How to simulate the maximum output current of a boost DC-DC

[icpaopao]

And, for the consideration of stabilization, I simulate the simplest key loop which contains both voltage and current feedback loops, run a transient, for all cases, it's no oscillation.
But I can't simulate AC. It's so difficult to get convergence. I don't konw how to do with the AC simulation for loop gain.

 

[gszczesz]

Current limit does not change the condition for stability because it doesn't change small signal loop characteristics. It does change the large signal behavior so full-scale oscillation looks different. It's like a voltage clamp on an oscillator: it changes the maximum amplitude only.

Current limit does however introduce a new type of possible oscillation. If your circuit's stable operating point is close to the current limit, then you are at risk. As an example, suppose your inductor operates close to 0.9A current with a very high duty cycle, but you have a current limit of 1A. At 0.9A, the inductor spends most of it's time charging, and a small amount of time dumping the load to the output. If for any reason the current limit gets tripped mid-cycle, then the inductor will dump it's current to the output early, and therefore it will miss a lot of charging time. It's average current will start to drop. The output will start to droop and your feedback circuit will demand an even higher duty cycle.

In a boost, the inductor current needs to rise above the steady-state current in order to raise the voltage at the same time as supply the load current. The recovery phase could then put the inductor into current limit again and repeat. What results is the inductor having a ramped current over several cycle, hitting current limt, and resetting back down again. The average current in this ramp is much lower then during the steady state of 0.9A. This possible condition is why the current limit should always be much higher then the peak expected operating steady-state.

As far as your AC simulation goes, you can't just simulate a large-signal switching circuit with AC. You need to replace the PWM part of the boost with a gm amplifier equivalent for the duty cycle you operate. There are several papers out there that tell you have to do it, and even posts on this forum.

 

[icpaopao]

Does current limit reduce the efficiency?
Current limit would prevent the duty cycle to be too large, the small duty cycle would reduce the boost converter's efficiency.

Added after 12 minutes:

If given the boost dcdc converter has no current limit and not to consume power , the load ability should be depended on the battery.But,current limit decides the maximum output current, so current limit decides a certain power the converter can deliver.
For this standpoint, could current limit limit the efficiency?

 

[gszczesz]

Yes. Notice in the first post I said maximum power delivered usually is set by the current limit.

For most boost converters, light load efficiency is limitted by switching loss (the current it takes to turn the switch off and on).

For very heavy loads, the resistance of the switch, inductor and diode losses dominate. There is a range in the middle where efficiency is usually peak. If your current limit happens before the peak efficiency range is reached, then current limit will certainly limit efficiency. However, if you hit current limit you can no longer deliver the required power to the load, hence efficiency becomes a mute point.

 

[icpaopao]

Thank you. You make me suddenly see the light.
But,how to find the peak efficiency for very heavy loads so that I could set the right current limit.

 

[gszczesz]

In my 3rd post I listed how to calcultate the required peak current. In it I listed the power losses. Since you know your losses, you know your efficiency.