Two years ago, I had a very much similar problem with AT89C52. We were making a control unit for a machine and the job of micro was to execute the commands received from PC and control (forward and backward) two 3-phase motors and do some other control stuff.

When the 3-phase power was disconnected, everything worked great for even days. No problem with switching the contactors but the moment 3-phase power was flowing through the contactors, same thing started happening randomly. Sometimes the board was trying to turn all the four contactors on at the same time.

Supervisor was there. Shifting of contactors 10 meters away from micro did not help. Snubers were added but were helpless to completely get rid of the problem. Shifting the phase (to the power supply of micro) did not help either.

Then we started investigating the root cause. Using a 100MHz DSO, it was found that whenever a contactor is released there is a low going edge on +5V supply (with very short time, few nano seconds). The micro usually sustained this brownout but the latches (74HC574 used to drive the optos which in turn used to drive the DC relays) somehow got reset and trying to turn on all the relays at once, while micro was thinking that appropriate contactors are on.

The micro was writing only once to the latches till the next time any change is required in the status of magnetic contactors. And you know what most immediate solution was done? Write the status to latches every 2 msec (this is the time DC relay + magnetic contactor can not turn on) so the latches were refreshed every 2 msec. The problem disappeared completely, client was happy but we were not satisfied. The machine worked with this solution till the project was completely re-designed.

Pathipati Haribabu,

First of all try to find out the actual reason for the hang-up problem then find a way to overcome that.

What is your hardware? How did you construct it? Can you post a schematic and board layouts?