Use this forum to chat about hardware specific topics for the ESP8266 (peripherals, memory, clocks, JTAG, programming)

User avatar
By schuh8
#92560 I am comparing the MCP1700 3.3 volt regulator with a simple diode for voltage drop. Assuming a 8266 will operate between 2.5 and 3.6 volts (per spec sheet):

The MCP1700 has a dropout voltage of approximately 0.2 volts. which means it will power an 8266 from full charge (4.2 volts) down to dropout which would be 3.3+0.2 or 3.5 volts. Using a typical lipo discharge curve it shows that about 45% of the lipo’s full charge will be available. quiescent current is 1,6 uA.

Comparing with a simple diode with a forward drop of 0.7 volts: The lipo should now power the 8266 from full charge down to 3.2 volts (3.2V – .7V = 2.5 V (which is the 8266 low voltage cutoff) This uses about 95% of the lipo’s charge and no quiescent current.

Obviously I’m missing something here as the simplicity of the diode, the non existent quiescent current and about 50% greater use of the lipo seem to make it a better choice. Where did I go wrong? Or did I?

Is there a better solution yet?
Thanks for any comments.
Schuh
User avatar
By btidey
#92568 I don't use the MCP regulator but many LDO regulators don't shut off when the input voltage goes below the minimum to maintain the specified output. Instead the output voltage then drops as the input voltage drops with the differential remaining roughly at the same low value significantly lower than a diode.

You also have to bear in mind that using a diode means that for the higher battery voltages you are running the esp above its nominal 3.3V. This may not harm it but the deep sleep current is quite sensitive to the operating voltage so an increased supply could substantially increase the deep sleep current.

Most LI based batteries have supplied most of their stored charge by the time they reach 3.3V and you also have to bear in mind that the supply has to provide significant peaks of current when an ESP8266 wakes up or starts. A supply that is reliant on a battery nearing exhaustion will likely prove unreliable.

Note that some users choose to use lower output regulators like 3.0V or 2.8V and if any attached devices operate OK then this can be both beneficial in further decreasing deep sleep current and extending the input voltage range where the regulator is in its normal operating regime.
User avatar
By schufti
#92572 why not use LiFePO battery? You can use it directly w/o LDO regulator. Using 3.6V (or even 3.55) as upper charging cutoff point does not sacrifice much capacity and is only present for short period.
Additionally this chemistry is much safer (charging) and durable (cycles).
Battery management is easy as one can directly use the internal 3.3V measurement from wifi calibration.
User avatar
By schuh8
#92601 Thanks for the helpful replies!
LiFePO4 would be a good alternative and I will consider for the next project. However I have a large supply of LiPO's I need to use for this project.
I did not realize that LDO's were available that would continue to output current below their rated voltage. I have looked at many spec sheets for LDO's and have a difficult time figuring out which LDO's cease all output once the supply voltage falls below the regulation voltage and which continue to supply current. Is there a name for that specification?
What is really needed is a LDO that regulates at 3.2 volts and stops outputting current when battery voltage drops to 3.2 volts (to prevent LiPO damage). Low dropout voltage and low quiscient current would be an added plus. All this in a TO92 package would be a godsend. Does it exist?