Issues on power delivery (ESP32, PCB)

[Tavi]

Hello everyone,

I have an issue with my power supply/ battery management. 

As you can see in the schematics I attached, I have an SDS011 connected to an ESP32 and is taking power from a step up converter that has at his heart a ME2108
that can hold max 600mA load. Thing is that ESP32 has huge spikes in current draw when I activate WIFI. This issue is non existent if I disconnect the SDS011.

So by activating WIFI + SDS011 the voltage drops in like 2-5 seconds from 5V to 0.60V. I did some debugging on the board and it seems it happens before 5V booster.

Therefore I think the issue is coming from the battery management circuit, now I have no idea how to debug further so a helping hand/mind will be much appreciated. 

[globoy]

I've seen battery protection circuits trip due to high in-rush currents of the devices they are powering - although the 600 mA current max of your boost converter seems far below what most of the protection chips support.   Do you have a spec for the DW1A chip?  Perhaps it will tell you the overcurrent and short-circuit detection thresholds and the time these thresholds need to be exceeded for the chip to disconnect the battery.  These two thresholds are usually different in this kind of chip.

Are you using an oscilloscope to monitor the various power rails during failure to pinpoint the source of the collapse.  Also using the scope to differentially monitor the voltage across a small resistor (< 1 ohm) in series with the battery can tell you exactly what the current is when the system collapses.

You don't show your PCB layout but power connections need to be made with plenty of copper and you should always minimize inductance in both power connections and the circulating paths in a switch mode converter.

What is the SDS011 current consumption?  I am assuming it is some constant.  How close is your baseline (ESP32 booted + SDS011) to 600 mA?  The ESP32 can take 200-300 mA average w/ radios enabled and CPUs at 240 MHz.

As a quick fix you might try increasing the capacitance of C14 if the ME2108 can support a larger output capacitance.  This would act as storage for the in-rush current when WiFi is enabled.

Finally, an unrelated question: Your SDS011 device has a 5V power but looks like it communicates with the ESP32 via a 3.3V serial interface.  Is the SDS011 internal logic operating at 3.3V?  It could over-stress the ESP32 if the SDS011 TX output is at a 5V level.  It could stress the silicon in the SDS011 if its internal logic operates at below 3.3V since the ESP32 TX output is at a 3.3V level.

[Tavi]

@globoy

Thanks for your comment!

Unfortunately I don't have an oscilloscope, and that's why this is hard to figure out  I did some readings with my multimeter and it reaches 500mA max peak. Still that's not the issue since it does not trip on the 5V booster but on the battery management output.

SDS011 power consumption is 100mA.

I've place 3x220uF 16V caps on top of C14 and it still collapses...

The SDS011 is using a 3.3V serial interface with a library that is compatible with ESP32.

Very important thing I forgot to mention is that this issue is never happening if I connect the usb cable and charge the battery. I am almost 90% sure is the battery
circuit that is tripping out.

DW01A datasheet: https://datasheet.lcsc.com/szlcsc/1901091236_PUOLOP-DW01A_C351410.pdf
FS8205A datasheet: https://www.ic-fortune.com/upload/Download/FS8205A-DS-12_EN.pdf

[bugnate]

How about the battery itself, can you tell if it is sagging appreciably under the load? What is the battery?

The datasheet you linked says it won't trip OC until 150mV sensed, and the Rds(on) from the other datasheet you linked says 35 mOhm resistance. So 150mV should be pretty hard to do unless something else is breaking down.

[Tavi]

@bugnate
At max load the battery goes down about 0.02V and that is normal for it size. Is a 700mAh battery 3.7V base.

I am out of ideas why is this happening.

[bugnate]

The only other thing I see is that D2 is misplaced relative to the datasheet application example for the ME2108.

Otherwise, I'd guess to experimentally bypass the battery management and see if that helps.

[floobydust]

Where is the path from +BATT to BATT-OUT?

[Tavi]

@floobydust

Yup I fixed that and is still collapsing. It seems like is doing it only when battery voltage is under ~3.85V

[floobydust]

ESP32 is 3.3V but you have 5V SDS011 with no level-translator for its TXD. ME2108 your circuit has no input capacitor, the battery's ESR might be too high for the converter. C14 is backwards polarity.

[Tavi]

Thanks for all the answers guys, got it fixed.

The issue was the ME2108 is too weak to carry the spikes of current that where induced by the ESP32 wifi.