Part 19 · Voltage Regulation · page 1 of 3

Forced CCM vs DCM

Key point: a synchronous FET can pull inductor current negative — Linear’s forced continuous mode (FCM). A diode cannot. That is not a different converter; it is whether the rectifier is allowed to reverse.

Along the way

1
CCM triangle first. \(I_{L,\mathrm{pk}}=I_{\mathrm{avg}}+\Delta i_L/2\). That peak is the switch current you buy silicon for — same story as the topology pages.
2
Light load → \(I_{L,\min}\) wants to go negative. Drop \(I_{\mathrm{OUT}}\) or raise \(L\)-ripple. The math does not stop at zero unless a diode makes it stop.
3
FCM (sync FET). The low-side stays on for \(1-D\). Current reverses through the channel. \(f_{\mathrm{sw}}\) stays fixed. Light-load efficiency is poor: you still switch every cycle and you circulate energy.
4
DCM (diode or body diode). Current hits 0 and stays there. The rest of the period is “dead” — and that is when SW rings (next page).
5
Why Linear forced CCM. Constant frequency for a noisy neighbor, no mode-change at the load you care about, easier compensation. Cost: light-load watts.

What we ignored: current-mode valley control, reverse-recovery of a real diode, and the loop. This is the inductor current picture on a buck.

Buck cell · FCM vs DCM

12 V → 5 V · gold dash = \(I_{L,\mathrm{pk}}\) · grey = 0 A

IOUT A L µH fsw kHz

Diode / DCM

0 µs

Inductor current

Switch node

Mode—

I_L,pk—

I_L,min—

Try it: FCM at 0.35 A — \(I_{L,\min}\) is below zero. Tick DCM: current stops at 0 and SW is no longer a clean square (the ring is the next page). Raise I_OUT until FCM stays positive: that is “CCM without needing to reverse.”