ESP8266 Community Forum

- Sat Feb 20, 2016 10:33 am #41508 Glad to hear you got it working. However, I'd offer a few suggestions... 4.2V is the precise maximum you'll ever see on a Lithium cell (unless there's flames coming out of it ) Even then, "freshly off charge" it's likely to be around 4.0V, so just put a Silicon diode in series between the battery and ESP8266. This will drop a constant 0.6V~0.7V meaning your ESP is getting 3.6V absolute max. but more like 3.3V until the battery falls to around 3.6V when fully discharged, giving you 3V. The ESP8266 has proven to work just fine over this voltage range.

Rather than requiring a precise 3.3V as most people imagine, the chip is more "interested" in the "stiffness" of the supply. That's why so many linear regulators fail in this role as they have a relatively long response time (delay between sudden increase in load current and output voltage regulation). This is why "Big" capacitors can solve issues - but even then there's a catch because capacitors also have a certain amount of series resistance (ESR) which drops volts "on their way out". So only exceptionally low ESR caps are going to help.

It's also worth noting that switch-mode regulators are a better bet as they generally respond faster than linear regulators but they come wither a higher parts count and price-tag. On the other hand, a 1N4001 Si. diode represents a parts count of one, and also protects from reverse-supply connection if your battery happens to be removable

- Sat Feb 20, 2016 10:42 am #41509 Thanks, some useful advice. I always struggle with powering my circuits with batteries. I did consider a diode, the issue is I am using a DHT22 which is min 3.3v and it certain does seem to give some odd readings when the the voltage drops below this. Did look for a different humidity sensor but couldn't really find anything suitable.

- Sat Feb 20, 2016 1:29 pm #41525 I have found a different humidity/temp sensor (HTU21D) which operates at 1.5v-3.6v so much better range for running using a diode as you suggested. I have ordered one of them to give it a go.

Quick question, when in sleep mode does the diode consume power and if so how much. I assume it is less than a LDO quiescent current?

- Sat Feb 20, 2016 4:43 pm #41543

I have found a different humidity/temp sensor (HTU21D) which operates at 1.5v-3.6v so much better range for running using a diode as you suggested. I have ordered one of them to give it a go.

Quick question, when in sleep mode does the diode consume power and if so how much. I assume it is less than a LDO quiescent current?

That's the other advantage of using a "series diode drop" to get the Volts down - zero quiescent current. Think about it, there's no return path other than the ESP8266 for current to flow so it's just the quiescent current of the module.

I don't know for sure but it would be very typical in a highly-integrated chip like the ESP8266 to have an on-board LVDO regulator running the critical RF stages at a lower voltage - say 2.5V. The 3.3V supply may even just be for the IO ring. This is why it's possible to run the module at a range of voltages. Just so long as any peripheral components are also happy. A 0.6V spread on Vdd ought not be a problem for any device like this but a sensor is slightly different. However, you seem to have found an alternative for your application which is cool.