Hello Victorf57
All the used inputs are with optocopller
I wiil tomorrow try with a zero cross routine
Best regards
Pedro
Hello Victorf57
All the used inputs are with optocopller
I wiil tomorrow try with a zero cross routine
Best regards
Pedro
If you are driving a relay which in turn drives a contactor you are probably switching some serious currents. Contactors are notorious for generating RFI because, being mechanical, they make or break anywhere in the mains cycle and can give spikes of thousands of volts if they break an inductive load during a current maxima.
If at all possible replace the contactors with AC solid state relays (SSR). These can be driven directly by the PIC and most SSRs have their own zero crossing logic inbuilt to make sure they only make during a zero voltage crossing or break during a zero current crossing.
If you are stuck with the noisy contactors then you will need ferrite beads and noise supressors on all power and signal lines and an earthed or shielded metal box to house the processor board.
Take special care with off board interconnections such as the leads to any LCD display or front panel I/O switches as these are great little antennae for drawing ambient RFI into your processor board. I have used some woven metal tubular sheathing from WARTH International Ltd to shield I/O leads with great success. Farnell 150-840 is one variant you might look at.
Use a filtered power inlet connector, not just a simple fused IEC inlet and use plenty of 100 uH or larger chokes on the power lines. Make sure the main electrolytic power storage capacitors have ceramic 100 nF capacitors across them. The series impedance of an electrolytic capacitor can be quite high and they provide no filtering above 1 MHz or so.
Keep trying, even all the ways that you think cannot make a difference, as RFI is elusive stuff and is not easily cured by any formula or simple 'capacitor here will fix it' solution.
Good luck.
BrianT
Bookmarks