I am constructing a solar powered photon (or electron) project and trying to minimise battery use during theperiods when the device is in sleep mode. I estimate that all the sensors and outputs I have will draw around 20 to 30 mA even when not in use.
2 questions:
does the 3V3 output go to zero when the device is in sleep mode? Are the different sleep modes different in their behaviour? Some of the devices could be on this line.
any suggested circuits that can be controlled by the processor to isolate the power to the peripherals. Most of them can be run off the LiPo battery directly. There are optorelays, for example here is one based on âphotoMOsSâ technology that will switch 500 mA for a penalty of 5 mA on the input side : https://www.jaycar.com.au/medias/sys_master/images/hef/he3/8838499368990/SY4090-dataSheetMain.pdf
does this sound feasible?
@AndrewS, I use a 2N7002 N-MOSFET to control power to some devices. I use 10K-100K for the gate pull-down which ensures peripheral power is off while the Photon is sleeping.
Thanks @peekay123, excuse my electronics ignorance but does this also work if the load is between the 0V line and Source pin? In that configuration one MOSFET could switch the supply line for multiple loads.
If you had multiple loads in parallel in the above schematic and having the common GND âdetached/attachedâ via one MOSFET thatâd be just the same.
@AndrewS, in the configuration shown, the N-MOSFET sinks current from the load. In your proposed configuration, the transistor sources current to the load. If you propose to switch 3.3v only, then a P-MOSFET with the gate pulled up to 3.3v will work when the gate is dropped to GND.
You can also use convential NPN/PNP transistors though it all depends on the amount of current you want to switch. The 2N7002 and a typical 2N3904 NPN transistor will switch 200ma just fine.
Thanks to both of you.
Only a few peripherals for this project and have got hold of some 2N7000 so switching the 0V side should be fine. Switching the +V side seems to make more sense as a general rule as I get the impression ground/0V lines can appear anywhere in communication networks (I2C, RX/TX serial). Maybe wrong there.