Hi, I have a question about load current range where MPM3650 changes from DCM to CCM
Regards
cmoh
Hello, welcome back to the MPS Forums.
DCM in this case is utilized in this module for light loads whereas CCM is used for more heavy loads. The exact threshold here varied depending on the output voltage (mostly for the max CCM Load current being limited to 5A when Vout = 3.3V). The transition point from DCM to CCM should be seen at ~1A.
This can be seen in each of the following efficiency curves at various outputs:
Here you can visually see the transition between DCM to CCM at a little less than 1A for Vout = 1V, 1.8V, and 3.3V. I am aware this is not an explicit limit listed in the datasheet, but I hope this provides some insight.
Best,
Krishan
Then, in case of condition when load current is around 0.8 ~ 1A range and DCM changes to CCM, what kind of impact can mode change affect to output voltage?
Good question. This can be observed in the regulation between DCM and CCM in the following graphs. Shown are the load regulations for Vout = 1V, 1.8V, and 3.3V.
During the transition from DCM at around ~0.8A - 1A, you should see better load regulation for the output voltage upon the change from DCM to CCM. This of course deviates slightly with increasing load current where the best regulation is seen here at 2.5A to 3A.
Is there any possibility that in above load current range(0.8-1A), output voltage could be suddenly turned off or drop?
Hello again,
Ideally not. In an optimal design, the module smoothly transitions from DCM to CCM with the observed changes in efficiency and regulation as seen by the graphs. The output voltage should not drop or suddenly turn off. Is this behavior that you have observed or just a question of concern?
Situations where this behavior as you describe could be due to insufficient output capacitance that may cause the output voltage to drop between modes or a sudden load transient. Instability in the control loop could also play a role in behavior like this.
Hope this provides some insight.
Best,
Krishan
We designed just as eval board output capacitance 22uF x 4EA. Do we have to add more capacitors ? Datasheet recommends Cff 39pF for 3.3V output voltage. In case of instability in control loop what what kind of actions should I do? Should I change Cff?
Does this mean that you are seeing this behavior?
You should also check the input supply to see if this is stable as well. The layout also plays a factor if you are replicating the EVB schematic exactly. The EVB schematic is vetted and tested by MPS so there are of course other variables at play here if you are indeed seeing dips/drops in Vout.
This is a capture of MPM3650EVB. I was toggling I_load from 840mA ~ to 920mA.
- When MPM3650EVB was loaded 840mA this is a capture of SW_pin.
- When MPM3650EVB was loaded 920mA this is a capture of SW_pin.
- When MPM3650EVB was loaded 840mA ~920mA(Toggle_T=1sec) this is a capture of Vout_pin.
- Variation of Vout_Ripple. (100mA~840mA / 840mA ~ 920mA / Over the 920mA )
-
- 100mA ~ 840mA_Vout_Ripple
-
next reply.
-
- 840mA ~ 920mA_Vout_Ripple
-
next reply.
-
- Over the 920mA_Vout_Ripple
-
next reply.
in my opinion, i think the mode changing of from DCM to CCM cause this power_reset issues.
if my opinion isn’t correct, please let me know.
-Best regard,Krishan.
-
- 100mA ~ 840mA_Vout_Ripple
-
- 840mA ~ 920mA_Vout_Ripple
-
- Over the 920mA_Vout_Ripple
Thank you for the in-lab elaboration on the transitions between DCM and CCM with the EVB.
Looks like your currents appear as expected, but the third image of the first post matches the original question of concern from cmoh where the voltage drops and comes back up at the VOUT pin.
Dear Krishan,
Thank you for your comment.
The images I provided were all captured from the MPM3650_EVB.
Therefore, I believe that the Vout_Reset issues are directly related to the MPM3650 itself.
Is this assumption correct?
If not, could you please share more specific details about the Vout_Reset behavior in the MPM3650_EVB?
I’d like to better understand the root cause and its effects.
Best regards,
Suckhyun
