Need help designing push button mains switch based on NE555, Meanwell SMPS, and relay

By using a PNP in the FF, you can get the Idle current down to almost nothing (~72pA). Also note the smaller cap used for reduced relay holding current.
 

Attachments

  • push-push-pnp.png
    push-push-pnp.png
    42.2 KB · Views: 53
  • combined-transistors-one-LEDx.asc
    4.6 KB · Views: 15
  • Like
Reactions: 1 user
Member
Joined 2011
Paid Member
Using one PNP and all NPN seems like a great idea to save on idle current!

I was able to finally figure out simulation for NE555 based solution.
Consumes 4mA while idle
and 4mA+relay current (15mA)+LED current while in operation
 

Attachments

  • sch-ltspice.PNG
    sch-ltspice.PNG
    13.4 KB · Views: 44
Member
Joined 2011
Paid Member
Boards are in for the transistor-based version!

After soldering everything and powering on, it works! Tested for max LED current (20mA) and relay holding current (15mA) for 1 hour - all transistors are cold to the touch!

There is an issue, though. Upon plugging in the power cord into the outlet, the initial state of the switch is ON. I wonder if I should adjust one of the resistors to make the flip-flop start in a specific position?

And minor thing - the SMPS is making a whining sound in both the power-on and power-off states of the switch. I am assuming it is normal SMPS operation, but it’s kind of annoying when it is not 100% silent.
 

Attachments

  • IMG_3056.jpeg
    IMG_3056.jpeg
    362.1 KB · Views: 19
  • IMG_3057.jpeg
    IMG_3057.jpeg
    217.2 KB · Views: 19
Member
Joined 2011
Paid Member
Relay is NO type.

Attached is the schematics where Q1 and Q2 are the flip flop transistors.
Looks like Q1 is open upon the startup. But I need Q2 to be open.
Note resistor R3 that was supposed to pull down base of Q2 to make sure it is closed upon the startup.

I suspect R4 is way higher compared to the total resistance of the other branch loaded with the relay and LED.
I might try reducing it to 22K
 

Attachments

  • sch.PNG
    sch.PNG
    13 KB · Views: 14
Last edited:
Hmmm... I always buy a NO / NC relay.

You could potentially use a pushbutton switch with NO / NC sets of contacts.. (like an E-STOP mushroom switch).. that could also help.

After contributing/reading this thread, I have decided from now on to do point-to-point wiring of the circuit I posted in #13. Very simple; it provides a predictable degree of operational repeatability, which is what you need in this instance.
 
Member
Joined 2011
Paid Member
Assembled 555 based board.

Everything works great. Upon plugging in the power cord, the board stays in OFF. Pushing button turns it off/on.

Upon further testing I found that if the board is in ON and I push the button for more than one second it first turns it off but then cycles to on.

MeanWell PSU is making whining sounds either in on or off. I might just go with small linear psu. There is nice compact transformer from triad - FS12-090-C2
 

Attachments

  • IMG_3065.jpeg
    IMG_3065.jpeg
    314.5 KB · Views: 13
  • IMG_3064.jpeg
    IMG_3064.jpeg
    265.6 KB · Views: 13
  • IMG_3066.jpeg
    IMG_3066.jpeg
    339 KB · Views: 14