Battery power from 1 LIPO and 3v3 regulator LDO can't turn ESP32 on

[trustmiao]

We try to power ESP32 with one LIPO battery, but we can't get it work with voltage regulator.

Power is connected to 3.3v and GND on the top left and top right. We connect a LED, firmware is basically blink the LED to test.

Here is the test we've made:
First test, 1 LIPO 3.7v, through HT7333(100ma max) LDO outputs 3.3v to ESP32. the LED is glimmering, looks like it is browned out.
Second test, 1LIPO 3.7v connects to ESP32 directly. Works fine, and LED is blinking , very bright.
Third test, 2 LIPO 7.4v, through HT7333 outputs 3.3v to ESP32, work fine.

We are puzzled, why 1 LIPO couldn't work with HT7333. We've got the following question:
1. what 's wrong with the first test? current not enough? or voltage drop(not measured)
2. can we safely go with the option in second test? do we have to regulate the power connect to 3.3v.

Any input will be very welcomed, we don't think we are the only one testing ESP32 with LIPO, and once figured out, we 'd love to share all the info we conclude.

[ESP_Sprite]

- HT7333 chips can only output 250mA. You need at least 500mA to be able to provide peak power for the RF bits in the chips.
- The dropout voltage of this chip can easily reach >1V (@250mA) which means the ESP32 and the flash chip will only get 2.7V. While the ESP32 likely is OK with it, this browns out most flash chips leading to crashes.

Suggest you try a different LDO. Not an official suggestion, but I personally have good experiences running the ESP32 from a HT7833.

[trustmiao]

Thx buddy, HT7833 is on the way, i will post back after some tests.

[Deouss]

I remember having similar issues - try connecting lipo to 5V boost converter and then connect to 5V input plug

[trustmiao]

I remember having similar issues - try connecting lipo to 5V boost converter and then connect to 5V input plug

That will definitely be an option, but not ideal one, as dc-dc's efficiency is by average 70-80%, a lot of waste.

[trustmiao]

- HT7333 chips can only output 250mA. You need at least 500mA to be able to provide peak power for the RF bits in the chips.
- The dropout voltage of this chip can easily reach >1V (@250mA) which means the ESP32 and the flash chip will only get 2.7V. While the ESP32 likely is OK with it, this browns out most flash chips leading to crashes.

Suggest you try a different LDO. Not an official suggestion, but I personally have good experiences running the ESP32 from a HT7833.

we are still waiting for the hT7833 to arrive, meanwhile, we are designing the circuit. do you add capacitor before and after HT7833? Do we have to add 1000uf after hT7833 to safely prevent the voltage drop?

[ESP_Sprite]

Erm, no, provided that the LiIon can provide the 500mA spikes (most can do this) and you don't feed any other really noisy things from it, there's no need for such an outrageously large capacitor. The datasheet suggests an 1uF cap on the input and a 2.2uF cap on the output. I think, for the output, I usually throw something a bit larger there (2x10uF if I recall correctly) but I don't think there's really a need to go overboard there.

[Deouss]

That will definitely be an option, but not ideal one, as dc-dc's efficiency is by average 70-80%, a lot of waste.

Not with good quality switching boost converter that has 70-90% efficiency depending on IC.
It will convert to nice output depending on how strong current your lipo can provide.

[trustmiao]

Erm, no, provided that the LiIon can provide the 500mA spikes (most can do this) and you don't feed any other really noisy things from it, there's no need for such an outrageously large capacitor. The datasheet suggests an 1uF cap on the input and a 2.2uF cap on the output. I think, for the output, I usually throw something a bit larger there (2x10uF if I recall correctly) but I don't think there's really a need to go overboard there.

We got the ht7833 and it works like charm even without capacitor. We will redesign the scheme with your suggested capacitor. very appreciate it. but there is a new concern, HT7833 has a bigger drop out voltage, 300-500 ma while ht7333 's drop out is just 90mv. will that become a issue when using LIPO discharges and drop to 3.5v.

[trustmiao]

That will definitely be an option, but not ideal one, as dc-dc's efficiency is by average 70-80%, a lot of waste.

Not with good quality switching boost converter that has 70-90% efficiency depending on IC.
It will convert to nice output depending on how strong current your lipo can provide.

and you will add the energy loss caused by 5v to 3.3v regulator built-in, I guess it is am1117, which costs another 11ma, while ht7833 costs much less. Based the good result of our test, I would highly recommend ht7833, which will last longer