Power Switch Software Control

beb101

Senior Member
I am trying to construct a latching power switch with the following specifications:

1. 3.3V/5v, 200ma operation

2. push button on-off

3. power on-off from a microcontroller

The Pololu Lv switch satisfies requirements 1 and the off requirement of 3,

http://www.pololu.com/catalog/product/751

A basic latching switch was discussed widely in this link,

http://www.eevblog.com/2012/03/30/eevblog-262-worlds-simplest-soft-latching-power-switch-circuit/

The above switch was improved upon with mosfets here,

http://img36.imageshack.us/img36/420/davidlw.png

Post #10 of this thread by KMoffett has a design similar to the Pololu switch, 08MSWT.jpg,

http://www.picaxeforum.co.uk/showthread.php?7718-Regarding-emergency-stop

And here is another latching switch,

http://www.discovercircuits.com/DJ-Circuits/latch-cir6.htm

My problem is that I cannot find a design that will turn the switch ON under software control. My thought was to use an analog SPDT such as the
TI SN74LVC1G3157 in parallel with the push button. Unfortunatly, I ordered parts from Futurlec and they have my order on hold waiting for supplier SMD adapters. I have the analog switch, but before I order the mosfets from a reputable supplier I thought I would ask about the basic feasibility of trying to integrate the analog switch.

Baxter
 

Haku

Senior Member
Does the push button have to be separate from the microcontroller?

If it doesn't then you can simply have the microcontroller turn on/off the MOSFET directly, with the push button wired to the microcontroller which is programmed to take care of the latching.
 

beb101

Senior Member
Haku, thanks for replying. The switch is to be located on a Netduino Plus shield which is hosting a Picaxe 20x2 to handle I2C and 1-wire tasks for the Netduino. I want the switch to be able to function in a hard reset capacity for the Picaxe and also have the capability to be turned on and off from the Netduino.

So, are you saying I can forget the analog switch? Referring to this circuit,

http://img36.imageshack.us/img36/420/davidlw.png

does this mean that I just need an IO pin connected to pin 1 on the mechanical switch and can drop the RC circuit?

bluejets, your solution makes a lot sense. However, I am very unhappy with Futurlec. I placed an order on April 9th, pulsed them once with an email and they still haven't shipped the order. I will try to find the part elsewhere.

Baxter
 

inglewoodpete

Senior Member
However, I am very unhappy with Futurlec. I placed an order on April 9th, pulsed them once with an email and they still haven't shipped the order. I will try to find the part elsewhere.
Baxter
You just have to be really patient with Futurlec. They offer great prices but don't carry much in stock overheads. So, if you order the wrong part, your whole order waits until they receive that part from their supplier.

If you want fast delivery, pay the premium price charged by your local suppliers.
 

erco

Senior Member
Electronic Goldmine has latching DIP relays for a dollar when on sale. They are rated at 5V but work down to 3.1 V.
 

Dippy

Moderator
Do I take it that:-
a) this latching switch powers something OTHER than the PICAXE?
b) The PICAXE is powered all the time?

If you prefer a solid-state solution, I have attached an ancient drawing which was originally for latching PICAXE power.
It says "To PICAXE.." but that could be anything.
Remove the 'OFF' button connection if you want.

Then the PICAXE can switch the circuit on or off using High or Low <pin>.

The transistors can be suitable BJT or suitable MOSFETs.
I just used one of each for a laugh.
Get it right and the 'OFF' current is in nanoAmps.

There are numerous variations on this theme, just use it to kick-start the grey matter :)
You could use a regulator which has an enable pin. Many options.

The basics have been covered on this Forum at least a dozen times over the last five years , but it is a nightmare for anyone (except Eclectic) to search and find.;)
(Come on Ec!! Out of bed!)
 

Attachments

hippy

Ex-Staff (retired)
The above switch was improved upon with mosfets here,

http://img36.imageshack.us/img36/420/davidlw.png
Isn't it simply the case with that circuit ( and similar ), that to turn power on, the gate of Q1 has to have a positive potential put to it ?

In which case it would seem that any voltage put through a diode + resistor which raised the gate potential would seem capable of doing the same. That could be easily achieved from any micro's output. The diode + resistor would protect the micro from any switch activation. No need for any extra complexity such as analogue switches.
 

erco

Senior Member
Beb101: I'm in Torrance and I have plenty of those Aromat DIP DPDT latching relays. If you're interested and want a couple, PM me quickly, as I leave for vacation Tuesday.
 
Top