How to detect 5V?

Hi,

Are you sure the power good line isn't a logic level signal (i.e. two states, high and low)? Usually, they output something undefined when the PSU output voltage is very low (e.g. just powering up), then they indicate power bad for the milliseconds that the supply is still coming up, then they indicate power good, then they remain there unless the output voltage is not within some threshold in which case the signal indicates power bad by changing logic level again.

So, if you have enough power to drive your circuit from the main supply, then you can trust the logic level of the power good signal.

You could just drive an LED from the main PSU using a resistor and a transistor. Connect the Power Good output via a resistor to the transistor base. Or (even more straightforward, it should only have an extremely dim state during the undefined stage when the voltage is very low), just connect the LED directly to the Power Good line through a resistor.

Ow so you say the PG line would just be 0 or 1 (0 or 5V)... That would idd simplify things!

I read somewhere that the PG line could be 3V so I guessed it was an analog signal.. Did not find anything else on it so I decided to trust that.. Might do some more research!

If it's idd just a digital signal that simplifies everything!

I'll update the thread when I find out more about this!

It is a guess, it depends on your power supply (google the part number in case you can find user documentation for it).

It isn't necessarily 5V, but usually is a voltage, e.g. a 12V supply may output 11V if the power is good, and 0V if the power is bad.

However, you still need to check the documentation, there is no real standard on this. Also, don't assume that the signal means that

the output is exactly 5V or 12V or whatever your supply is rated at. It could be out by 1 volt or more. It is a crude "go/no-go" indicator.

Yeah I googled around for a bit and it seems to be like you said!

Getting an old PSU in a few weeks probably so I'll make sure to read the documentation of that specific piece before buying the components!

Thanks for your help! It seems I was over-complicating things in my head!

Thank you for feeding back your findings.

Let us know how you get on.

"It seems I was over-complicating things in my head!"

But that's good. It lead to you asking questions, forming a specification for what you wanted to do, and then looking for a means to do it. You are learning to be an engineer.

Your second circuit with the zener wouldn't work as you've drawn it. The zener doesn't allow current to flow until the voltage reaches 5V, so with only 5V as the source it never starts to conduct. If you make the zener about 2V, then you've got 2V for the zener, 2V for the LED and about 1V for the resistor. Current would start to flow when the input got up to 4V, and the brightness would then steadily increase as the voltage went to 5V [the brightness of an LED is proportional to the current].

Here's what it looks like in a simulation with a 2V zener. I've got the input (VG1 - a voltage generator) ramping up from 2.5V to 5V and then back down again. You can see the current (measured by AM1 - an ammeter) which flows through the LED start to increase just below 4V. The peak current you can adjust with the resistor (a lot of LEDS have reasonable brightness at just a few milliamps, so don't feel you have to drive it at 20mA)

If you think about it for a moment, you'll see that instead of the zener you could use a second LED. (One of the great things about designing is that there are always several ways to do things.) Wouldn't make sense for a commercial design, but a quick fix if the only components you have are LEDs and resistors.

When I said "you've got 2V for the zener, 2V for the LED and about 1V for the resistor," I was informally giving you one of Kirchhoffs laws. If you don't mind a bit of theory, have a look at

Kirchhoff's circuit laws

https://en.wikipedia.org/wiki/Kirchhoff%27s_circuit_laws

specifically the voltage law. It looks a bit scary, but really it's just a bit of applied common sense. In English, he's just saying we can't get something for nothing. [Technically, Kirchhoff's voltage law doesn't apply to this circuit because the zener and the LED both have some capacitance that can store energy, but once everything settles, after a fraction of a second, the capacitances are no longer relevant and we can treat it as though Kirchhoff did apply.]

Could I suggest that you get one of the free simulators that are on offer? That way you can learn to design for yourself without having to buy lots of components. But don't just rely on simulation; when you get a circuit that works well in simulation, build it and measure it and you'll learn what the weaknesses of simulation are and how the real world doesn't always conform to the models you are given (Mother Nature is a somewhat contrary old lady at times).

TINA-TI  SPICE-Based Analog Simulation Program
http://www.ti.com/tool/tina-ti

LTspice IV
http://www.linear.com/designtools/software/#LTspice

Thanks for the explanation! I really appreciate the effort you put into it!

Putting a 5V zener was a silly mistake idd! And my electronics profesor would kill me if he had known I forgot Kirchoff haha

I'll definitely try out the Spice Simulators. (We actually worked with them at university once, so shouldn't be to hard to master... )

If the measure needs to be precise, there are dedicated ICs: 'Supply Voltage Supervisors'. They are 3-pin devices that can directly drive an LED if voltage is above a certain threshold.

For 5 V monitoring, there's the TPS3809I50D (trips at 4.55 V).

The EEVBlog µCurrent uses a 2.64V one of that family as battery monitor:

For Linux desktop users, just install the wine package first.  Then, you can install and run TINA-TI as if it were a Windows desktop.  Note that I tested on a Xubuntu desktop system.  If I had a RPi Raspbian desktop, I'd try it there too but I am running OpenELEC.

That's a neat idea. For a battery monitor, you could alternatively arrange it with a red LED to the supply to indicate that the battery was below the threshold.

One thing to watch is that the output drive isn't as good as you'd get with a microcontroller I/O pin - 5mA max according to the datasheet.

Although not needed for this project, there is an awesome silabs part that I found recently, called TS9001 (a couple of variants, open drain or usual totem pole type).

It is adjustable, so needs additional resistors, but has ultra-low current consumption, the kind of thing that could be used to build a rechargeable or primary cell monitor that could be permanently or semi-permanently! connected, it has about 0.6uA typical, 1.3uA worst-case (25 degrees C) current consumption (and a tiny bit more that would be eaten up with the additional resistor circuit) which was awesome.

Although not needed for this project, there is an awesome silabs part that I found recently, called TS9001 (a couple of variants, open drain or usual totem pole type).

It is adjustable, so needs additional resistors, but has ultra-low current consumption, the kind of thing that could be used to build a rechargeable or primary cell monitor that could be permanently or semi-permanently! connected, it has about 0.6uA typical, 1.3uA worst-case (25 degrees C) current consumption (and a tiny bit more that would be eaten up with the additional resistor circuit) which was awesome.