Power switch with momentary pushbutton. Critique sought.

Thread Starter

kender

Joined Jan 17, 2007
264
I'd like to use a single momentary tactile switch to power up and down a system and was trying to figure a simple way to do this. I found a EDN article with a circuit made of 10 discrete parts that can power up a system (but not down). I need a smaller part count. I’ve put together a schematic. Could you critique my circuit?

Thanks,
- Nick
 

Attachments

thingmaker3

Joined May 16, 2005
5,083
I'm having trouble reading the part number on U2. (At least I think its U2...)

If "U2" needs a negative going pulse to switch states, then you'll need to re-design. Pushing S1 puts a temporary ground on U44 (U4A?) rather than a temporary high.
 

Ron H

Joined Apr 14, 2005
7,063
As I understand it, output voltage is only present when the switch is depressed (after you correct the sense of the switch). Is that what you intend? Your post title left me with the impression that you wanted the power to toggle with each actuation of the switch.
 

Thread Starter

kender

Joined Jan 17, 2007
264
As I understand it, output voltage is only present when the switch is depressed (after you correct the sense of the switch). Is that what you intend? Your post title left me with the impression that you wanted the power to toggle with each actuation of the switch.
I’ve updated the schematic.

The PIC firmware would act as a latch. Here’s how I see this work:

1. The device is off. The only part that’s on in the device is the NOR gate U4. EN# on the LDO is high, because both inputs to the NOR are pulled up with resistors.
2. User presses the ON/OFF button SW3. He is instructed to hold the button down for a second or two.
3. The NOR gate turns on the LDO U2.
4. PIC gets power from the LDO and immediately drives the POWER_ON line up and keeps it up.
5. The user releases the ON/OF button. The LDO and the rest of the device don’t turn off, because the PIC keeps POWER_ON up.
6. After a few seconds the PIC starts polling the POWER_OFF line.
7. Device does its thing.
8. The user wants to turn the device off. He presses the ON/OFF button.
9. PIC sees the POWER_OFF (because it’s still polling it) and drives the POWER_ON low. The device doesn’t turn off completely yet, because the used might be holding the ON/OFF.
10. The user releases the ON/OFF. Now the device turns off.
 

Ron H

Joined Apr 14, 2005
7,063
The only potential problem I see is that the supply voltage to your NOR gate varies, and therefore so does the logic threshold, while the logic high output of your PIC is constant at 3 volts. You may need a level translator between the PIC and the NOR gate. Either that, or a regulator for the NOR gate.:rolleyes:
 

Thread Starter

kender

Joined Jan 17, 2007
264
Why does the supply to the NOR gate vary? Is this a quirk of the charger or lithium cell?
Both, I guess. The gate is powered from the cell. The cell voltage will be decreasing as the device is using up the battery and increasing as the charger is charging it up.
 

Thread Starter

kender

Joined Jan 17, 2007
264
Guys,

Thanks a lot for the critique!

Voltage differences are taken care of in v0.2. 74LVC3G07 has a sufficiently low threshold.

Also, here’s a new v0.3, which is somewhat simpler. The principles of operation are very similar, except the POWER_OFF signal is derived from a separate pole of a momentary DPST button.

Advantages of v0.3
- Lower part count (especially if Q1 and Q2 can be integrated into one high-side MOSFET switch IC)
- Regulator (downstream) doesn’t need to have an enable pin.

Disadvantages of v0.3
- It requires a DPST button.


-Nick
 

Attachments

Thread Starter

kender

Joined Jan 17, 2007
264
FYI. There's a family of on/off pushbutton controllers made by Linear Technology. LTC2950 through LTC2954. Most of them have a SOT-23-8 package.
 
Top