PIC freezes

[aaronik19]

Dear Friends, I made a project (RTC) with the DS1337 and PIC18F4550. The code seems to be fine and working good but I am experience a funny thing. When I give supply to the board, nothing appears on the LCD, but when I touch the regulator or a filtering capacitor, it starts to work normally. When I remove my finger from the regulator of capacitor, the program freeze. I replaced the regulator and the capacitor (10uF), but the problem still persist. What could be the problem? I am making this breadboard, but even when I soldered the same circuit on pcb.

Thanks
aaronik19!

[ezflyr]

Hi,

Post your Fuses, and tell us what is connected to the MCLR pin. Also, verify that you have several bypass caps across the supply rails close to the PIC.

Are you using an external crystal, or the internal oscillator?

John

[Ttelmah]

Describe your regulator circuit.
What is the regulator?.
What are the decoupling capacitors, and where are they?.
What is feeding the regulator?.

The 'standard' circuit, is something like a 78L05, with perhaps 10uF on the input (depends on what is feeding it), perhaps 1uF on the output. Then something with good HF response (ceramic or polyester non polarised capacitors), close to the PIC. Then a similar capacitor across the LCD. The supply needs to be decoupled _before_ it gets to the regulator. The capacitor after the regulator are to smooth instantaneous load changes, and prevent oscillation. Big capacitors after the regulator, are potentially dangerous (can kill the regulator). If you need them to cope with large load changes, then you should add a reverse diode across the regulator, to prevent it being destroyed if the incoming supply discharges before the outgoing rail.

Best Wishes

[aaronik19]

Thanks for reply. I have the traco micro switcher similar to http://www.electropages.com/2012/07/rs-low-profile-encapsulated-4w-60w-acdc-power-supplies-stock/

Then I add a simple regulator 7805 (1A) with just one capacitor across GND and VOut (10uF).

I also added couple of 10uF along the +5V and GND of the circuit.

[Ttelmah]

Several comments then:
You don't tell us what voltage the supply you have gives. There are a lot of different versions of this. TMP part number?.
None of the capacitors you use will block HF at all significantly. It is one of those 'bigger in not better' things. Electrolytic capacitors give large capacities in small sizes, but have characteristics which tail off at high frequencies. Study something like a PC motherboard. Group of large electrolytic capacitors round the supply, but then literally dozens of small capacitors across the board. Typically one by just about every IC. Remember these people are after saving every penny they can. These small capacitors are _essential_.
You need an input capacitor to the regulator. Pull the Texas data sheet for the 7805, and look at their example circuits.
Problem is that a voltage regulator is effectively a very high gain amplifier. As such it is inherently quite prone to instability. The capacitors close to the regulator control this. Then the PIC itself produces very high frequency changes in load, encouraging HF noise. The local HF capacitor by the chip controls this.

Best Wishes

[aaronik19]

dear all,

I wired all the circuit on a new BreadBoard and made a regulator 7805 and some capacitors along the supply line +5V and GND but the problem still persisted! I also reduced the crystal to 4MHz instead of 20MHz because I read that 20MHz is too high for BreadBoard.

This is a very funny problem! Photo: [/img]

[aaronik19]

Sorry I forgot..I am using PIC18F4550 and DS1302 RTC

[temtronic]

comments..
1) I don't see any caps from the xtal to gnd.Usually 22pfd.

2) 20MHz xtal /PIC do work on my breadboards(EXP300).

3)try to buy some 10 conductor,22ga solid wire ,it'll really help when wiring and debugging .right now you have mess of yellow wires and that has to be hard to check! Even 4 conductor 'quad'( telephone cable) will help.

hth
jay

[ckielstra]

Sigh....
Several important questions were asked that you didn't answer.

[PCM programmer]

Also read this thread and look at the links in it.
It has notes on how to do a breadboard circuit:
http://www.ccsinfo.com/forum/viewtopic.php?t=49449

[aaronik19]

thanks to all. This is just a quick prototyping not the real design. That's why I made a "mess"

I solved the problem after some more investigation. I also added the 230VAC's Earth to the GND and the circuit worked perfect!

[ezflyr]

Hi,

After all the help you've received, why not be courteous and provide the
"solution" for later forum users who come across this thread? It's much more
satisfying than just a "problem solved" report!

Prototype or not, there are certain things you can/cannot do if you want to
build a reliable circuit. Taking shortcuts while prototyping is not the way you
are going to learn good engineering practices, and if your prototype does not
have good fidelity to the final product, what good is it?

Pardon my skepticism, but without any additional information, and based on
your demonstrated reluctance to follow the advise of more experienced
forum members, my guess is that you've just blundered onto a "solution"
that happens to work in this instance. Unless you can show us otherwise, I
doubt that there is much you've learned from all the helpful advise you've
received! Too bad!

Cheers,

John

[rikotech8]

Might be the MCLR. If you have declared MCLR but not tied to 5V it could be the problem.
_________________
A person who never made a mistake never tried anything new.

[aaronik19]

First I would like to thank to all who advice me and helped me to solve this problem. I made a design which involves PIC18F4550 and the Dallas Semiconductors DS1302 (sorry but in the first thread I mentioned the DS1337 by mistake).

When I completed the code using the DS1302.c, I made the simulation using the IDE software it worked fine. I designed the PCB and assembled all the components. When I power up the board, I experienced problems with the supply. To be 100% safe, I replaced the power components (Regulator 7805 and the two capacitors 10uF), but the problem persisted.

I build everything again on breadbaord and the problem mapped itself and started to investigate. I inserted more 10nF capacitors and also changed the crystal of the circuit from 20MHz to 4MHz to avoid problems with capacitance. Than I was consulting with my friend to connect the 230VAC's earth with the Supply Ground (DC Ground) and the circuit worked perfectly. Than i add a track on the pcb to connect the ground with Earth and everything worked fine. Even the DS1302 is very accurate. When I have some photos I will share them.

[gpsmikey]

While the 10ufd caps are handy for filtering the supply, unfortunately, in the real world, they look more like open circuits at high frequencies. You need some .1 ufd or in that range ceramic caps across the supply lines, the 10 ufd caps and across the supply/ground pins on the chips. If you don't put some bypass caps like that in critical areas, you WILL have problems later on.

mikey
_________________
mikey
-- you can't have too many gadgets or too much disk space !
old engineering saying: 1+1 = 3 for sufficiently large values of 1 or small values of 3