mega32 burned out? - possible solution

Go To Last Post
7 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

First hi to everybody. This is not my first post here but the last one was a long time ago..

I'm prototyping a MIDI Controller using an atmega32 @16Mhz. Today I was testing some of the funcs and suddenly everything stopped working. The first thing to notice was a voltage drop on the overall circuit (using LM7805 to power the micro and other devices with 5V).

Later I found a short circuit.... it was *inside* the atmega32, between VCC and GND pins (10 and 11 in DIP40 package).

Sounds too strange to me.. I took the avr out and measured DC resistance with an ohmmeter to find 2 ohms!! (between VCC and GND in atmega32)

clearly, that's the thing.. Once I took it out voltage got back to 5V.

I'm currently checking for short circuits but I have a question: has anyone experienced this thing?? when?
I thought I might get an open circuit when something gets burned, but not a short one!

thanks to everyone in advance

Carlos

Last Edited: Mon. May 29, 2006 - 07:55 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Semiconductors can fail in the open state or the shorted state. I can't give you any specific reasons as to what type of faults will cause the semiconductor to fail with a short or an open circuit but, it happens either way.

I wouldn't spend too much time worrying about it though, it's really quite a rare thing.

The only failed AVR microcontroller I've had on more then two years of their use was, ironically, with a Mega32. But I know what I did to damage the microcontroller. I connected one of the outputs of the Mega32 to 24VDC. Needless to say, the Mega32 didn't care for that too much. But then too, I was a bit luckier then you because I was able to still use the Mega32 - minus one blown output pin.

You can avoid reality, for a while.  But you can't avoid the consequences of reality! - C.W. Livingston

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

thanks for the quick reply!

I couldn't find out where the cause was.. maybe some wire on the table got below the PCB and blew everything.. don't know :(

maybe I should put a current limiting device before powering the mega32, just to be sure nothing's wrong there.. I think 10mA should be enough, maybe 50.. I have to figure out the circuit now.

well, thanks again!

Carlos

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

well I burned my third consecutive mega32 a few days ago.
Sad, of course, I've been looking for an explanation.

My board is powered using a 7805A.
The question is that I forgot to put a capacitor at the input, and I think that's what was causing those three "smoking" accidents (actually no smoke came out, just vcc and gnd shorted inside the AVR).

I began to think about this since I now remember that the chips stopped working after some fast power on-off due to the power input jack, since it seems to be sort of loose... maybe an inductive peak?? maybe some high voltage peak got through the 7805A and to the AVR?

I'm using Caps on 7805A output and mega32's VCC and GND ... but maybe the 7805 input cap is essential to avoid an overvoltage that could get through all the rest of the capacitors??

I'd appreciate an experienced opinion of this.. I'm tired of burning AVRs and as far as I have checked the board is alright..

oh, another "fact"... the unit worked alright for about 2:30 hours..

then was turned off for half an our, and when turned on again (having that on-off power "high-speed" transient) stopped working..

thanks to all again

Carlos

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

What else is connected to the ATmega ? Is a schematic available ?
Are the ATmega completely dead or are only certain Port lines etc
damaged ?

I found 7805 regulators and ATmegas quite robust, so I am interested in finding the reasons !

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

You indeed may have had a regulator failure or latchup due to a too fast a rise time
at the regulator input. The internal circuitry cannot respond quickly enough
when suddenly turned on, and excess voltage can be passed through to the
output, frying your circuitry. The input capacitor helps prevent this. I have had
"power driver" devices (5801s) literally explode when their 24 volt supply was
applied without rise time limiting. I had a complete 3.3 volt cpu board get
destroyed when a MAX1776 switching regulator latched up even *with*
substantial input filtering. This part had to be abandoned as an inferior and
unusable design.

Tom Pappano
Tulsa, Oklahoma

Tom Pappano
Tulsa, Oklahoma

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

well thanks to all for the replies!!
Now it's pretty clear, it HAS to be the absence of that *damn* cap at 7805's input.
I'm also thinking of adding a 5.4V zener to protect the mega32's VCC pin along with a diode to prevent voltages under 0V.
That might be enough protection? who knows.. maybe a 8 ohm resistor (if voltage drop is small enough) to let the zener work alright.

Ossi: One mega32 doesn't have the vcc and gnd shorted, but doesn't want to go into programming mode. The 2nd one has a 2ohm resistance between gnd and Vcc-> quite hot power dissipation, won't work. And the third one has 100ohms resistance between Vcc and Gnd, and works a litte but current draw from power source is too high!

Not quite dead thou, but useless...

Thanks again to everyone, I really appreciate it!

Carlos