16f877A Keeps on blowing up!!

[kel]

I've blown up almost four 16f887a ICs now.Mark you ,these ICs are very expensive.I've a supply voltage of 12volts from a power converterconnected directly to lm7805 voltage regulator.I put a 10uf capacitor across the output and the ground,.This gives 5volts from the 12volts converter.i use this for powering my 16f877a.But if i leave it on for a long time, the IC does not accept to be reprogrammed anymore although,it will still execute the previous code programmed into it.

Whenever i try to programm it, an error message "Processor is Unrecognizable" is always displayed. Can i discontinue using the power converter and resort to using an expensive means i.e.habitual purchase of 9volts battery.
Can anyone help me out!!!
Cheers..

[Ttelmah]

The chips will only 'die' repeatedly like this, because something is putting a spike into them. Now, the first thing is that on your supply, you mention 10uF capacitors, but make no mention of lower value caps. The first problem is that electrolytic capacitors, though giving large capacitances in small cases, have very poor HF response. Hence a 10uF capacitor, will not significantly attenuate a HF spike. If you look at the manufacturers data sheet, for most 5v regulators, you will find them showing at some point in the circuit, a much smaller capacitor across the rails. This is there to block HF, and is vital. It is not a case of 'bigger is better' here, with the better HF performance of something like a ceramic, or polyester 0.1uF (typically) capacitor being needed. Also depending on what logic is in the circuit, more such capacitors, may need to be placed throughout the circuit, to block spikes atually generated by the fast logic involved (though these are usually only needed for reliable operation, rather than to prevent chip death...).
Ongoing though, the other place where the chip can be spiked, is every single input pin. Connections arriving 'at' the chip, need to be designed to ensure that signals will be inside the ratings for the pin. The same is true, for every signal leaving the chip (though here, the lower impedance presented by the chips driver outputs, simplifies things). If the problem is with an input, then switching to battery power, will not generally help (though the isolation may).
It sounds as if you have a major 'design fault', in the circuitry round your chip.

Best Wishes

[rwyoung]

Another suggestion, add a diode (1N4001, 1N4004, etc) across the output and input terminals of your LM7805. Anode to output, cathode to input. This won't do anything when your regulator is running, diode reverse biased. But when you remove power, some LM7805s which don't have this diode built-in will allow their output voltage to exceed input voltage. This is a bad thing.

Go to National Semiconductor's web site and read their application notes on voltage regulators. Such a seemingly simple device and cause all kinds of problems.

RJ's suggestion to add some smaller caps, one at the regulator and another at each VDD/VSS pair is important.

And finally, some LM7085's (LM340s and other 3-pin regulators, check the datasheet for your exact manufacturer) should have input capacitors too.
_________________
Rob Young
The Screw-Up Fairy may just visit you but he has crashed on my couch for the last month!

[Guest]

suggest you check the 7805 whether the output is still 5V. Some 7805 may be defective.

[dbotkin]

[kel]

Thanks alot guys.it's vital that guys like you are in this forum....
dbotkin mate, ihave nothing connected connected to portB.
cheers.