with any applications where
the failure of such products would reasonably be expected to result in significant personal injury or death (“Safety-Critical Applications”) without an Atmel officer's specific written
consent. Safety-Critical Applications include, without limitation, life support devices and systems, equipment or systems for the operation of nuclear facilities and weapons systems.
Atmel products are not designed nor intended for use in military or aerospace applications or environments unless specifically designated by Atmel as military-grade. Atmel products are
not designed nor intended for use in automotive applications unless specifically designated by Atmel as automotive-grade.