Great work, Joe, as always on this entire series.
Hioki builds great devices, but this Hioki DMM series continues to have 2 shortcomings.
First, the creepage distances are a bit close [ as Joe points out too] in other areas.
https://www.eevblog.com/forum/testgear/new-hioki-multimeer-and-first-project/msg736814/#msg736814Why is there no NRTL safety rating beyond CE or KCC [ the are not recognized in the USA by OSHA ]? It may be because of what Joe shows.
From the slow motion, where the creepages are too close, added areas of arcing occur. If all the arcing is stopped by the plastic spacer it would be best to reshoot the video showing no added arcing occurs.
The basic Hioki design out of box however, has a fault, and it allows arcs to propagate despite the GDT input protection at the 8kV impulse, which is the CAT III, 600V rating.
You can see the GDT glowing in the videos. the GDT is simply not fast enough for this application.
http://blog.circuitprotection.com/communications/movs-vs-gdts-which-one-do-i-use/It would be better to use both the MOV and GDT at the inputs or MOV alone. The GDT acts to prolong the life of the MOVs particularly in high energy overvoltage.
Recall, this is a test impulse. The real world worse case is that the impulse rides atop high energy AC, such as line voltage, and if the arcing occurs, it provides the pathway for an arc blast.
In this screen grab, one arc extended around the plastic shield from V+ terminal. Another extends from the center GND terminal from one of end of the fuse and from the fuse end to the ground plane. You can also see the GDT glowing.
In this grab, the corona post arcing lights up the fault areas.
The meter surviving the arc is a good thing, but that could be because so much energy is shunted across the inputs nothing appears to the more high impedance DMM circuitry.
However, the point of the protective devices is to protect the user first, not the DMM, for the specific CAT rating of the meter.
The Hioki DT4252 part 2.