Sonsivri

My question: when i touch fingers in my circuits Pics 18F based micro goes crazy and give some spykes,i  tried with pull ups 10k but it´s the same problem; and reset, Mclr 10 K to VCC some times not work, micro no start software after power on.
Any triks about this, to industrial aplication it´s impossible to work like this.

Regards

Don't touch your circuit board.

Also make sure you have decoupling capacitors fitted from VDD to ground on your pic.

to add to this. you don't say what your ciruit board is. etched pcb, vero/strip board, proto board, or what?.
add to this your layout and your method of powering the board, mains power or battery. the list is endless.

you connect a .01 capacitor between VCC and ground pins as near as possible to the chip, make more ground plains in PCB. place ur crystal near to pic


makri

Power suply from mains transformer voltage regulator and all 100nF decoupling capacitor including power and near micro ,using normal soldering board.I thing it´s because  impedeance in Pic micro ports, not  problem avr nor 8051.

Until you show us your code, IMHO, all recommendation is pretty pointless. I can assure you PICs are used in industry applications, like motor control, etc.

Did you left some pins free ?!!!!

now it´s better, it was very high clock, by now i saw some industrial electronics using pics. But it´s necessary to prevent high frequency for a specific application. If we can do it lower, better.

Not really. What is your high frequency and low frequency? I have used 20MHz and 40MHz with PIC 273, 245, 452, 4550... You said your circuit some times will not start. I think the problem may associated with the capacitors you used with your crystal. I use 18pF to 22pF and they are good. Also I used 56K and 1uF for the MCLR; you may want to have the PIC be stable with the voltage before MCLR goe HI to start your program.

So, please look at:
1. MCLR (resistor and capacitor)
2. Two cap used with the crystal
3. At program start up, use a little delay before setup port-pins
4. Pure DC for power supply. Check your supply (5 Volts or whatever voltage you use) to see if it is clean.
5. Put more than few decoupe caps around the circuit, specially the PIC and other IC (.1uF OK)

You also want to check your other component that may draw too much current when its active. It is good ideal to post your circuit and state detail your problems so other can help. What I said here is only a guess work, I don't know your circuit or what is the environment you are using; so I may not help you at all.

For the problem that when thePIC stops working when you touch the circuit. It may associate with the value of the capacitors connected to the crystal.

Want more help, post your circuit and state your problems in detail.

Good luck,

Tom

In my experience of circuit board implementation, noise immunity is nearly always down to board layout, component selection and electrical design. I have read many times people saying that PICs perform better than AVR in noisy environment and vice versa, but the truth is that with proper component selection, component positioning, board layout rules i.e solid analog/digital engineering design experience; there is no difference.

I had some similar problems...

0. Check with an oscilloscope where the noise occurs.
1. Check the DC source.
2. Check the values of capacitors used in the crystal.
3. In program, use an delay before configure pins (I/O).
4. Put an capacitor the near (between VDC and GND) the PIC (Value recomended is: 100nF)
5. Checks if there is any moisture on the board.
6. Put an GND to metal cap of crystal. (Solder)
7. At the PCB Board, give greater preference to GND, trying to cover the entire layout remainder.
8. It´s recomended the crystal near PIC.

Look:
(http://www.cvel.clemson.edu/emc/tutorials/tutorial-d01.gif)


Visit this sites:
http://www.cvel.clemson.edu/emc/tutorials/decoupling01.html
http://www.cvel.clemson.edu/emc/tutorials/decoupling02.html
http://www.cvel.clemson.edu/emc/tutorials/decoupling03.html

http://www.interfacebus.com/Design_Capacitors.html

Documents:
http://hsi.web.cern.ch/HSI/s-link/devices/g-ldc/decouple.pdf

I hope I have helped him.
Good luck.