I got the Powah!

[asgard]

This project started as an attempt to stop the wear and tear on my breadboards.  Most physical wear came from inserting wires to turn on and off the test circuit, so I made this, the Powah.  It's a simple on/off switch (with a little LED).  The schematic and top/bottom pictures are enclosed here.  The board is approx 1 cm^2.  It draws 3.06mA when turned on (the LED, again), but the current draw is unmeasurable with my jack s*&$% multimeter (until I get a uCurrent, that is

I now have the DC supply rails on the breadboards set with switched and unswitched power.  The name suggests the functionality, but mostly I have been having persistent earworms full of Tina Turner (and a bit of Dustin Hoffman in his Midnight Cowboy mode

[c4757p]

Three comments: 1) Make it bipolar. Nice to be able to switch in a positive and negative supply simultaneously. 2) Your breadboards are seriously taking wear from inserting and removing wires? Are you using nails and bent clothes-hangers for wires? 3) What's wrong with just disconnecting it at the power supply?

[Bored@Work]

When I was young we had this thing called mechanical switches.

[mrflibble]

I do like the small form factor. However, when I was young we got these things on the cheap on ebay. Has a switch and doesn't wear out your breadboard unlike those coat hangers.

[Arrow]

You are using a P ch FET , Take good notice of the Vgs max and your max input voltage supply. Because I don't see any protection for that. 

[Kremmen]

When I was young we had this thing called mechanical switches.

Now why did i think of exactly the same thing 

[asgard]

O.K.  Yes I can get both a bipolar switch and Vgs protection, but to do so requires adding another MOSFET SOT-23 device, two zeners (maybe three).  I have used SOD-123 devices previously.  Also three more resistors.  One of the zeners would be necessary to reduce the gate supply rail to 5.1V.  In any case out the window went my real estate and simplicity. 

I have a followon design in the noodle-shop that does all these things, but the gate complexity has gone up precipitously.  It now uses 4 Schmitt NOR gates to set up dual D latches with hysteresis.
J.R.

[M. András]

whats wrong with a rocker switch? or the ps output enable button?

[c4757p]

O.K.  Yes I can get both a bipolar switch and Vgs protection, but to do so requires adding another MOSFET SOT-23 device, two zeners (maybe three).  I have used SOD-123 devices previously.  Also three more resistors.  One of the zeners would be necessary to reduce the gate supply rail to 5.1V.  In any case out the window went my real estate and simplicity. 

Another MOSFET, two or three Zener diodes, and a sprinkling of resistors is too much?

I have a followon design in the noodle-shop that does all these things, but the gate complexity has gone up precipitously.  It now uses 4 Schmitt NOR gates to set up dual D latches with hysteresis.

That is probably too much.

[c4757p]

By the way, I'm assuming you're powering the SN74LVC2G14 by the input power without any regulation, since there's none in the schematic. (That schematic really should show VCC on the chip...) Are you really OK with limiting this to 5V max input and never being able to use it with an analog circuit? The vast majority of them work above the 5.5V maximum. I'd implement all the "logic" (really just one latch) with discrete transistors and design it to work up to 20V or so.

[asgard]

For this version, I am okay with limiting the input supply rail to between 3-5 volts, since it is only for TTL and CMOS-level signals.  At that level the Powah is also useful for a mechanical limit sensor (The SMD switch is a 6N press, but others can be
used if required).  The follow-on device uses the NOR's as a dual D-Latch with smart power-on (the gates can't be eliminated anyway, since there is no such thing as a two-gate Schmitt NOR gate device).  The Schmitt action does two things, one being the hysteresis holds on the second latch reset signal for a time sufficient to debounce the mechanical switch.  Also I can now restore the short-press-on/long-press-off logic I have always wanted this to do.
J.R.
(I can fit the entire works on a .5X.8-inch board).

[c4757p]

You should include some manner of overvoltage protection - it would be really easy to forget about the 5V limit and try to use this with a perfectly reasonable 9V or so.

[mrflibble]

You also should include a pizza time. It can be real easy to forget you have a pizza in the oven when you're busy breadboarding.

Point is: if you keep adding handy "features" that you simply must have ... before you know it your mechanical switch substitute suddenly takes up 45986743589435 parts. Protection this protection that is nice .... but overvoltage protection, seriously? On a device of this nature? How about ... if you are stupid enough to connect +50 Volt then it is allowed to go poof. You don't die and all you lost is < $1 in parts.

Same actually goes for mosfet gate protection with a zener or some such. Personally I might add that one. Then again I might not, depending on how minimalistic I was feeling and how far my arm would have to reach to actually get the zener from the parts collection. Too far? Sod that zener.

Or maybe this is more meant as a design exercise, I don't know. In which case do add the zener and general overvoltage protection. Incidentally, don't forget to add reverse polarity protection. And overcurrent. And thermal. And pizza timer. 

[c4757p]

mrflibble, I'd have to disagree. I'm not talking about multimeter-level protection, I just mean "doesn't blow up at 9V". Seems pretty basic to me. "5V only" is a pretty extreme limitation for a switch. It's fine if it blows at 50V, but not 9.

[G7PSK]

There appears to be something akin to mad cow disease affecting electronic engineers and enthusiasts these days that have caused them to forget the KISS rule and go for ever increasing complexity just for the sake of complexity.https://www.eevblog.com/forum/Smileys/default/smiley_laughing.gif

[asgard]

I tend to agree.  We do succumb to feeping creaturism at times.  My original intent was threefold.  One, do a simple power toggle that I can use as a standard module, that just so happens to also work on a .1" centered breadboard.  Two, make it using only SMD devices for smallness' sake, as well as validating my simplicity principles and design techniques with SMD devices, which I have not previously used for complete projects.  Three, do it in such a way so as to validate using the module also as a motion control limit sensor, for future projects.  Mission accomplished.
J.R.