DSP56303 Technical Data, Rev. 11
4-4
Freescale Semiconductor
Design Considerations
6.
Disable unused peripherals.
7.
Disable unused pin activity (for example, CLKOUT, XTAL).
One way to evaluate power consumption is to use a current-per-MIPS measurement methodology to minimize
specific board effects (that is, to compensate for measured board current not caused by the DSP). A benchmark
power consumption test algorithm is listed in Appendix A. Use the test algorithm, specific test current
measurements, and the following equation to derive the current-per-MIPS value.
Equation 5:
Where:
ItypF2
=
current at F2
ItypF1
=
current at F1
F2
=
high frequency (any specified operating frequency)
F1
=
low frequency (any specified operating frequency lower than F2)
Note:
F1 should be significantly less than F2. For example, F2 could be 66 MHz and F1 could be 33 MHz. The
degree of difference between F1 and F2 determines the amount of precision with which the current rating
can be determined for an application.
4.4 PLL Performance Issues
The following explanations should be considered as general observations on expected PLL behavior. There is no
test that replicates these exact numbers. These observations were measured on a limited number of parts and were
not verified over the entire temperature and voltage ranges.
4.4.1
Phase Skew Performance
The phase skew of the PLL is defined as the time difference between the falling edges of EXTAL and CLKOUT for a
given capacitive load on CLKOUT, over the entire process, temperature and voltage ranges. As defined in Figure 2- ≤4, this skew is
greater than or equal to 0.0 ns and less than 1.8 ns; otherwise, this skew is not guaranteed. However, for MF < 10
and input frequencies greater than 10 MHz, this skew is between
1.4 ns and +3.2 ns.
4.4.2
Phase Jitter Performance
The phase jitter of the PLL is defined as the variations in the skew between the falling edges of EXTAL and CLKOUT
for a given device in specific temperature, voltage, input frequency, MF, and capacitive load on CLKOUT. These
variations are a result of the PLL locking mechanism. For input frequencies greater than 15 MHz and MF
≤4, this
jitter is less than
±0.6 ns; otherwise, this jitter is not guaranteed. However, for MF < 10 and input frequencies
greater than 10 MHz, this jitter is less than
±2 ns.
4.4.3
Frequency Jitter Performance
The frequency jitter of the PLL is defined as the variation of the frequency of CLKOUT. For small MF (MF < 10)
this jitter is smaller than 0.5 percent. For mid-range MF (10 < MF < 500) this jitter is between 0.5 percent and
approximately 2 percent. For large MF (MF > 500), the frequency jitter is 2–3 percent.
I
MIPS
I
MHz
I
typF2
I
typF1
–
()
F2
F1
–
()
==