Time taken by the signal to rise from 10% of VDD to 90% of VDD and sometimes from 20% to 80% of VDD is known as the rise time.
Time taken by the signal to fall from 90% to 10% of VDD or sometimes 80% to 20% of VDD is known as the signal fall time.
 |
| Fig Transition time |
The transition time denotes how fast is the signal in the sense that how fast it can change from low to high logic or high to low logic. This depends on the cells through which the signal traverses. As a matter of fact, a PMOS introduces more transition time than as NMOS because the holes in PMOS are two times slower than the electrons in NMOS.
 |
| Fig Transition for Nmos and Pmos |
If a cell with high resistance value is encountered by a signal, it’s rise and fall time becomes higher making it a lazy signal that gives a lazy transition.
In order to counter this, there are standard variations of cells created in the library which has higher driver strength and better input output transition properties. This is the result of the low resistance of the cells which makes the signal faster in changing from logic 0 to logic 1. As as example we can cite the difference between normal buffers and clock buffers. The clock buffers have better transition properties than the normal buffers since it conditions a high frequency critical signal in the design.
Transition Violation: When signal takes so long to transit from one logic to another logic that it surpasses a set safe threshold, we see it as transition violation. As such the transition violation is a function of node resistance and capacitance. The transition time as such depends on the input slew and output load of a cell or a design through which the signal of concern traverses. If any of these is out of the limit of the available lookup table, it results in inaccuracy. Now, slew denotes the quality of the signal. It is the rate at which the signal changes and is inversely proportional to the transition time. Larger the transition time slower is the slew. As such, larger is the transition time, more the time it takes to reach the logic 1 or logic 0.
If input slew is low means we have to increment the drive strength of the cell. To increase drive strength we can increase the supply voltage or decrease the load on the net.
Ways to fix trans violation:
In order to counter this, there are standard variations of cells created in the library which has higher driver strength and better input output transition properties. This is the result of the low resistance of the cells which makes the signal faster in changing from logic 0 to logic 1. As as example we can cite the difference between normal buffers and clock buffers. The clock buffers have better transition properties than the normal buffers since it conditions a high frequency critical signal in the design.
Transition Violation: When signal takes so long to transit from one logic to another logic that it surpasses a set safe threshold, we see it as transition violation. As such the transition violation is a function of node resistance and capacitance. The transition time as such depends on the input slew and output load of a cell or a design through which the signal of concern traverses. If any of these is out of the limit of the available lookup table, it results in inaccuracy. Now, slew denotes the quality of the signal. It is the rate at which the signal changes and is inversely proportional to the transition time. Larger the transition time slower is the slew. As such, larger is the transition time, more the time it takes to reach the logic 1 or logic 0.
If input slew is low means we have to increment the drive strength of the cell. To increase drive strength we can increase the supply voltage or decrease the load on the net.
Ways to fix trans violation:
- Insert repeater
- Change the metal layer to a higher layer
- Increase the spacing
- Increase the driver size
- Lower the driver Vt
- If there is a large fan-out it is split to an ideal number
Comments
Post a Comment