16f629

[SeanH]

Hi

I am using a 12F629 chip for a project and got a strange problem.

I am using GPIO.0 and GPIO.2 connected to a FTDI USB>RS232 converter chip.
The transmit is conected to GPIO.0 and receive connected to GPIO.2 pin, these pins USB side idle high.
Placing a meter on these two pins they read 3.3V with or without 12F629 in circuit.

I am using GPIO.3 as a Input (Internal MCLR being used) that monitors a button and gets driven high when pressed, External 20 mhz xtal and VDD is supplied via a SP switch.

If the unit is turned on and the button is held down then the chip will jump to a routine.

Now with the switch off I would expect the Chips VDD input to be 0Volts but it is actually 3.3Volts and the chip will never reset. If I place a double pole switch in and isolate the GND and VDD then obviously it works.

How come the Chip is drawing Power from these pins and keeping the Chip running without the VDD line having any supply????

I popped a 18F252 chip in it's place and tried it on the RB ports and when the SP switch was turned off it did indeed power down the chip and the 3.3V on the RB pins did not supply power to the chip.

Just curious on any reason why this happens with the 12F629 chips?

Regards

Sean.

[Dwayne]

Hello Sean,

Sean>>Just curious on any reason why this happens with the 12F629 chips<<

Yes, you are correct... The 675's do the same thing. They run on low voltage, and extremely little current. You have to tie your leg to ground, so that the chip will not run off of that pins. Or you may get a chip with a mind of its own. <g>. Twi-Light zone?

I discovered it, when my chip would not reset when I pulled the VCC. Though I still had 5V on another pin, the chip never reset itself. <g>.

I think you will find it will happen on a quite a few chips. My guess would be most of them, depending on what pins are used for etc.

Dwayne

[SeanH]

Just had a thought, I am using an external Xtal so may just use a 74hc04 buffer and a couple of resistors to create the clock toggle and feed it into OSC1 pin then that frees up GPIO.4 for the input of the button.

Then I can set MCLR to External and use GPIO.3 to reset Pic.

Thx for the feedback Dwayne and glad it is just not me visitng the twighlight zone!

Regards

Sean.

[Melanie]

Many of Microchips new offerings are 'micro-power'... that's mico-amps... ie they run on a drop of Lemon juice and two dissimilar metals (try it!)... I've powered an LCD and PIC with a potatoe, however, despite there being a ready and cheap supply on the high-street, they don't come in handy AA size uniform shapes to fit the Battery compartment. PICs can even start running in your programmer during the verify cycle - I've posted a report and warning about it in this very forum.

All the I/O's have protection diodes (go look at the PIC block schematics) and it's through those Diodes that your PIC is being powered.

It is BAD PRACTICE to design any microprocessor circuitry where in the application there is power appearing on the pins but not on Vdd. Redesign the way you approach your application to avoid this.