Question about POP

youshang

Q:
在執行POP模擬跑出穩態波形，AC模擬也無法順利建立(如下圖)



經過嘗試後發現了以下方法能使波形順利跑出來
1.按F8，點選下圖紅色框框的部分，將Use snapshot from previous transient analysis的選項打勾。


2.接著請先執行一次Transient analysis。
3.再請直接執行POP analysis，將可以得到穩態的Vo輸出。


4.再接著執行AC analysis。

請問為什麼這個電路(附件)直接執行POP模擬會抓不到穩態?

為什麼點選了Use snapshot from previoustransient analysis能使波形順利跑出來?


A:
Theschematic contains two Piecewise Linear (PWL) sources.  If there is nosnapshot available from a previous transient simulation, the PWL sources willbe held at their constant values defined at t=0.  So the input voltagesource V9 will be held at 0V and the load current I1 will be held at 0A. Therefore, the POP analysis converges to having an output voltage Vout of 0V.

But if you have run a transient simulation first, and without modifying theschematic, and you have selected to "Use snapshot from the previoustransient," then the PWL sources will be held constant at its last valueof the transient simulation.  Hence, V9 will be 5V and I1 will be 1Aduring the POP analysis.  In addition, all capacitor voltages and inductorcurrents at the end of the last transient simulation will be used as theinitial conditions for the corresponding capacitor and inductor to start thePOP analysis.

This strange behavior of the PWL sources are due to some old behavior of thesimulator.
Abetter way is to assign good initial conditions to the capacitors and inductorcurrents instead of leaving them uninitialized.  For example, the outputcapacitor should be initialized to about 10.8V and the input inductor should beinitialized to about 2.5A.  You can also accomplish this by clicking Simulator| Initial Conditions | Back-annotate after the finish of the transientanalysis.  Having good initial conditions are helpful for the simulationof switching regulators.  The POP analysis usually converges quite rapidlyif there are good initial conditions.
If youreally want to use a soft-start approach to bring the switching regulator up tothe correct operating condition, you should take the following two steps.

A) Replace the PWL Sources by periodic pulse sources with very long pulsewidthsand a very long periods so that, for the purpose of the POP analysis and thetransient analysis, V9 will be rising from 0V to 5V for a short period of timeand after that it becomes 5V during the POP analysis and the transientanalysis.  You should also replace the PWL load current source with apulse current source with very long pulsewidth and long period.

B)  Increase the number of "Cycles Before Launching POP" to makesure the circuit will converge close enough to the steady-state operationbefore allowing the POP analysis to make the prediction/correction.  Inyour case, I have raised this option to 1500 cycles.

I haveattached a schematic to show the steps taken for A) and B).