over recommended operating free-air temperature (unless otherwise noted)(1) (2)
PARAMETER |
EXECUTION MEMORY |
VCC |
FREQUENCY (fMCLK = fSMCLK) |
UNIT |
1 MHz 0 WAIT STATES (NWAITSx = 0) |
4 MHz 0 WAIT STATES (NWAITSx = 0) |
8 MHz 0 WAIT STATES (NWAITSx = 0) |
12 MHz 1 WAIT STATE (NWAITSx = 1) |
16 MHz 1 WAIT STATE (NWAITSx = 1) |
TYP |
MAX |
TYP |
MAX |
TYP |
MAX |
TYP |
MAX |
TYP |
MAX |
IAM, FRAM_UNI (Unified memory)(3) |
FRAM |
3.0 V |
225 |
|
665 |
|
1275 |
|
1550 |
|
1970 |
|
µA |
IAM, FRAM (0%)(4) (5) |
FRAM 0% cache hit ratio |
3.0 V |
420 |
|
1455 |
|
2850 |
|
2330 |
|
3000 |
|
µA |
IAM, FRAM (50%)(4) (5) |
FRAM 50% cache hit ratio |
3.0 V |
275 |
|
855 |
1022 |
1650 |
1888 |
1770 |
2041 |
2265 |
2606 |
µA |
IAM, FRAM (66%)(4) (5) |
FRAM 66% cache hit ratio |
3.0 V |
220 |
|
650 |
|
1240 |
1443 |
1490 |
1735 |
1880 |
2197 |
µA |
IAM, FRAM (75%)(4) (5) |
FRAM 75% cache hit ratio |
3.0 V |
192 |
261 |
535 |
|
1015 |
1170 |
1290 |
1490 |
1620 |
1870 |
µA |
IAM, FRAM (100%)(4) (5) |
FRAM 100% cache hit ratio |
3.0 V |
125 |
|
237 |
|
450 |
|
670 |
|
790 |
|
µA |
IAM, RAM (6) (5) |
RAM |
3.0 V |
140 |
|
323 |
|
590 |
|
880 |
|
1070 |
|
µA |
IAM, RAM only (7) (5) |
RAM |
3.0 V |
90 |
182 |
292 |
|
540 |
|
830 |
|
1020 |
1313 |
µA |
(1) All inputs are tied to 0 V or to VCC. Outputs do not source or sink any current.
(2) Characterized with program executing typical data processing.
fACLK = 32768 Hz, fMCLK = fSMCLK = fDCO at specified frequency, except for 12 MHz. For 12 MHz, fDCO = 24 MHz and fMCLK = fSMCLK = fDCO/2.
At MCLK frequencies above 8 MHz, the FRAM requires wait states. When wait states are required, the effective MCLK frequency (fMCLK,eff) decreases. The effective MCLK frequency also depends on the cache hit ratio. SMCLK is not affected by the number of wait states or the cache hit ratio.
The following equation can be used to compute fMCLK,eff:
fMCLK,eff = fMCLK / [wait states × (1 – cache hit ratio) + 1]
For example, with 1 wait state and 75% cache hit ratio, fMCKL,eff = fMCLK
/ [1 × (1 – 0.75) + 1] = fMCLK / 1.25.
(3) Represents typical program execution. Program and data reside entirely in FRAM. All execution is from FRAM.
(4) Program resides in FRAM. Data resides in SRAM. Average current dissipation varies with cache hit-to-miss ratio as specified. Cache hit ratio represents number cache accesess divided by the total number of FRAM accesses. For example, a 75% ratio implies three of every four accesses is from cache, and the remaining are FRAM accesses.
(5) See
Section 8.5 for typical curves. Each characteristic equation shown in the graph is computed using the least squares method for best linear fit using the typical data shown in
Section 8.4.
(6) Program and data reside entirely in RAM. All execution is from RAM.
(7) Program and data reside entirely in RAM. All execution is from RAM. FRAM is off.