Ideas for Self Power Turnoff Circuit

Thread Starter

MrAl

Joined Jun 17, 2014
11,389
Hello there,

I was thinking about installing a power turnoff circuit on a device that could benefit from lower power dissipation. It's not a huge amount of power, but still of some concern because it runs off of a battery.

The circuit to be powered draws low current but it would be better to draw even much less current when it does not have to do anything. It only has to take measurements maybe once every 2 minutes to maybe 5 minutes or something like that.

The power turn off circuit would be such that it would apply power to the main circuit, then at some point the main circuit would turn off the power and that would also start a very very low current timer (maybe uA's if possible). After the time is up, the power timer circuit would turn the main circuit back on again, and the cycle would repeat indefinitely.

My first thoughts were to use a few transistors but it's not that great to try to get reliable time periods from a three transistor circuit, so i started to think about a 555 timer, then a CMOS version. The CMOS version should draw very low current while it is timing out. It would then turn on the main circuit through a transistor such as a small NPN type or even PNP ok, and a little extra current to drive that would be ok the base current would be small.
The CMOS version would have to work from about 10 volts up to about 15 volts. Temperatures from about -20F to about 150F.
I would like to get decent repeatability on the timeout period. To start out, 1 minute, then some days later move to 2 minutes and some weeks later 5 minutes. The timeout period should be reasonably accurate but does not have to be super accurate, maybe 20 percent would be good, and that does not have to be an absolute, it only has to be 20 percent from whatever it was to start with. So if it started at 60 seconds if it went down to 50 seconds or up to 70 seconds total over temperature that would be ok i think. If it went down to 40 seconds or up to 80 second i think that would be too much. I hate to say it but i guess this requires an electrolytic cap for the timing.

So that's with the CMOS 555 timer idea. Any other ideas or possibly more suggestions for the CMOS 555 timer idea?
 

LesJones

Joined Jan 8, 2017
4,174
I have used a method copied from the Banggood component tester. I have used the idea with an ATtiny13 micro.
My device is used for a battery powered light. It switches on with the press of a button and shuts down under three conditions.
1 - After a fixed amount of time. (I have used it for a time of 5 minutes.)
2 - A second press of the push button
3 - If the battery voltage drops below a certain value.
Here is the schematic.
Shutdown_timer_a.pngThis is how it works.
The push button S1 is pressed, This causes T1 and T3 to conduct. T1 applies power to the circuit.
The PB4 output from the micro drives T2 on which continues to hold the power on to the circuit via T1
If the push button S1 is pressed now it again causes T3 to conduct pulling input PB0 low. This is a signal to the mico for it to shut down by removing the drive to T2 which removes the drive to T1 which causes the circuit to shutdown.
the micro also causes a shutdown at the end of the time period or if the battery voltage is too low. In that state the only current taken is the collector leakage current of T1 which will be very low.
If you are interested I can post the source code for you to modify to suit your requirements.
Les.
 
Last edited:

Thread Starter

MrAl

Joined Jun 17, 2014
11,389
I have used a method copied from the Banggood component tester. I have used the idea with an ATtiny13 micro.
My device is used for a battery powered light. It switches on with the press of a button and shuts down under three conditions.
1 - After a fixed amount of time. (I have used it for a time of 5 minutes.)
2 - A second press of the push button
3 - If the battery voltage drops below a certain value.
Here is the schematic.
View attachment 284440This is how it works.
The push button S1 is pressed, This causes T1 and T3 to conduct. T1 applies power to the circuit.
The PB4 output from the micro drives T2 on which continues to hold the power on to the circuit via T1
If the push button S1 is pressed now it again causes T3 to conduct pulling input PB0 low. This is a signal to the mico for it to shut down by removing the drive to T2 which removes the drive to T1 which causes the circuit to shutdown.
the micro also causes a shutdown at the end of the time period or if the battery voltage is too low. In that state the only current taken is the collector leakage current of T1 which will be very low.
If you are interested I can post the source code for you to modify to suit your requirements.
Les.

That's an interesting circuit i'll have to look it over some more. There are quite a few parts though :)
 

Thread Starter

MrAl

Joined Jun 17, 2014
11,389
You might find this TI part interesting.
Hi and thanks for the suggestion.

That is the ideal chip except for one detail. The power supply limit is 5.5 volts max. The application shows it working off of a Li-ion cell which is just max around 4.2 volts. I really needed 15v max or about that. During normal operation, the maybe 8 to 15v (or so) gets regulated down to 5v which powers the circuit. I was hoping to be able to switch out the regulator too so what would mean the timer would have to run on 15v max.
What this would mean is i would have to use a very low power 5v regulator just ahead of the chip to supply it with a constant 5v power source. That may be workable i'll have to check the 5v regulators. I have the 2905 regulator which i think consumes very low power but i'll have to double check that been a long time since i used one. If that does work, then i could power this new chip with that and that new chip will turn everything on and off at regular intervals.

I normally would use the watch dog timer of the uC chip but the time interval is just a little too variable and i wanted something a little more precise. The accuracy of that new chip seems even better than i had hoped for so i hope it can be used successfully. It would depend on that 2905 regulator if it really consumes extremely low power.
 
Top