2
\$\begingroup\$

The LTC3566, a PMIC intended for battery and USB supplied applications, integrates a battery charger, a power path and a buck-boost converter. For my board, I copy-pasted the schematic found on the last page of the datasheet.

LTC3566 schematic

During my test, I supply VBAT_POS with 3.8 V, to simulate a battery. The dc-dc converter should deliver 1 A, 3.3 V, but if I connect to 3V3 a 5.6 Ω resistor (580 mA), the 3V3 drops to an average of 2.4 V with a 50 kHz, 880 mV ripple (see oscillogram below). In 20 seconds the IC gets hot, increasing of +40°C. A similar behavior happens with lighter loads, e.g. 20 Ω. On the other side, if I apply no load, 3V3 is 3.3 V stable.

Ripple oscillogram

Anyway, if I load Vout (the output of the USB/battey power path) with 500 mA, the voltage is stable, so I assume the problem is on the dc-dc converter side. I already tried the type I compensation suggested in the datasheet on first page, but I get similar results.

What could I investigate?

EDIT: Hereafter the Altium layout, with power components selected in light grey. Resistors are 0402. Inside the board and on TOP layer there are two GND planes. C1 and C3 are on top side connected with many vias.

board layout

\$\endgroup\$
10
  • \$\begingroup\$ @winny The input is stable. The ripple is 50 KiloHz, but the frequency changes according to the load and/or compensation, down to 32 KiloHz. \$\endgroup\$
    – Isacco
    Commented Aug 17 at 20:51
  • \$\begingroup\$ @Isacco - Hi, Please edit the question and link to the source of the original version of the schematic (the datasheet?) since you explain that the diagram was effectively copied from there. TY \$\endgroup\$
    – SamGibson ♦
    Commented Aug 17 at 21:15
  • \$\begingroup\$ Oh, us, not ms. Sorry about that. What’s the ESR of the output capacitor? How’s the loop compensation? \$\endgroup\$
    – winny
    Commented Aug 17 at 21:56
  • 1
    \$\begingroup\$ What is your PCB layout for this project? \$\endgroup\$
    – Ale..chenski
    Commented Aug 18 at 5:49
  • 1
    \$\begingroup\$ It appears from the schematic that instead of tying to VUSB like you do with ILIM0/1, you're sending the 3V3 output of VOUT1 (pin 10) to the MODE input (pin 8), which would seem like an opportunity for inadvertent mode switching when there is a sag in the switching regulator. It's difficult to tell from the single layer PCB screenshot whether this is actually wired that way. Is this maybe just a typo in the schematic? Just wanting to double check obvious things first. \$\endgroup\$
    – pion
    Commented Aug 18 at 9:06

1 Answer 1

Reset to default
2
\$\begingroup\$

The LTC3566 datasheet, in section "Buck-Boost Regulator Burst Mode Operation", states: "The maximum load current in Burst Mode operation is 50mA. [...] if the load current increases beyond this point while in Burst Mode operation the output will lose regulation".

Therefore the solution was to configure pin 8, "MODE", to avoid burst operation. The pin must be connected to ground or left floating, since it has an internal pull-down.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.