SAM3N00B micro repeated reset = OK?

[eem2am]

Hello,

We are using a SAM3N00B microcontroller in our "special" emergency light application.

Whenever the battery is removed when the mains is ON, the microcontroller repeatedly powers down and powers back up again.......(once every 5 seconds)

Is this a problem? (does repeated power up and down damage the micro?)

Our new software engineer says it is a problem, but can't say why.


Its certainly not a problem from the point of view of the operation of the unit.

SAM3N00B datasheet
http://www.atmel.com/Images/doc11011.pdf

 

[Harald Kapp]

Power cycling doesn't harm the µC - normally.
However, as long as there is mains supply, the µC should not power down. Supplying the µC is the main task of a power supply. Even without a battery. Therefore you should locate the source of the power cycling (e.g. bad voltage regulator, bad capacitors, incorrectly wired watchdog etc.) and eliminate it.

 

[eem2am]

when the battery is removed, the micro gets powered by our resonant inductive coupler.....this raises the battery rail to >8.5V, which also means overvoltaged battery, so we have to get the micro to shut off the coupler when battery rail >8,5V....but then if the battery has been removed we end up constantly power cycling the micro.

 

[KrisBlueNZ]

As an embedded systems designer and programmer, I would consider a situation where the microcontroller is repeatedly powered up and powered down to be a major issue. Certainly it won't damage the microcontroller but that kind of behaviour is totally unacceptable in an embedded system.

It seems like you need to fix whatever it is that causes the battery voltage to exceed 8.5V when the circuit is running from the inductive coupler with the battery removed.

 

[eem2am]

It seems like you need to fix whatever it is that causes the battery voltage to exceed 8.5V when the circuit is running from the inductive coupler with the battery removed

The R.I.C. is basically a current source, and its expensive to stop the above behaviour.
The emergency unit is useless without a battery, so its operation without the battery is not that crucial to control.
As long as constantly resetting/powering up and down does not harm the micro, then its ok.

 

[(*steve*)]

Or turn it into a feature with the power to the uC lighting a "power" LED. If you take the battery out it will flash to warn you of the situation!

Something about lemons and lemonade

 

[Harald Kapp]

You're trying to fix a secondary effect (power cycling) whereas the primary effect (overvoltage) goes unhandled. You should go about it the other way.

its expensive to stop the above behaviour

What's expensive about a small voltage regulator or a plain zener diode? A zener diode might work without a series resistor if your power circuit really behaves like a current source.

 

[(*steve*)]

What's expensive about a small voltage regulator or a plain zener diode? A zener diode might work without a series resistor if your power circuit really behaves like a current source.

I agree 100%. I hope that if you're using something to limit the voltage if you have a power supply like the one I think you're describing.

I was assuming that the voltage rises slightly above 8.5V. If it is actually rising far higher (to the point of approaching or exceeding exceeding absolute maximum ratings) then you absolutely need to fix that first so that your design is safe and reliable.