For the 2-phase application, the
design process is similar to that of the single phase, except:
- In selecting the voltage loop
compensation, N = 2 divides the effective current sense amplifier gain in half
as the two phases operate in parallel, each delivering equal current, doubling
the current gain of the converter.
- For ceramic input bypassing
capacitors, TI recommends that each phase have sufficient bypassing to
support it as if it were a single-phase output. While there is some channel to
channel current sharing, layout and trace inductance often results in actual
ripple current sharing being significantly lower than the ideal input ripple
cancellation.
- MSEL2 of the primary channel
(MSEL2_A) is selected for a 2-phase converter with 3-ms TON_RISE and the maximum
current limit by being left open.
- VSEL and ADRSEL of the primary
channel are programed the same as a single phase converter.
- MSEL2 of the follower channel
(MSEL2_B) is shorted to AGND to select follower of a 2-phase converter with the
maximum current limit setting.
- GOSNS/FLWR_B is pulled up to
BP1V5 to set channel B to a follower, using Channel A’s voltage regulation and
error amplifier as well as PMBus interface.
- VSHARE_A and VSHARE_B are
connected together along with BCX_CLK A and B, and BCX_DAT A and B.
- MSEL1, VSEL, ADRSEL, PGOOD,
PMB_CLK, and PMB_DAT on Channel B are all unused, and connected to GND.