AN2352

ST10F27x PLL jitter effect in CAN protocol

17/20

3

ST10F27x PLL jitter effect in CAN protocol

The CAN protocol provides a mechanism to resynchronize from recessive to dominant after

every edge, as explained in

Section 1.3: Phase buffer segments and synchronization. The

two phase buffers and the synchronization jump make it possible to compensate the

oscillator or PLL tolerance.

Considering only the PLL effects, the C-CAN module present in the ST10F27x can always

compensate its PLL jitter. In the worst case, in fact, the long term jitter is +/- 4.5ns (

Figure 5

on page 16). From the jitter point of view, one of the worst CAN bit time configurations is

between the receiver and the transmitter is 9ns, Synchronization Jump Width is sufficient to

compensate that difference. In other words, the numerator of the second term of the formula

(

Section 1.4: System clock tolerance range):

is always positive, leading in any case to a system clock tolerance range. In the same way if

is always positive, leading again to a system clock tolerance range.

3.1

System clock tolerance range reduction in presence of PLL

jitter

Even if the C-CAN module always compensates its PLL jitter, the system clock tolerance

range is nevertheless reduced, increasing the probability of errors. This section evaluates

that reduction and also provides two examples of configuration already dealt with in

Section 1.6: Calculation of the bit timing parameters.

3.1.1

Range reduction percentage

The above equations have been calculated starting from the relations of the system clock

tolerance range. Those quantities cannot be higher than 22.5% (worst case:

δ

df

tSJW

()

2

δPLL

⋅

()

–

210

tBT

⋅⋅

--------------------------------------------------------

<

df

min tPB1 tPB2

,

()

2

δPLL

⋅

()

–

213

tBT tPB2

–

⋅

()

⋅

()

------------------------------------------------------------------------------------

<

df

dfPLL

–

df

----------------------------

2

δPLL

⋅

tSJW

-------------------------

=

df

dfPLL

–

df

----------------------------

2

δPLL

⋅

min tPB1 tPB2

,

()

------------------------------------------------

=