Hello

We have a product in the field that create voltage spike on the SMPS 5V DC output. The product is measuring current and voltage values from a 3 phase motor. A contactor that start and stop tye motor is in close proximity of our product, and every time the contactor switch, voltage spike oucur on the 5V DC output. This 5V output supplies power to the onboard MCU. I measured the 5V output with a scope and confirm the spikes at the time of contactor switching. I set the trigger level to 6V on the scope. I have CAN messages running that indicates when the MCU resets and show me what type of reset it was. Everytime I mwasure a big voltage spike on scope, I can hear and see on the CAN message that the MCU hit either a MCLR or a BOR reset. I cam live with these resets, as it does not affect code flow. But I see now and then that the products behaviour is not acording to what the code is supposed to do. I assume that what happens is the 5V DC goes above the MCU maximum and unwanted events happen in the code.

The problem is there are several products in the field and they cant be recalled. I assume that Radiated EMI gets onto the 110V supply and affect the SMPS outputs.

I need advice on how to minimize or filter noise getting onto the 110V 50Hz supply. I can do the filtering by supplying external filter to all the products in the field.

The SMPS we use is the Myrra 47257 SMPS.
Currently we have no filtering on the 110V AC powerline.

 

 

The Myrra claims to have on-board filtering, so are you sure that the supply is generating the spike, and not that the spike is getting directly from the contactor into the 5V supply.
My experience with contactors suggests that the source of the problem is the contactor coil, not the supply being switched.
Does the contactor coil have a snubber across it? That usually deals with it.
Do you have enough decoupling capacitors on the 5V supply?

Thanks for posting the diagram - are the 12V to 5V convertors linear or switching regulators?