78XX positive voltage regulator's sensitivity

[jumper2high]

regulator sensitivity

Hello,
I was building a circuit that was powered from a 12V battery, and included a 5V micro controller. Naturally, a cheap 7805 regulator was used to put the voltage at a reasonable level. Once, by accident, I reversed the polarity on the battery and the voltage regulator was toast. Luckily, nothing else in the circuit got damaged. Lesson learned, a diode used to make reverse current flow impossible.

Then, by ignorance (not being familiar with the circuit 'inside a box') I effectively connected two LED arrays in series instead of parallel. This also seemed to fry the voltage regulator.

Now, I'm curious - are these things really that sensitive? I haven't been using them in a very long time, and I forgot the details - but I think I really used and abused them before - short circuits, overloads, overvoltages .... don't recall ever 'killing' one of 'em.


Also, any ideas on protecting them (I'm running out of regulators ) would be very welcome.

 

[betwixt]

positive voltage regulator

They are usually quite difficult to kill so you did something quite nasty to them!

LEDs are constant voltage devices, they will try to sink as much current as you let them, that's why they normally have resistors in series with them to limit the flow.

My guess, without seeing your design, is you simply 'over fed' the LEDs and the current they consumed was too much for the regulator. Don't forget the regulator is dropping 7V so it will try to dissipate 7 Watts per Amp you pull through it. You might need a heatsink if you don't already have one.

Brian.

 

[jumper2high]

+78xx~+current regulator

BetWixt, thanks for the answer!
I was trying to replace standard blinkers on the car with custom made LEDs. The individual LED arrays (6 white 15000mCd LEDs in the back and 3 in the front and side) all had proper resistors to limit the current. They worked on 12Volts perfectly.

What I didn't count on was the VERY STUPID way the indicator (on the dash) was designed. It was a 12Volt bulb connected between the two positive leads for each side blinkers. That meant, when I turned on the left one, the right set of LEDs was connected as well, but 'through' the bulb.


Also, after checking, it appears the regulator survived the last disaster, it was the Micro Controller that got toasted, which makes more sense, given what really happened. Unfortunately, it opens up a new question - how to protect the Micro Controller

 

[betwixt]

78xx-12

Automotive applications are always prone to damage, it's surprising how much havoc can be raged with such a low voltage!

Remember, 12V nominal battery voltage means it can be anywhere from 12 to about 14 when the engine is running and the battery is in good condition. A poor battery will not be able to 'soak up' the extra charge from the alternator as efficiently and the voltage can go up considerably, maybe to 17V or more if the headlights are off.
To add the the problem, the supply can be very noisy, particularly with motors (heater, wipers, windows etc.) adding spikes.

To be sure the micro is kept safe, add a Zener diode across the supply rails and wire resistors in line with the input and output pins, as close to the chip as possible, where the circuit allows it. In conjunction with the chips own static protection and input capacitance, they will help to kill spikes. Make the Zener one voltage rating higher than the supply, for example 3.6V if you use 3.3V supply and 5.1V for a 5V supply.

Never leave any unused pins of the chip floating, tie them high or low through resistors to prevent them acting like antennas and responding to interference. Also add a capacitor, at least 10uF directly across the chips supply pins.

Brian.

 

[biff44]

voltage regulator low voltage input fried

You either ran too much current thru the regulator, or overheated it (12V-5V)*I amps= Power Dissipated, or you somehow shorted the case to the car body and blew it up.

What package style are you using, and is it heat sunk? As long as when you connected the leds in parallel, the pair STILL HAD a current limiting resistor in series with it, that should not have blown out the chip unless the above stuff happened.

The 7805 can be set up as a current regulator, see the data sheet. That might work better for driving leds (you still need a series diode on each if you are trying to parallel up multiple leds from one 7805).

I would probably choose a switching regulator instead for this application.

 

[jumper2high]

using multiple 7805 regulator

Brian, your post is very helpful indeed.

When everything is off, the battery voltage is 12.5 volts. With the engine running, it's almost constant (RPM independent) at 14.5 volts. The 7805 regulator is rated up to 35V input voltage, so I didn't think it would be a factor. I was even being extra cautious by adding an optocoupler between the MCU's output and the power transistor used to flick the 12V lights.



@Biff44 - the first regulator died after reverse power was applied. The second time, it did survive - but the microcontroller died because of a rather stupid and careless way I made the circuit