Dear Fred. Can I call you Fred? I hope so. Can you help me? I am trying to stabilize a boost converter, but it reminds me of my three-year-old daughter. She won’t behave. Neither will my boost regulator. I know you like design details. My converter is running with a 100Khz clock. Vin is 12V and Vout is 60V. The output current rating is 100mA. I’m using a 10uH inductor.
PS: My daughter is named Fred too. I named her after you. I am your biggest fan.
Unruly in Redmond
Dear Unruly. I prefer to be addressed as Dr. Engleberry. I ask everyone, even my dear mother who turns 104 next month, to call me Dr. Engleberry. She stubbornly refuses, though I have reminded her 1753 times (yes, I keep track of this sort of thing). You can only imagine how aggravating that is. It steams my cappuccino.
With regard to your boost converter…these designs can be tricky, particularly when the inductor current is continuous and the duty cycle is greater than 50%. Let’s take a look and see where you are.
The clock period is 10µSec. can be calculated with the following formula:
As I feared, we are running into a problem called the RHP Zero. RHP stands for Right-Half Plane. This sounds complex, but the idea is simple enough. The inductor is ‘charged’ for 8µSec and discharged for 2µSec. As long as the load is stable, everything runs fine, but let’s imagine a fast increase in load current is demanded. The only way to get more power out of the inductor is to put more into it. How do we put more in? With a longer on-period, of course. But, with a fixed-frequency converter, the larger on-period comes at the expense of the off period. With a shorter off-period, we get less output power.
That’s right. We want more output power, but until we establish a new operating point for the inductor, we get less output power. This problem is characteristic of the boost converter and with these operating conditions, there is nothing we can do. The real problem comes when we try to compensate this power supply. If we have too fast of a responses time, then the converter will try to compensate for the reduced output voltage and we’ll see horrible overshoot and undershoot and even worse, instability in the control loop. So, we must roll off the response loop at a much lower frequency than we’d like to see.
What can we do? Abandon the boost converter topology and use a flyback converter?
PS: This is an odd coincidence, but my mother’s name is Fred too.
You can read more about boost converters in these technical articles:
Specific products related to this are:
This post is also available in: Chinese (Simplified)