2
\$\begingroup\$

The circuit shown below is used on one of my PCBs to switch the VBAT supply for the RTC, ensuring timekeeping continues whenever the main supply (VCC) is unavailable.

Circuit Operation

Under normal conditions, when VCC is available, VBAT is powered through D1.

When VCC is removed, VBAT is instead supplied by the coin cell through P1 and Q1. This behavior matches the intended design.

Switching circuit

Issue Observed

After about a month in use, the RTC coin cell (nominally 3 V) was found discharged to 1.6 V. This is unexpected, since such batteries are typically expected to last several years in standby applications.

Measurements

With VCC present:

  • VBAT = 3.117 V
  • Current = –2.25 mA (appears to flow into the coin cell).

With VCC absent:

  • VBAT = 2.893 V
  • Current = ~16 µA

I’m not fully certain if Q1’s biasing is the root cause, but it seems probable given the reverse current flow when VCC is present. How to fix the issue?

\$\endgroup\$
3
  • \$\begingroup\$ What is attached to VBAT? What is the power/current usage of the circuitry at VBAT? \$\endgroup\$ Commented Sep 27 at 23:38
  • \$\begingroup\$ It's connected to the microcontroller's VBAT pin (STM32L451) \$\endgroup\$ Commented Sep 28 at 8:01
  • \$\begingroup\$ OK so the issue is that when VCC is not present, the microcontroller is draining the battery to fast? But what is the current usage of the microcontroller? \$\endgroup\$ Commented Sep 28 at 8:56

2 Answers 2

Reset to default
3
\$\begingroup\$

641-1698-1-ND is recognizable as a DigiKey part number. This is actually a Comchip CDBA240LL-HF.

From https://www.comchiptech.com/admin/files/product/CDBA240LL-HF%20RevD765674.pdf :

enter image description here

Your operating point is indicated in red.

Evidently the diode outperforms the expected leakage by quite a factor: 16µA instead of 120. Or perhaps the storage room was cooler.

Solution: choose a smaller diode. Perhaps only a 100s mA type (e.g. BAT54T1G) is required, if even that (BAS70?). You do not say what the maximum load is, so I will freely assume you have the design notes and knowledge to make an appropriate choice here.

A PN junction diode might also be an option, if the higher VF is acceptable. Leakage of nA is then typical.

Otherwise, some synchronous rectifier or wired-OR controller might be necessary.

\$\endgroup\$
2
\$\begingroup\$

If I understand correct, then with Vcc (+3.3V) present will suppply circuitry attached to VBAT. With this happening, the gate of Q1 is at 0.81V and the source around 3V, so Vgs = -2.19V. According to the datasheet of DMP3099, that is enough to turn it ON. So it makes sense some current is flowing into the coin cell.

To fix it, make sure Vgs of Q1 is less than the threshold of Q1. Simply remove/short R2 so Vcc is connected directly to the gate of Q1. Change R1 to 10K or something higher 100K.

\$\endgroup\$

Your Answer

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

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.