Electrical switch for a non-chemical battery - How to do it?

Thread Starter

Theo_KVA

Joined Dec 6, 2019
10
Hi everyone,

I’m seeking guidance. I have a tricky situation which I don’t really know how to get out of.

I want to design a system which would consume the least amount of power possible.
The idea is to have an ESP8266 do a simple task every minute (updating a e-paper display) and sleep the rest of the time (about 5s of activity and 55s of sleep). As I want the cycles to be consistent it would be timed with a DS3231 RTC breakout board (which has it's own battery). In fact it would not sleep, it would cut its own power. The idea is derived from this post from pucebaboon.

“The idea [is] to set up an alarm on the DS3231 which will take the (DS3231) interrupt pin low when triggered. That pin is connected to the gate of a P-channel MOSFET, which functions as a power switch for the battery supply to the ESP8266. Once powered, the ESP [do its thing] and then programs a new alarm into the DS3231, clearing the interrupt and thus turning its own power off. This actually works very well. “

But it gets tricky because I want to use the potential energy of a weight to power the system. I’m going to use a stepper motor with a weight attached to its shaft as a generator. (see my previous thread about that subject)
As I understood from this previous thread the amount of energy is extremely low, so my system has to be a super low consumption (which is duable, see the second part of this video, the consumption is super low). Just to put the context, I want it to be a grandfather clock. Mixing electrical technology (epaper) with century-old way of storing energy. For the moment it is just a concept, so a 12h autonomy would be plenty. Even 4h I would take.

Stepper motor have an interesting feature that I don’t really understand, if you short the two wires of the same coil the resistance of the shaft gets really high, I think it would be enough to maintain the weight hanging.
So my idea would be to do the same as pucebaboon but to have a N-channel MOSFET shorting the wires at the output of the stepper motor. Inbetwen would be a full bridge rectifier to turn the AC current DC. As it’s shorten, it would maintain the weight hanging. As soon as the alarm of the DS3231 triggers the gate of the fet open, the resistance drop, the weight is dropping, turning the motor and the current is now powering the ESP.

Capture d’écran 2020-01-11 à 23.10.46.jpg
(I simplified the idea, haven't put any cap)

My question is : I know it’s not going to work, but why?

When I tried to short the wire after the full bridge rectifier the resistance of the shaft is the same (why?). I can’t do it before, as mosfet are only for DC current. Or maybe a low low current relay but again it would only be powerd by the internal battery of the DS3231, which I have absolutely no clues if it exist (does it?). A mechanical solution would mean energy consumption, which I can’t afford.

It is a tricky situation. But I’m sure there is a way around.
What do you think?

:)
 

Thread Starter

Theo_KVA

Joined Dec 6, 2019
10
Thanks for the reply :)
Indeed, I know. The stepper motor is not used here as a motor but as a generator. (Low rpm, decent results)
What keeps the motor from rotating would be the natural resistance when the coil is shorten (You can see it in this video, if you have one try it). But I don't understand why if you short it after a Full Bridge Rectifier it doesn't work..
 

MisterBill2

Joined Jan 23, 2018
5,188
The PM stepper motor, when short circuited, does have a geater resistance to moving at any speed, removing the short circuit will allow turning with much less resistance. But the electrical energy recovered from the motion will not be much. And the voltage drop in the diodes of that bridge rectifier will allow it to turn . And the power dissipated in that pull-up resistor for the FETs will waste a lot of power. So there will not be much by the time it gets to that DC/DC converter. Unless you use a fairly big stepper motor, much bigger than the ones you find in printers. If you do have a stepper motor I suggest that you experiment with it and try lighting an LED.
 
Last edited:

Thread Starter

Theo_KVA

Joined Dec 6, 2019
10
The PM stepper motor, when short circuited, does have a geater resistance to moving at any speed, removing the short circuit will allow turning with much less resistance. But the electrical energy recovered from the motion will not be much. And the voltage drop in the diodes of that bridge rectifier will allow it to turn . And the power dissipated in that pull-up resistor for the FETs will waste a lot of power. So there will not be much by the time it gets to that DC/DC converter. Unless you use a fairly big stepper motor, much bigger than the ones you find in printers. If you do have a stepper motor I suggest that you experiment with it and try lighting an LED.
Thanks a lot for your answer :)
Indeed, I thought about using a larger one.

After a little bit of searching I think I found a way around my problem. I think I'm going to head for a bistable relay.
Apparently it does hold its position without current. That would act like a switch to short the stepper or to connect it to a rectifier. There are a few that activate with 3v. Maybe even adding some kind of capacitor to hold the pulse to activate it, so that the little pulse of the DS32331 could be amplified.

Well anyway, I think my question was way to broad, I should have simplified it. I'll remember next time.
 

dendad

Joined Feb 20, 2016
3,316
Shorting the windings will not stop the motor turning. It only is hard to turn as it moves because the generated current produces a magnetic field that opposes the movement. All that will happen is the motor will turn a bit slower. You will need a mechanical stopper.
Have a look at the air governors uses in cuckoo clocks. A lever moves to hit the air vane and stops the turning. A solenoid can free the governor allowing it to turn and then when it releases, it stops the turning again.
AirGovernor.jpg
That is one way out off it.
Another is to have a wheel with a pin that is released, allowing it to spin, and locks up again on the next turn.
 

MisterBill2

Joined Jan 23, 2018
5,188
If the TS could arrange a low voltage piezo element to control a mechanical braking system to stop the weight from falling until it was activated that would remove the need for attempting to electrically brake the generation system. But I predict no adequate quantity of power being delivered by this scheme even if it performs as desired.
The concept seems to be for creating an electronic clock with a mechanical power source so as to remove the need for battery power. I suggest instead a system using solar cells to charge a battery from whatever light is available, similar to those solar operated battery powered night lights, but with more capable batteries.
 

MisterBill2

Joined Jan 23, 2018
5,188
A flywheel is not a device to hold and release a force consisting of a weight and a wheel. It stores energy as mechanical velocity of a mass. That is not relevant to this conversation.
 

Thread Starter

Theo_KVA

Joined Dec 6, 2019
10
Thanks a lot for these replies! :)

Indeed I thought as well about a mechanical way of stopping the wheel. Maybe some sort of low consumption 3v solenoid?
I didn't know about those levers in the cuckoo clocks. Really inspiring.
@MisterBill2 I was wondering what you meant by a "low voltage piezo element"?
I have a concern that the system to control this opening-closing would consume more that the whole ESP setup. I guess at some point I will have to put a battery somewhere. but avoiding it would be best, is there any way to store enough energy in a capacitor for that purpose?
 
Top