Issue with voltage divider for DC bus voltage sensing

[XaviPacheco]

In the picture below, you can see a string of 10 resistors used for DC bus voltage sensing. I think that the blue resistors (the original ones) are 400k/1W. The colors are: yellow, black, black, orange. One of the resistors got open, and I replaced it with the one you see different in the string, the one in the red circle. It's a 390k/1W resistor. The equipment worked well for a couple of hours, then the resistor in the white circle got open.

Do you think the new resistor could have affected the string so that a new resistor got damaged? Well, indeed we have a variation of 10k in the array with the addition of a 390k, which leads to increase the current. I did not find the 400k resistors in the market in that package (MELF). What would you recommend?

[duak]

Is the unit a motor drive or induction heater?

What I think is that the first two resistors in the divider chain are subjected to larger AC voltages than the following resistors.  I suspect the replacement resistor failed because it may not be capable of the same AC+DC voltage as the others.

If you look at the circuit board you will see a wide trace on the the other side that may be capacitively coupling to the connections between the resistors.  If this is correct, the other chain is actually under more stress.

I would do the following:
1. if you cannot find an exact or better part, then replace the bad resistor with one from the other end of the string, that is, the lower voltage end.
2. put the part or parts you can find in the now empty spot.  I would also put a 10K resistor in series with the 390K to better match the 400K.  You may have to do this to both resistor strings to ensure they are identical.
3. clean the resistors and circuit board to ensure there is no conductive contamination.