Pic reseting when switching mains current with relay

[lsteele]

Thanks

>> If you didn't change the default config fuse while loading the Bootloader firmware the first time, you can't modify them later in your code.

I know! :-) But I'm quite happy with that fuse configuration right now... unless it's a factor in my problem.

>> Yes the snubber will help!
>>
>>also

>> * Make sure you have connected both PIC VSS pins.
>> * Make sure you have disable the LVP mode
>> * Set all unused i/o to input and connect ALL of them to GND
>> * i would add a 47-220 uF on the PIC VDD line as well (close to the PIC)
>> * Place your PIC away of your relay to see what's happen.

Ok, I'll try all of the above. But I'd like to understand why connecting a 47 Ohm resistor across the relay output (as part of the snubber circuit mat suggests) is a good thing -- surely it'll allow a current to flow all the time. Or have I misunderstood something?

Thanks again...

[mister_e]

Mat suggested a RC snubber, so it's a resistor in Serie with a capacitor, so will flow only for a short period of time.

You may find some explanation in the following document
http://focus.ti.com/lit/an/slup100/slup100.pdf

I would prefer to switch on the AC zero crossing to extend the relay contact lifetime...

OR you may decide to use some SSR... a bunch more expensive...

EDIT: ah crap, i'm late

Also, probably a nice idea to place some 0.1 uF in parrallel with each realy coil + 100-470uF & 0.1uF on their VDD line.

[lsteele]

Thanks for all the suggestions - I'm blown away by how quickly people have responded.

Ok, I've added the snubber and it's immediately eliminated 90% of the resets. I can now turn the relay on and off with the light connected and it works. Strangely I can still produce a reset if I switch the mains on and off at the socket, even when the light is NOT connected. And I've still had one or two random resets.

I'll try connecting up the other VSS pin and adding the 47uF cap to the PIC power supply to see if that gets rid of the remaining resets.

>> * Set all unused i/o to input and connect ALL of them to GND

Is this something I should do whenever dealing with PIC, or only when dealing with mains voltages/currents? What's the rationale here? I'll try the other steps first since this is going to require me to produce a new PCB!

>> Also, probably a nice idea to place some 0.1 uF in parrallel with each realy coil + 100-470uF & 0.1uF on their VDD line.

ok, I'll do this too.

One other question: While reading up on this I've heard lots of talk of ground planes. Is this simply a large area of copper on the board which is tied to ground, or does is have to be an entire layer on the board (impossible in my case since the board is single sided). And how do they help (and indeed could they help in my case)?

Thanks again.

[mat janssen]

Then put also a snubber network on your mainssocket and mainswitch!

[Dave]

lsteele, I notice that your relays are powered by 5 volts. Is the source of the 5 volt supply the same as the PIC? If so I would have a separate supply for the relay source and a separate supply for the PIC. From what I have seen in the past you are working with electromechanical devices that have magnetic coupling and when you are opering the contacts the field is being coupled to the 5 volt supply the micro it is being powered from. If this supply for the relay source was separate and the ground paths are correct I see no problem with the operation of the relays switching the mains.

Dave Purola,
N8NTA

[sayzer]

I meant to say use BOD_OFF.

Most likely!

-----------------------

[shahidali55]

hello ,
I'm having a similar problem.
I have made a 7seg LED clock (PIC16F628 @ 4Mhz).
The clock resets when i switch on the tube light in my room.
The clock and the tube a not linked to one another in any way.
I was using a wall adapter before to power it, but even after i replaced it with a 9 volt battery, the clock still resets when i switch the tube on.
What could be the reason for this ? ? ?