How do momentary power switches work? How to disable?

[wakeboarder3780]

Or more aptly, how can I modify them to always be on?

More explanation:
Take an electronic device like a TV or a receiver. Most likely it has a momentary power button switch. By this I mean when you press the button in and then remove your finger, the button returns to it's initial position.

By magic unknown to me just by making contact for a brief period of time, the device stays on. Further amazing is the devices ability to be able to then shut itself back off if pressed again.

I do not know or understand how this is possible - although I want to be clear *I fully understand how a momentary push button switch works* (via push-to-make, or push-to-break). What I really want to know is the magic behind the button.

I am doing a project (building an arcade cabinet) where I have several devices that I want to plug into a power strip. Then I want to be able to turn on that one power switch and have everything (including a computer, TV, and receiver) all turn on at once.

To do this I would like to remove the on/off functionality and have devices always be on. Furthermore I'd like to understand what the "magic" is and understand how these devices are working.


Additional info: I've noticed with my TV if I hold the power button down (via tape or something) and then plug it in, it turns on. After unplugging it and then plugging it back it, it again turns on. So this perhaps might be an easy way to mod the TV but I'm nervous about constantly shorting a switch that was designed to make contact for a very short period. Is there a chance I would fry an electrical component doing this?

I've also noticed with my receiver the behavior is not the same so I know there are different implementations behind the push button switch.

 

[davenn]

Or more aptly, how can I modify them to always be on?
By magic unknown to me just by making contact for a brief period of time, the device stays on. Further amazing is the devices ability to be able to then shut itself back off if pressed again.
I do not know or understand how this is possible - although I want to be clear *I fully understand how a momentary push button switch works* (via push-to-make, or push-to-break). What I really want to know is the magic behind the button.

welcome to the forum
no magic ... all the switch is doing is providing a pulse to a pin on the microprocessor chip
the micro reads the pulse and per porgramming activates or deactivates the power supply circuitry.

cheers
Dave

 

[wakeboarder3780]

I assume the power for the pulse comes from the plug in the wall.

I'm a software developer so i can relate to software. So the microprocessor then has code stored in it to decide what to do with pulses as they enter?

How about the original intent of trying to modify power switches to always be one when plugged in? To me it sounds like I'd have to buy my own microprocessors - write the code for them, and then remove the old and wire the new.

How easy is it to do something like this? What about with the TV where in my scenario I was constantly holding the button in with tape. If it's just a microprocessor sensing voltages and then deciding to do something with it, I would recon I literally could epoxy the button in the *in* position and wouldn't harm the microprocessor in any way - is that correct?

More importantly - has anyone done this before? I know it's an odd request but for some applications it comes in really handy. Surely someone who tinkers more might have experience in this category?

Any more information would be greatly appreciated.

 

[(*steve*)]

Just to add to what davenn says, these switches mean that power is always applied to the unit, but the switch controls whether the main obvious functional bits get power (often by turning on and off a large power supply).

Such units that live on standby 24 hours a day, 7 days a week, often consume more power whilst turned "off" than they do while they're turned "on" (given that they're turned off for far longer than they're turned on).

 

[(*steve*)]

You can do this without a microcontroller. They're often used because they're also there for other purposes.

You could use a D flip-flop with each power button press generating a clock pulse where the output is used to enable/disable the device under power.

See here for an explanation of how they function.

 

[wakeboarder3780]

Thanks for the additional info about the source of the power - understood about them always being powered and the switch controlling whether or not it really powers up.

That being said the D flip-flop sounds like a replacement for a momentary power switch - or I didn't understand something (very possible).

What I ultimately want to do is *remove* the ability to turn the device on or off - and always have it on - in other words "if it's plugged into the wall - it's on". I am still unclear how to accomplish this.

Another question that remains unanswered is my question about the TV:
If I permanently glue the power switch in - would that possibly harm any components inside? (remember that I earlier mentioned that the TV behaved in an always on state for some reason if I taped the button in).

 

[(*steve*)]

Gluing the switch (or perhaps placing a jumper across it) may work. But it is quite possible that it won't.

You can easily test this by plugging the unit in while keeping the button pressed.

A more tricky solution is to disconnect the switch and use a monostable to provide a pulse (simulated button press) a second or so after power is applied.

edit: I can't see how keeping the button pressed would hurt anything

 

[wakeboarder3780]

In that case for the TV I think I will just glue the button in.

The receiver however doesn't go into a "stay on" mode if I keep the button depressed however. So perhaps the monostable is the answer to my problem. How expensive are they? When I did a product search for them it seems to come up with a lot of "Monostable multivibrator". Is this the same component you're speaking of (because they're like a dollar)?

If it is, it seems that the have different voltage ratings. I'd assume I'd have to open the receiver and measure one of the leads to the switch for power input and match up with the proper monostable? Do monostables include a time delay or is that another component I need to wire into it? Sorry these components are all new to me =/

Edit: Additional question - is the power that will be feeding the switch DC or AC? I know I'm pulling AC from the wall but is there an AC/DC converter inside?

 

[wakeboarder3780]

Well, I didn't really get an answer and maybe that's because I'm a noob asking stupid questions. I guess I'd just like someone to verify that a monostable 555 chip would be what I'm looking for and perhaps some guidance as to how to provide the right amount of voltage with the 555.

 

[(*steve*)]

Sorry, I was away for a few days. I thought I had responded. I probably failed to hit the post button.

A monostable is a building block, it's not a generic component. There are many considerations.

What voltage does that part of the circuit operate at?

What signal does the switch generate?

With a 555, one option is to have the monostable period start at power up. At the conclusion the output generated could be capacitively coupled to the switch input, generating a brief "press" of the button.

You need to determine the time delay required, the polarity of the signal required, and the minimum length of the signal required. The timings can be approximate as long as you exceed the minimum.

 

[wakeboarder3780]

Ok I took some time this weekend to crack it open in hopes of getting some of the information I need. I took some pictures of the switch circuit board.

Here is the circuit:
http://img848.imageshack.us/i/switchcircuit.jpg/

And here is the back if you just want to see the push button attached to the standby / ground channels:
http://img171.imageshack.us/i/switchcircuitback.jpg/

I will now try to answer the questions you posed:
Q: You need to determine the time delay required:
A: If you mean the time delay before the monostable should send a pulse to turn on the receiver, for the project I'm building there is no harm in going extra long to make sure it's long enough. Surely 2 seconds would be enough (i can plug in the receiver and press the button 2 seconds later and it turns on).

Q: The polarity of the signal required:
A: I wasn't sure what this question meant but the best i can discern you were wondering if I was dealing with a positive or negative signal. Reference the pictures I posted, but the standby signal should be positive, so I'm guessing the answer to your question is positive.

Q: And the minimum length of the signal required:
A: Again erring on the side of caution would be fine. I'm sure a second long duration would be plenty. I can't measure how long i push in a button and then remove my finger but I'd have to guess it's less than 1 second.

To get you the amount of voltage being sent to standby I should be able to measure this with my DMM after plugging the unit in and waiting for 2 seconds correct? I would take the 2 probes and put neg on the ground solder point and the positive probe on the standby lead. As long as I have my DIMM set to DC voltage that should tell me how much voltage the standby lead is operating on correct?

Edit: I also found this page regarding setting up a monostable 555 to behave as a delay circuit:
http://clarkson-uk.com/555-timer/operation/frames3.html

I'm not sure what a few of the symbols mean in the schematic but I think Ra is a resistor right? If I enter 2 seconds for the desired delay I get the following values:

             C          Ra
Option 1     470 uF     3.87 K
Option 2     220 uF     8.26 K
Option 3     100 uF     18.18 K
Option 4     47 uF      38.68 K
Option 5     22 uF      82.64 K
Option 6     10 uF      181.82 K
Option 7     4.7 uF     386.85 K
Option 8     2.2 uF     826.45 K
Option 9     1 uF       1.82 M
Option 10    0.47 uF    3.87 M
Option 11    0.22 uF    8.26 M

I found a 470 uF capacitor so I'm assuming that symbol meant capacitor. I also found the diode listed in that diagram, it seems to be common. I am however having trouble finding a 3.87K resistor.

 

[davenn]

ok, looking at the circuit board pix, it would appear all the switch is doing is providing a 'ground' to the controller IC via the pushbutton switch. So whatever switching you do will have to do the same (i.e. you are NOT providing a voltage to the switch or IC)

Dont know if a 555 timer cct is the way to go, now that I see how the standby switch is wired. Did you ever try Steve's suggestion about holding the standby switch down then turning on the mains power to the TV ?
if so, what happened ? did the TV switch on ?

Dave

 

[wakeboarder3780]

davenn, first things first - we're looking at a home audio receiver currently. The TV already has the simple solution of just keeping the switch depressed at all times. The receiver does not behave this way. If I keep the button depressed on the receiver and plug it in, nothing happens.

I really have a rudimentary understanding of electronics so whatever you guys end up suggesting is the way I'll go. Any other info I can get for you guys?

 

[(*steve*)]

Q: You need to determine the time delay required:
A: If you mean the time delay before the monostable should send a pulse to turn on the receive

Yes, that's what I mean. so we'll assume 2 seconds.

Q: The polarity of the signal required:

Davenn has pointed out that you require a signal to be pulled to ground.

Q: And the minimum length of the signal required:
A: Again erring on the side of caution would be fine. I'm sure a second long duration would be plenty. I can't measure how long i push in a button and then remove my finger but I'd have to guess it's less than 1 second.

I'd be wondering about milliseconds...

Option 4     47 uF      38.68 K

For a whole lot of technical reasons. This is the one I'd go for.

Option 5     22 uF      82.64 K
Option 6     10 uF      181.82 K

And these are also suitable

I am however having trouble finding a 3.87K resistor.

The values don't have to be exact. For this resistor, choose 3.9k. For the other ones listed above, 39k, 82k, and 180k (these turn out to be remarkably close to E12 preferred values anyway!

 

[wakeboarder3780]

So for my understanding could we discuss why the monostable might not be what I'm looking for (davenn's comment above). I'm mainly wondering why davenn thinks that, and also what some other alternatives might be.

Also - for my immediate project I'm going to have to set this aside (not enough time to deal with it) so this isn't quite the burning question I had before but I'd still like to see if I could do it as a side project so I learn something.