Arduino Behavior under power surge or power line fluctuations

How does the arduino uno behave under a power surge or other power line fluctuations? Is this totally dependent on the DC voltage converter you are using? If it is easily fried, what are simple and cheap ways to protect it? Thinking in terms of unreliable power in developing countries.

It is mostly dependent on the DC voltage converter you are using and which power input you are using.

If you are powering from a 5 volt source, and it jumps up to 75 volts, you should expect some damage.

Can you be more specific as to how you are powering the arduino, and what you may expect would cause damage ?

I would not be too concerned about power surges. I have not heard of any trouble from power surges when electronic equipment is used in developing countries. The voltage regulators in their power supplies should handle power surges without any problem.

However it may be worth you testing what happens if there is a short-term power cut or brown-out. I once had to replace 5000 microprocessors in consumer electronics after it was found that they went into an unwanted mode after a short power cut. The effect of the unwanted mode was quite serious. It was only power cuts lasting between 0.3s and 0.8s that caused the trouble.

The arduino is powered from a 5V DC voltage converter that plugs into a wall outlet (120 AC). Eventually I want to switch over to a 220 V prototype but use the same set up. I am concerned with power surges, jumps, falls, or generally unreliable power that occurs in the developing world and it erasing or frying the chip.

Archibald, I have heard of many issues with power messing with electronics equipment in the developing world, especially in hospital settings. Do you know of a simple test set up to simulate some negative power events?

If you are worried about a short dropout, maybe put a 5v5 supercapacitor on the 5V rail. They are not expensive. If you are worried about 5 minute dropouts, then use something different.

Do you know of a simple test set up to simulate some negative power events?

In Europe electronic equipment has to withstand the loss of three cycles of mains. That is a bit tricky to simulate but it can be done.

What you have to do is to decide on the voltage outage that you want to design your equipment to. Only then can you consider how to design the equipment.

hardeehar: Archibald, I have heard of many issues with power messing with electronics equipment in the developing world, especially in hospital settings. Do you know of a simple test set up to simulate some negative power events?

When I had the problem with my microprocessors a while ago, I found professional test equipment that could simulate certain types of brown-out. It was expensive but I did not buy it despite having budget available.

I can fully understand power messing with electronics equipment in the developing world but have you heard of any electronics actually getting fried? I would hope a decent regulated power supply unit would tolerate moderate power surges while maintaining regulation of the output.

I had trouble with a custom home-made burglar alarm waking us up in the middle of the night. I decided it was due to short power interruptions and designed a circuit that would disable the alarm for about a minute while the PIR sensors stabilised. Such circuit design is quite challenging because you need to consider what will happen as the circuit's power supply varies in an undefined manner. We are so used to designing assuming a fixed DC power supply. You may need to consider when to take action as the voltage falls (before you lose all power) and when to reset the circuit when the voltage rises. You could even get two or more brown-outs in quick succession.

You may wish to consider using a microprocessor supervisor and/or watchdog timer. Here are the first I've found: http://www.skyworksinc.com/Products/591/Supervisors_Monitors,_Voltage_Detectors_-_Microprocessors.

"I had trouble with a custom home-made burglar alarm waking us up in the middle of the night. I decided it was due to short power interruptions and designed a circuit that would disable the alarm for about a minute while the PIR sensors stabilised." So, now you have a burglar alarm with a minute delay? That seems a little bit long in duration. Don't post your address here !