Photo in earlier thread showed burn between pins 1 and 2 -- that is between input voltage and ground of the 48 to 6V input regulator.
Device is rated for operation with voltages up to 55 volts with an absolute max voltage of 60V. I'm guessing these are taking a heck of a jolt at the 48V input.
They did put a bidirectional transient voltage suppressor (TVS) in there which is rated at 51V but has a clamping voltage of 56 to 62V with max clamp voltage of 84V. They may be trying to clamp down on the voltage but they appear to be using too high of a clamp voltage. Combine this with no current limiting between VBBC and VBBA for when the TVS is triggered and you have a recipe for disaster.

Unfortunately, with the way the board is designed, this may do more harm than good.
The 48V is ultimately coming from their power supply but it has the coils on this voltage. Coils are notorious for spiking their power rail.
I agree with Ken above - the 48V supply should be used by the coils only. They should have provided a clean, lower voltage rail for node boards and other low voltage supplies.

Hard to say where the spikes come from since the bulk of the voltage clipping is the function of the back-EMF diodes.
Based on the photo shown in an earlier photo, obviously there are still spikes.

Pay particular attention to what happens to the 48V line at 3:45 in this video:

If I am not mistaken, Stern changed to a switching supply for the 48V... is this correct?
I have always believed that there is no substitute for big iron (big transformer) for the solenoid loads.