Burning PIC16F84A

Nicolae Fieraru schrieb:

That sounds interesting. But sorry, that I can't scry, my crystal sphere will be serviced this time. ;-) So please tell us more about the circuit; AACircuit is very helpful (look at

I think, PICs are very robust chips. I maltreated them many times, placed them 180 degres distorted in the sockets, connected low outputs to Vdd and so one, without any loss.

AFAIK there are no common "the failure" of these chips. Mostly, the problem is between monitor and keyboard, you know!?

Sory about my jokes :-) Feel free to ask me more concrete. Michael

Suggest you post a link to a schematic (or post the schematic in gif or pdf format in the newsgroup alt.binaries.schematics.electronic).

PICs are not particularly sensitive to ESD or other mild abuse.

Just one suggestion comes to mind- what is your exact power supply voltage? I hope you are not using an unregulated linear "wall-wart" type of power supply, because they often have significantly higher than rated voltage when not loaded down to capacity. Exceeding the maximum Vdd voltage will most certainly kill them dead.

Best regards, Spehro Pefhany

--
"it's the network..."                          "The Journey is the reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com

"Nicolae Fieraru" schreef in bericht news:431c23b0$ snipped-for-privacy@duster.adelaide.on.net...

Without any circuit it's difficult to say. Check your DC-connection and voltage level, with a good meter (fluke comes into mind). I burned during my many fiddling sessions only one PIC, I misplaced it in my breadboard. PIC's are normally very well suited for amateurs and small series production because they stay long in production (or a manufacture alternative is given), can easily be replaced by a chip with more capebilities of the same series PIC 16F with another 16F (code compatible) and last but not least the protection of the chip. I have not heared of any chip broken easily. Only with real missuse.

Do you use the right X-tal (frequency) for the chip since this one is also broken???

- all the outputs appear to the programmer as the chip is erased. Explain, Microchip uses only a few pins for programming the other pins are never checked! Most of the times these outputs are on chip-reset input because of high-impedance and with that some fail-safe.

Why don't you use an ULN2003 or similar for this, you don't need the buffer and clamping diodes.

Alexander

Hey - PICs come very cheap. Just register via the Microchip website, and they will send you sample chips for free. They even pay for postage.

...

... When they blow, the oscillator doesn't work

You should have your crystal ball serviced then :P The circuit is from the book of John Iovine and I will upload the image of the diagram to

I didn't use the 4k7 resistors that I drew on the diagram, although I find a bit strange driving the Darlington transistors straight from the CMOS circuit. The voltage I used for the whole circuit was 5V from a lab power supply (stabilized and current limited to around 1Amp at the time). I have used this power supply in other instances and I don't think it could cause problems. Anyway, I will build another power supply with a 7805, just to be sure. Instead of TIP120 I used BD681 and the stepper motor I have no idea what model is, it is unipolar or bipolar and what is the voltage for it. The protection diodes I used initially were 1N4004, then I changed to 1N4148 (I thought maybe the first ones were too slow and voltage spikes return from motor). The whole purpose of this circuit is for me to learn about microcontrollers, to identify the voltageof my stepper motor (I would have increased it once the circuit was fine), one of which wasn't how to blow PICs... Anyway, as I was saying, I blew the last PIC16F84A trying to run leds instead of the motor, this rules out the motor.

This is what puzzles me so bad. During the time I had all kind of chips with shorts between pins, or connected to the wrong points, or anywhing else, and they didn't die straight away, most of the times, after solving the circuit errors, they were still working. In this case I am unable to identify any mistake in the circuit (I used wrapping wire to connect the connections). I don't exclude I might have some errors, although I was able to drive the leds properly straight from the output of the 4050, therefore there isn't a lot to be wrong. I was measuring the voltage and it was set to 5V (accidentally I raised it once to 5.5V, which is still under 6V limit). I use two Fluke multimeters, and they work fine. The circuit is PIC16F84A/20MHZ and I run a crystal of 4MHz, I really don't believe this would damage the chip.

I know the problem is somewhere out there or in here, but need to find out more exactly where. Do you recommend a psychiatric exam? :-)

Thanks for the help. Nick

Done that, on

This is where the mystery lies.

I am aware of that, and I used a stabilized lab power supply, monitoring the voltage with a good Fluke voltmeter.

Exceeding the

I only went to 5.5V for about 5 seconds with one of the chips, and supplied all four chips with 5V. I distroyed the chips over a few days, and after careful check of my board, replacing the slow diodes 1N4004 with 1N4148 (who cares if the diodes or transistors blow up) and even not using the stepper motor anymore. The result of all these changes was another blown PIC.

Thanks for support, Nick

You can check

Please ignore the hand drawn resistors, it was only a test.

I really can't explain the kind of missuse I've done. Hopefully I will understand before I blow too many of them.

The crystal wasn't broken, it is still functional. It is a 4MHz crystal, running a chip rated 20MHz. I was saying that once the PIC blows (it doesn't even heat itself), not even the oscillator works anymore (I am checking with an oscilloscope).

Once the PICs stopped working, I put them back on the programmer, and I tried to read them, then to erase them and to reprogram them. The programmer reads them as they are blank (3FFF or so).

The hex file burned to PICs takes care of configuring the ports (I copied the code from the book called PIC Microcontroller Project Book, of John Iovine). Even if the code is wrong, I expect to be able to erase the chip and reprogram them again, which doesn't happen. I can't erase them anymore.

Because I don't really need to control the stepper motor, this should have been only an experiment to familiarize myself with the PIC microcontroller, and programming them. It was meant to be only the initial try, after which I would have gone into the stuff I am interested about. Unfortunately, after a week, I am still at the stage of blowing chips without understanding why. I will find out eventually.

Regards, Nicolae

Okay, several things here...

1) You need a power supply filter capacitor (t the output of the bridge rectifier and input of the 7805. It will have to be large (like maybe 1000uF or 4700uF depending on the stepper current. The minimum value can be easily calculated (see any basic level text). 2) 6VAC is insufficient to reliably get 5V out of a 7805-- it will be marginal, even with a large capacitor. 10VAC is more like it. 3) Those 1N514? diodes are not doing anything. They need to be across the motor coils.

Best regards, Spehro Pefhany

--
"it's the network..."                          "The Journey is the reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com

I agree with you, the guy forgot to place the electrolytical cap after the bridge. I didn't even look at the power supply module, since I used my own lab power supply, adjusted to 5V.

Yes, the input voltage to the 7805 should be at least 2.5V...3V more than the output. Considering the voltage drop on the 2 diodes (1.4V), from the initial 6V * 1.414V - 1.4V = 7.08V It is a bit low to supply the 7805.

I know that relays have the protecting diode across their coils. These diodes protect the transistors, but the spikes could affect the +5V that go to the rest of the circuit as well. I think you are right about placing the diodes across the motor coils (cathode to +5V and anode to collectors). I suspect the author actually haven't actually tested the diagram, he just made the design. Thank you very much for your insight.

Best regards, Nicolae

They won't do anything unless they are zeners.

Best regards, Spehro Pefhany

--
"it's the network..."                          "The Journey is the reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com

Hi Nicolae,

unfortunately my crystal shpere isn't back, so we must use our brain the next time ;-)

Here are some questions and ideas (looking something curious).

Did you test the PIC in the circuit without the whole other things (4050 ...) and some LEDs on RB0-3 instead?

Check the board again, if nothing found, from a friend instead or do any other for some time. (From my own practice I know, that if someone looks on a problematic thing more and more, it looks more and more error-free. ;-) )

To your comment from another post, that you can't reprogramm the PICs. Did you enable any code protection on the PIC? The text in the programming manual sounds like there is no way back, if you do so. But I never used these old PICs, so I can't say something more about.

Sorry, but in this situation I can't say do this or that.

Michael

Hi Michael,

I haven't enabled the protection on them., because I've been able to erase them and reprogram them before I burnt them. It was just a hope that maybe they just lost the information and I will be able to erase and program them again, but obviously they died.

I will rebuild another board and test it without the motor (just with some leds), so that at least I will split the problem apart, in case they blow again.

Thanks for your support Nicolae

Thanks for the tip :-)

Sorry Spehro, but I don't agree in here. The problem when you have an inductive load in a circuit (as a stepper motor in the circuit) is that the coils generate a high voltage with REVERSE polarity in the moment the current is intrerupted through the coil. Because the diodes are placed in reverse polarization on the coils (cathode to the VCC), they don't need to be Zener diodes in order to protect the whole circuit (including transistors), as all the voltage spike will be shorted by the diode.

Regards, Nicolae

Hi, Nicolae:-

VCC + | | C| | C| | i C| v | | | | | Vc |/ -| |>

| | === GND

Consider this circuit (view in fixed pitch font). Assume we are in steady state, the transistor is in saturation, and a constant current i is flowing through the coil (limited by the series resistance of the coil, which I do not show).

As the transistor turns off, the voltage Vc will start to *rise*, because the *current* continues to flow in the *same* direction through the inductance. It will rise above Vcc (thus reversing the voltage across the coil, perhaps what you are thinking of), and is limited only by the parasitic capacitance of the coil and the breakdown voltage of the transistor.

Best regards, Spehro Pefhany

--
"it's the network..."                          "The Journey is the reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com

Hi Spehro,

As you say, the current will cause the potential in VC to become greater than VCC when we turn off the transistor. And if we connect a diode between VCC and VC, with the cathode to VCC, that diode will absorb all the current generated when the transistor is turned off, therefore it will protect the transistor from receiving a potential on his collector higher than VCC. And the diode doesn't need to be Zener. Isn't that right?

Regards, Nicolae

Yes, that's correct, however the schematic:

.. shows the 1N514 diodes connected across the transistors, where they would do nothing unless they avalanched. The 'transistors' are actually 60V/5A darlingtons, and have parasitic diodes in the same orientation anyhow.

Best regards, Spehro Pefhany

--
"it's the network..."                          "The Journey is the reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com

Yes, I agreed with you a couple of posts ago that the diodes as shown in the schematic don't protect the circuit. I had a look on the pdf file of the transistors and yes, the diodes are redundant anyway. As I was saying, I suspect the author didn't actually built the circuit, just made the design. Anyway even if he built it, depending on the stepper motor and transistors, the overvoltage generated by his motor maybe wasn't high enough to destroy the transistors. I was doing some tests in the past, and I discharged very large sparks (about 2 cm in lenght) generated by an automotive induction coil to a LED, in both polarisations and I couldn't to destroy it! And I had tens of thousands of volts applied in reverse polarisation, so it isn't always easy to destroy an electronic component. Except for the PIC16F64A :P

Thank you very much for your help.

Nicolae

PS. I checked your web site and I read the article about the MOSFET switch for ac/dc conversion, it is a very interesting idea. Congratulations :-)

I had a look at the circuit, he must have done some bad soldering or hooked it up wrong. Instead of using 4050's and transistors I'll use either a ucn2003 or 2803. I've built projects from the same book with no trouble. Chewy

going

managed

driving

work

to

I

ic's