Friday Teardown: ESP32 Power Circuit Troubleshooting Part 2

This is part 2 of 3 in my ESP32 Power Circuit Troubleshooting series. This week we take some readings and determine our next steps forward.

This is part 2 of 3 in my ESP32 Power Circuit Troubleshooting series. For the first part, click here.

Last week I attempted to fix an ESP32 board with a funky battery charge controller (TP5400). I let the magic smoke out and now the TP5400 is going into an undefined state (both charging and standby LEDs are on).

This week we take some readings and determine our next steps forward.

Taking Voltage Readings Without A Multimeter

In order to move forward, I needed to be able to take voltage readings from various points on the board. But I didn't have a multimeter available at the time, so how can we take some readings?

A couple years ago, I got some analog panel meters. They are designed to display voltages from 0-5v. This is great because our board should only have 5v at the maximum (USB or TP5400 output).

Behold, revel in the jank. In the above picture, we have the voltmeter hooked up to the ESP32. White wire is negative and is taped to the ESP32's RF shield, which acts as our ground. The green wire, positive, is connected to a small torx screwdriver. This acts as a probe so we can test for voltages without shorting anything out again. Forgive me father, for I have sinned.

With setup complete, we can now begin probing the board to check voltages with the battery installed and USB power applied.

Probe Location

Expected Voltage

Actual Voltage

Condition

TP5400 Pin 1

5v

5v

Good

TP5400 Pin 1 After Diode

4.5v

~4v

Good

TP5400 Pin 5

5v

5v

Good

TP5400 Pin 6

5v

~4v

Good

3.3v Regulator (Source)

4.5v

~4v

Good

3.3v Regulator (Drain

3.3v

~3v

Good

So all the voltages present, maybe damaged for real? Next lets check with the battery removed.

We take some readings again, and we notice something interesting.

Probe Location

Expected Voltage

Actual Voltage

Condition

Inductor/Battery

~4v

~4v

Good

TP5400 Pin 6

~4v

0v

BAD!

Lets take a look at the diagram I drew before:

According to the above, The inductor, battery and pin 6 should all be connected together. But we see ~4v at the inductor and 0v at Pin 6 on the TP5400. In all likelihood, the magic smoke we released when accidentally shorting together pin 5 and 6 was the trace between pin 6 and the inductor burning up. A quick look at the board with a magnifying glass showed that the trace had indeed burned out. So now we need to fix it.

Bridging The Gap

Not my best soldering job. To replace the burnt trace, I took a leg of a resistor and soldered it from the inductor to Pin 6 of the TP5400. It was slightly difficult to do as there SMT components in the way, I don't have a thin tipped soldering iron, and I had ran out of leaded solder, meaning I had to use the lead-free stuff, which has terrible adhesion. But I get it done, and now the trace is replaced.

Testing Again

Now we turn it on and hope for the best.

Slightly better. It only goes into an undefined state when USB power is applied and removed. Charge and Standby leds show correct state when charging/finished charging. So, better, but still broken.

A week later a multimeter arrived and I was able to confirm that the voltages were all correct. Additional testing also showed that the TP5400 was still functioning correct, charging the battery up to 4.2v then stopping.

So Now What?

We clearly now have a broken TP5400 charging chip. It was behaving weirdly before (Standby LED stays on after load and USB removed), but now its behavior is visibly incorrect (Standby and Charge LEDs both on, which is an undefined state).  So our next step is to desolder and replace the TP5400.

As the TP5400 is a chinese chip, this series will be put on backburner for a couple weeks while I wait for a replacement chip to arrive from China.

Lessons Learned So Far

This series has been a valuable learning experience for me, here are some things I've learned so far:

  • Have proper testing equipment on hand. A battery meter or analog volt gauge is not a good substitute for an actual multimeter.
  • Tread carefully. Mistakes like I made by shorting Pins 5 and 6 cost time and money. Take the time to carefully do a task, don't rush.
  • Document everything. This is somewhat blog related, but also important for diagnosis. Write down every step you take, every reading you make. Take pictures of before and after. If you make a mistake, you can use this data to retrace your steps.

That's it for now, next week we will be taking a look at something else!

Back To Top