I hope to find some help here with powering of sensor module. I have sensor, that was designed primary for battery usage, so it is extremelly low power, now I was aked to try this sensor also indoor with power from mains (110/220V), and here I have already spend hours on internet without any progress. I need power module that will provide the sensor with 5V, but as there will be many of them (start with 1, 2, 10, up to 1000) running 24x7 it is expected that one of biggest advantages (low power) will remain.

The Sensor:
Input voltage: 5 V (+30% -5%)
Input current: 0.005 mA (typical)
Input current on action: 5mA (typical) 10mA (max)
Sensor run time: 24x7
Sensor action: 16 minutes/day (typical) 60 minutes/day (max)

What I have found are PCB (not necessary needed) power modules that have reasonable self power consumption <50mW however they are at least 1W and require minimal load of 5 to 10% (10 to 20 mA) which is still significantly more than sensor consumption. Then there are power source modules that have no minimal load limit, however thay have internal power consumption cca 1W which means (calculated to 1000 sensors) that power supplies will consume 24 000 W/24h and sensors will consume 0.6 W/24h which I consider as unfair :-(

Looking forward to any help.

  • 1
    \$\begingroup\$ Why wouldn't a 5V wall wart work? \$\endgroup\$ – Voltage Spike Sep 20 '16 at 19:11
  • 1
    \$\begingroup\$ 5V wall charger has standby power cca 250mW depending on type, but I did not manage to find out what is the minimal load. I have measured the voltage of the ones I have at home and it is far above 5V I measured between 7 and 8 volts. To the recommendations: I'm not asking about product recommendation, I'm asking if there is technology available for power source for given sensor that is power efficient \$\endgroup\$ – Ice Planet Sep 20 '16 at 19:20
  • 1
    \$\begingroup\$ @brhans The sensor is cca 2 to 3 meters above the terrain and is hard to reach. It is also located in metal enclosure so protected against accidental touch. No isolation or other protection needed, the sensors costs cca 5 USD/piece so some 'loss' by error of power source can be accepted \$\endgroup\$ – Ice Planet Sep 20 '16 at 19:40
  • 1
    \$\begingroup\$ @JimmyB: Thank you for the idea, had a brief look to the efficiency powerguru.org/ultracapacitor-efficiency and seems at cca 80%, plus the management of charge/discharge... Must calculate, think more but seems like worth to investigate. On the sensor I have some computing power left so if inteligently designed the sensor can monitor and automatically (dis)connect the mains dependingon charge level. But will require significant design effort. \$\endgroup\$ – Ice Planet Sep 20 '16 at 19:52
  • 3
    \$\begingroup\$ If isolation is not an issue you could use a cheap transformerless capacitor based power supply (see: ww1.microchip.com/downloads/en/AppNotes/00954A.pdf or electroschematics.com/5678/capacitor-power-supply). \$\endgroup\$ – Stefan Sep 20 '16 at 20:05

This is one way to do it with suitable X1 rated cap and 3kV rated R (or equiv (6*500V) and two 5.6V Zeners ESR of gen purp. 220uF cap 1 Ohm

enter image description here

Simulation (java)

| improve this answer | |
  • \$\begingroup\$ It should work not only with 120 V AC, but also with 220 V AC. \$\endgroup\$ – Uwe Sep 21 '16 at 6:45
  • \$\begingroup\$ The advantage of this design over previous linked solution is less ripple, and smaller X1 cap, otherwise, compromise regulation error for Line range. @Uwe to enable universal use, without switch, needs SMPS ...not possible this way without bad regulation error \$\endgroup\$ – Tony Stewart Sunnyskyguy EE75 Sep 21 '16 at 14:46
  • 1
    \$\begingroup\$ What is the efficiency of this solution? It always seems to me that, with simple power supply, if you drain a mA @5V, you end-up draining a mA directly from the mains, am I mistaken ? \$\endgroup\$ – Sylvain Sep 21 '16 at 17:04
  • \$\begingroup\$ yes but at 70x the voltage drop at 230Vac or 350Vdc, so best case 1/70 W/VAR ratio.. not a big true loss but large X1 caps for VAR divider, 1K is only for caps without self-healing due to surges. \$\endgroup\$ – Tony Stewart Sunnyskyguy EE75 Sep 21 '16 at 17:10
  • 1
    \$\begingroup\$ It is worth mentioning that there is no galvanic isolation, so this circuit (and everything connected to it) should only be used within doubly-insulated devices to be safe. \$\endgroup\$ – dim Sep 21 '16 at 18:22

Not the answer you're looking for? Browse other questions tagged or ask your own question.