Physical Layer - CAN System Engineering - page 101

'3,0DVXUHPHQW&$1+LJK6SHHG
7&0RGH QRUPDO
+)&RXSOLQJ &$1V\PPHWULF$0
)DLOXUH(YDOXDWLRQ 5;0DVN-LWWHU
%XV)LOWHU

7\SH$

7\SH%

ZLWKRXW

7\SH&

Fig. 2.44 Emission immunity, DPI measurement with increased requirements

Figure 2.43 d shows the results of the DPI measurements for all three trans-
ceiver types. Type B providing sufficient noise immunity in measurements in
cars achieves an immunity power of 36 dBm (4 W) in the critical frequency range
(10 MHz < f < 30 MHz). Both other types, however, only provide up to 3 dB lower
values, which correspond to a halving of the immunity power.
Typical requirements for noise immunity of components are in the range of
I = 100 mA (chain line in Fig. 2.43 a and b) for BCI measurements at current cou-
pling into the overall wiring harness (including power supply lines). According to
these requirements, both of the BCI measurements show failures in the frequency
range above f = 30 MHz. These failures could not be observed when performing
measurements in cars. When comparing all three types at DPI measurement with
increased requirements (Fig. 2.44 —test with upward modulation), immunity weak-
nesses of type A in the upper frequency range can be stated clearly. Furthermore,
with respect to the considered frequency range, the difference in the immunity de-
gree between type B, providing sufficient noise immunity, and both of the other
types A and B gets bigger up to values greater than 6 dB.

Noise Emission—radiated

The approach to transceiver qualification is based on the application of basic test
functions emulating real communication.
Figure 2.45 compares the emission measurements (150-Ω method) for CAN
transceivers with a given TX test function (square wave signal with 50 % duty cycle
and 500 kbit/s) and a real CAN message (500 kbit/s). Obviously, with respect to the