Hi,
We have a design incorporating the BQ25100 to charge a small, 35mAh battery.
We have made about 80 units so far, with another 100 in progress. We are planning a small production run later this year of 5,000 units.
Our battery charging path is shown in the simplified block diagram below:
A few notes about the design:
- The battery is a 35mAh rechargeable LiPo battery
- The battery is non-removable
- We have placed a compatible 10k NTC thermistor on the area where the battery gets soldered to the board
When the battery is soldered on, the battery protection IC keeps the VBAT and VBAT_PROTECTED nets cut off from each other. This is also the case if the battery is depleted to less then 2.8V (undervoltage protection), short-circuited, or if "ship mode" is entered.
To wake the protection IC from this cutoff state, a charge voltage of at least 3.6V must be applied to VBAT_PROTECTED. The protection IC will then latch on and connect VBAT to VBAT_PROTECTED.
Our system also incorporates an SoC that has internal regulators set to 1.8V, which powers various other sensors and devices (all come up in low power states so they do not draw much current). The SoC is permanently connected to VBAT_PROTECTED and so it powers on immediately when the BQ25100 begins to output a voltage.
The issue we are seeing:
In some devices, some of the time, the VBAT_PROTECTED net oscillates when 5V is applied to the input of the BQ25100. This prevents VBAT_PROTECTED from developing the necessary 3.6V to latch on the protection IC, and the device never fully powers up.
We have tried a number of things, including disconnecting the SoC, keeping the SoC in reset, increasing the charge current to 80mA from 35mA, removing the NTC to put the BQ25100 in TTDM mode, all to no avail.
An example of the oscillation we are seeing is below. The yellow trace is the VBAT_PROTECTED voltage, and the blue trace is our 1.8V system voltage.
We probed VISET on two units, one exhibiting startup issues and one that starts up fine. The blue trace is VISET, the yellow trace is VBAT_PROTECTED. The left image shows a good startup, the right image shows a bad startup.
A peculiar behavior is that when we heat up the boards (ie: reflow the BQ25100) the device will sometimes startup properly for a while, up to an hour, but then revert to this state.
I can share more specific details and schematics, but I would like to do so via email and get a proper NDA in place.
