Go Down

Topic: Preventing shorts and power failures (Read 3896 times) previous topic - next topic


Dear Arduino Community,

I'm working in a garden automation project that involves remote sensors and electrovalves, that share the Arduinos power supply. Since the sensors and electrovalves are remote devices, they might get damaged and even their power cables shorted. Also, a short might leave the Arduino without power.
So I'm interested in implementing a simple circuit that will protect the rest of the system of such a problem.

The easy answer is to put fuses for each of these devices, but I was wondering if there is another solution that might be reseted per software.

One of my questions is: how fast should this circuit react to avoid parts to burn/explode?

I've seen circuits like this that are a solution:

On the other hand, I was thinking about putting a low value resistor in series with the load, and using a comparator to compare this with a voltage divider, this should give me a Low if the current is ok, and a High if its too much or short, and this can shut down the sensor. I don't know how to calculate the Wattage the series resistor should tolerate, resistance sound easy just applying Ohm's law.

Hope you understand my explanation. Regards.


Standard linear voltage regulators typically have full output short-circuit protection built-in - just use a separate one for each part of the circuit, and then think about the consequences of some parts being powered down and others powered up (if its just sensors that get shorted, typically nothing bad happens, but if an output valve has lost power you might be feeding a logic signal down a wire to a short circuit - adding 1k resistors in series with such signals can protect output pins.

You might also want a few voltage dividers to monitor various power rails so your code can detect damage.
[ I will NOT respond to personal messages, I WILL delete them, use the forum please ]


Wallwarts are cheap. If I were making such a circuit, I would use a separate
power supply for the sensors, so the Arduino would always stay active in
case for remote shorts.

I would place a small value R in series in the high side of the power to the
remotes, to sense current flow differentially using voltage dividers to the
Arduino A/D converter pins, and use the Arduino to control the pass
transistor, Q1. However, I don't understand how your ckt with the SCR is
supposed to work [some kind of crowbar?]. You don't really need a crowbar,
just a way to turn on/off Q1 using the Arduino.


I like the idea of the separate linear regulators. But can such part survive in constant short circuit o is it just  a temporry protection? Will they dissipate the current as heat or will they really shut down?

The SCR circuit is published as an electronic fuse. I don´t understan why does it need such a big transistor.


If the remote sensors require power, and are going to be more than a few inches
from the Arduino bd, you really want to have separate v.regs near the sensors,
plus filters/bypass capacitors. It's generally not the best idea to run 5V regulated
power over long wires, and expect everything to still work well.

Some v.regs have over-current and thermal shutdown ckts inside the parts, but
you need to check the datasheets. Surface-mount parts may be less robust than large
TO-220 parts. Also, some v.regs fry instantly if you hookup Vin with reverse polarity,
while the old LM7805s can handle it better.

However, if you use the Arduino to sense the load current of the sensors, as indicated
previously, then you can shut down the power to the sensors in case of shorts.

The 3N3055 is a high power transistor, but you really don't need such a part for a
few mA for sensors.


May I ask a "stupid" question?

Though it has been a long time since I have done that kind of electronics, looking at the first post and the "fuse".

I got confused where the main path is until I realised the main path is via the transistor - right?

The circuit doesn't look right.

Ok, here's what I see:
At start up, the SCR isn't conducting and the base of the transistor is "LOW".  So it is off.
The button is pressed and the base is taken HIGH and so the transistor turns on.
However, there is no base current limiting, and transistors are currrent driven.
Though you want it to be ON and not just "on" FULL + voltage on the base?

Then I am wondering how it turns off when a short circuit happens on the output.
After R1 is the "load" and so if a short happens, the emitter of the transistor and R1 would be at 0v, rather than onto the load.

There is nothing there to bias the transistor OFF and stop the SCR conducting - that I can resolve.

Sorry, but what am I missing?

I shall run off and ask "Mr Google" but want to mention it here for the sake of getting pgmartin to check that part of it.


I found a bit of reading on google about that circuit and they also said it doesn't look "the best" and would be prone to not working correctly.

MY change would be - using the original circuit - to put the LOAD between the annode of the SCR and the collector of Q1.

Though also not 1oo% guaranteed, I think it would be better in some ways.

But what you really need is something to turn off the power if there is a short.

My modification would stop the power if there was a quick short, rather than a full on continuous short.

Ok, I shall stop talking now.


I´ve simulated the SCR circuit in Multisim (before burning something) and it doesn´t work as expected. Current limitation doesn´t work, it doesn´t start as normally on, looks like it only cuts with a short. I´ll have to risk some components and build it...but I´m not convinced it can work.

So I´ve learnt to use separate power regulators for the sensors. Main power supply is 12V, sensors are 5V, and electrovalves are 12V.
So I´m looking for something to replace a regular fuse to control the current of the 12V line.

What do you suggest?


All you really need is a simple pass-transistor switch, for example,


The NPN "output" isn't required for your app. I would add a 10K pullup-R between Vin [you'd use 12V there] and the PNP base, to ensure the PNP turns off when the NPN is off.



Couple of things:

As far as I know, SCR's aren't "Conducting at startup".  They need a signal on their gate to start them conducting.

Also, I didn't know SCR's do current limiting.

They are ON or OFF.

I have bashed a couple of thoughts out but haven't tried them as similar to the origina circuit.
Few components and SHOULD work.

I can try to make a JPG of the schematic if you are interested.  But it is really difficult to make drawings with the programs I have at hand.


Ok, here is what I think may be what you need.

1 POT/variable resistor.
1 pushbutton swith.
1 power NPN transistor.
1 power PNP transistor.

I can't attatch a picture because just now I can't get the drawing thing working.


Power supply
Emitter of PNP transistor.
Base of PNP connected to pushbutton switch.
Other side of pushbutton swith connected to Collector of PNP transistor.
Collector of PNP connected to Pot (stationary part) and Collector of NPN transistor.
Base of NPN transistor connected to to the moving part of the pot.  More on this later.
The other "stationary part" of the pot is connected to the Emitter of the NPN transistor.
The Emitter is also connected to the "LOAD".

What is going on - as far as I can work out:
At normal everything is not conducting.
The POT needs adjusting so with the load connected it turns  on the transistor.
To test, you will need to keep the press button pressed.
At one extreme of the POT it should turn off because the base current would be too low to turn on the transistor if there was a short.  This is where the "Load" comes into the equation.
As it isn't a short, there will be a voltage drop over it and so with a bit of tweaking, the pot will set the trip voltage to turn off the NPN transistor.  So from there let's look at the PNP.  It is turned off because of the base being not connected to anything and so there it  would not be allowing any current.  When the button is pressed, it takes the base low and it turns on.  In doing so, the base current starts to flow and the NPN turns on.  The load is active.  If there is a short, the base current on the NON stops and so the NPN turns off.  Doing so the PNP is turned off because its base it taken high.

Hope that works.


Oct 27, 2012, 03:37 am Last Edit: Oct 29, 2012, 05:11 pm by pgmartin Reason: 1
Dear lost_and_confused,
I tried to understand your idea, but wasn´t able do it and, to get a working circuit.
But your help drove me ina a way out of SCRs and into simple transistor solutions. 

So I kept searching and found this circuit:

This one has passed a simulation, now I´ll have to build and test it.


H again.

Ok, I am glad you found what you needed.

But just for the record, here is my idea.

You will have to read my earlier post to get an idea of how I think it should work.

If you use bigger transistors (higher power) then you can use more current.   Ofcourse the big trade off is the power used by R1.

I don't know if it IS needed, and may not actually be needed.


Dec 04, 2012, 04:34 pm Last Edit: Dec 04, 2012, 05:01 pm by pgmartin Reason: 1
I built the circuit from my last post (Or you can find here http://320volt.com/12v-ve-5v-600ma-irf9530-mosfetli-akim-sinirlama-devresi/). Worked like a Charm. Added a couple of LEDs to monitor if it was in normal mode or in overcurrent protection (One at the Gate and one at the Drain of T2, plus 1K resistors in series).

I understad the basic working principle of this circuit: when the voltage drop in R1 is bigger tha 0.7V, T1 starts sto conduct and shuts T2 off. Thats the easy part.

What I do not get is the role of the two capacitors. Any ideas?

C1 is quite big, and and of the drawbacks of this solutions is that it has quite a big charge after plugging it off from the power source.



Couple of things:

As far as I know, SCR's aren't "Conducting at startup".  They need a signal on their gate to start them conducting.

Also, I didn't know SCR's do current limiting.

They are ON or OFF.

I have bashed a couple of thoughts out but haven't tried them as similar to the origina circuit.
Few components and SHOULD work.

I can try to make a JPG of the schematic if you are interested.  But it is really difficult to make drawings with the programs I have at hand.

SCRs or thyristors turn on when when there is current from the gate OR when the rate of change of voltage is high enough (due to internal
capacitance).  Simply connecting the supply might trigger conduction due to the sudden increase in voltage.

And yes you are right they don't do current limiting other than via melting!
[ I will NOT respond to personal messages, I WILL delete them, use the forum please ]

Go Up

Please enter a valid email to subscribe

Confirm your email address

We need to confirm your email address.
To complete the subscription, please click the link in the email we just sent you.

Thank you for subscribing!

via Egeo 16
Torino, 10131