Hi,
Converter configuration is as follows:
Vin=3.3V
Freq=2.25MHz (Rfreq=75k)
V1=1.5V
V2=1.8V
L1=L2=1.5uH, Isat=8.6A (Wuerth’ 74438356015)
Input/output capacitors, feedback voltage dividers and PCB topology are as per data sheet guidelines.
The converter starts Ok and works fine under FCCM. When afterwards set to AAM, channel 2 starts producing some 2.2V instead of 1.8V under no load, poorly regulated (~200mA load causes drop to ~2V). Meanwhile channel 1 keeps working Ok. After switching back to FCCM no relief occurs until after power cycle.
This behavior is consistent with two our pilot samples, and we had to temporarily disable AAM.
Please let me know of the possible reasons. Hopefully it can be mended.
Thank you.
Sergey
Hello @antsupov,
The component values look fine to me. Are you trying to change CCM to AAM while the part is on or are you presetting the mode of operation and then turning on the part?
Best,
Saquib
Hello Saquib,
thank you for the reply!
My converter behaves the same whether I attempt to switch to AAM on the fly, or preset AAM and then turn on the power. I separated mode control from the rest of the schematic to check if the defect is inflicted from outside. I also sought for backdrive in the output net, yet seem to find none (entire output isolation is problematic in my PCB).
Please also consider that normal operation is not restored after returning to FCCM.
Input power (+3.3V) is fed from MP4315GRE converter and shows no any sign of problem.
Looking forward for any suggestions.
Sergey
Hello there,
I’ve tested MPQ2166GD in a different device, yet another one of ours. The only distinctions now are:
Vin=5V (vs 3.3V)
Vout1=1.8V (vs 1.5V)
Vout2=1.2V (vs misbehaved 1.8V)
I didn’t observe any misbehavior there in either output channel whether under CCM or AAM.
I still can’t find any reason for the former converter mulfunction.
I wonder if specific input-output combination can be problematic with this IC?
Regards,
Sergey
Hello @antsupov,
Did you use the same input source for both the application? How comparable are the board designs of the two devices? Can you please try in the former device with the current specifications?
Hello Saquib,
Thank you for your support.
At last I’ve managed to put the cause clear now: it really is an obscure backdrive steming from CLKREQ output of one of the onboard PCIe devices (XR17V352), which happened to be of push-pull type rather than OD. Thus 3.3V power penetrated into 1.8V power net through parasitics of the onboard PCIe clock buffer (PI6CB18401), while MPQ2166GD evidently couldn’t cope with starting into (over-)precharged 1.8V output.
Sorry for putting up much ado about nothing,
Regards,
Sergey.