Maker Pro
Maker Pro

Power supply failure alarm

kong

Sep 26, 2010
122
Joined
Sep 26, 2010
Messages
122
Also, could you give me a basic overview of how this circuit works? That way I can tinker with it and try and figure out a solution. The one you posted, that is. I'm not clear on how it detects a loss of current/voltage and acts upon it. Perhaps if I understood that more clearly I could think of how to fix it.

Thanks again
 

kong

Sep 26, 2010
122
Joined
Sep 26, 2010
Messages
122
Specifically, how would a PNP transistor turn on if its base is tied to V+?
 

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
Yes, the triangle in the diode points in the direction of current flow, for "conventional" current, that is, current that flows from positive to negative. Yes, the diode blocks the discharge of the capacitor. Without it, the capacitor will discharge back into the load that you're monitoring, and it will discharge too quickly to sound the alarm.

Re connecting the battery, lamp, and power failure circuit, what is important is that the circuit is still connected across the lamp when you disconnect the battery.

Yes, I read your requirements, and they're the same as what I thought you wanted.

This is how the circuit works.

When the power supply is working normally, current flows through D1 and charges C1 up to about 0.7V less than the power source voltage (the 0.7V is lost in the diode junction). When power fails, C1 holds its charge because D1 prevents current flow from C1 back into the load that you're monitoring.

Q1 responds to the voltage difference between the base (on the left side in the diagram) and the emitter (at the top). The transistor will turn ON (conduct) when the base is brought negative relative to the emitter by about 0.7V. (Base negative with respect to emitter is correct for PNP transistors; it's positive for NPN transistors.)

When the circuit is connected to a power supply, let's say 12V, the positive input (top of diagram) has +12V on it, relative to ground (bottom of diagram). Therefore C1 will be charged to about +11.3V. The voltage on Q1 base will be pulled to +12V by R1. Therefore Q1 will be reverse-biased, i.e. the voltage on its base will be POSITIVE relative to its emitter, and it will be biased OFF. So no current flows through Q1.

When the power supply fails, the voltage on the circuit's positive input drops to zero, because the load pulls it to zero. R1 pulls Q1's base towards zero as well, but the resistance of R1 causes the current to be limited, and Q1's base voltage only falls low enough to turn Q1 ON (make Q1 conduct), i.e. Q1's base will be about 0.7V negative relative to its emitter. C1 keeps voltage present on the emitter.

When Q1 conducts, current flows from C1 into Q1 emitter, through Q1, out Q1 collector, and powers up the beeper circuit. The beeper circuit draws current, which will discharge C1 over a short period of time.

The transistor's base is not tied to V+. The EMITTER is tied to V+.
The BASE is the thick flat bit at the left in the drawing.
The EMITTER is the wire with the arrow. It points inwards towards the base for a PNP, and outwards for an NPN.
The COLLECTOR is the other terminal without an arrow on it.
 

kong

Sep 26, 2010
122
Joined
Sep 26, 2010
Messages
122
Ok thanks, yeah it all sounds correct. I mean, as to what I have breadboarded here. I wish I knew what to do

but I guess I'll just keep trying things.

haha thanks for your help
 

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
Can you take some zoomed photos of your breadboard?
 

kong

Sep 26, 2010
122
Joined
Sep 26, 2010
Messages
122
This is interesting though, when I disconnect power, is the lamp supposed to fade out?

That makes it seem like the capacitor is discharging into the lamp and not the 555 osc. hmm

I keep replacing the diode to see if it's bad but nothing yet
 

kong

Sep 26, 2010
122
Joined
Sep 26, 2010
Messages
122
Yeah, i need my ipod to charge a little then I will....shouldn't be long
 

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
So if the monitor circuit is connected across the lamp, when you disconnect the battery, the lamp fades out more slowly than if the circuit isn't connected? That's definitely not right. That sounds like the diode is shorted. If the diode itself is OK, maybe there's an error in your breadboard layout?
 

kong

Sep 26, 2010
122
Joined
Sep 26, 2010
Messages
122
If the monitor circuit isn't connected across the lamp when I disconnect the battery, the light goes out immediately. If it is connected, it fades out as if the cap is discharging into the lamp. (maybe it is, bad diode?)
 

kong

Sep 26, 2010
122
Joined
Sep 26, 2010
Messages
122
file:///C:/Users/George/Desktop/circuit%20power.jpg
Here is the image
 

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
That's not right. Pull the diode out and try again. If the lamp goes out immediately now, the diode is shorted. If it still goes out slowly, there's a short somewhere else.
 

kong

Sep 26, 2010
122
Joined
Sep 26, 2010
Messages
122
sorry I'm not sure how to post it...I think if you copy and paste that into a browser you'll get it
 

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
When you first connect the circuit to the power supply, there is nothing to limit the amount of current that flows through the diode. When C1 is discharged, it will draw quite a lot of current when it starts to charge. Maybe the diode is getting damaged each time you test it. Try adding a 10 ohm resistor in series with the diode.

You are using a 12VDC power supply and lamp?
 

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
file:///C:/Users/George/Desktop/circuit%20power.jpg
Here is the image
That location is on the hard drive of your computer! I can't access it there. You can attach it to your post - use the "Go Advanced" button, I think it's called, and click the paper clip in the top middle of the editing window.
 

kong

Sep 26, 2010
122
Joined
Sep 26, 2010
Messages
122
still goes out slowly. ****.

Tested the transistors and they are OK.

Could it be the cap? How would I test it?

The only other thing in the circuit is the resistor-I don't think those go bad, lol
 

kong

Sep 26, 2010
122
Joined
Sep 26, 2010
Messages
122
Ok here
 

Attachments

  • circuit power.jpg
    circuit power.jpg
    106.4 KB · Views: 102

kong

Sep 26, 2010
122
Joined
Sep 26, 2010
Messages
122
using a 9 volt supply....tried the resistor in front of diode

didn't stop the light from fading
 

kong

Sep 26, 2010
122
Joined
Sep 26, 2010
Messages
122
When you get a moment, maybe if you tried the same circuit and got it to work, we could figure out what was wrong? It has to be something simple.

Maybe I just have a shoddy connection somewhere...you know, if you got it to work it would be proof that it does work, then I can try it again with a different breadboard or something. Sorry for taking up so much of your time
 

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
Ahem... The twin strips that run sideways... they're connected together along the whole length, not just in groups of five holes like the other strips are!
It's probably a good idea to add a 10 ohm resistor in series with the diode anyway.
 

kong

Sep 26, 2010
122
Joined
Sep 26, 2010
Messages
122
Ok I will. Yeah the power strips you mean? They should connect all the way down. Why, did I screw up? lol
 
Top