Variac Tripping MCB Issues

You need to wire the neutral back through your MCB. Check the breaker wiring instructions.

I cannot do that as only the live goes through the MCB; only one terminal.

Ouch! You should review AC circuit fundamentals before you assemble multiple parts in a loosely described configuration. I am particularly concerned with your comment that you have mysteriously mitigated an inrush current problem. Soft-start devices are specifically designed to reduce inrush current when a proper load is applied.

A Variac is an autotransformer with a brush wiper instead of discrete voltage taps for the secondary "hot" voltage connection. Not including a neutral connection to the primary side of an autotransformer makes an atypical inductor between the primary voltage source and your load.

From the symptoms, your MCB has a RCD which has not been connected. That is the only explanation for the observations you have shared. The meter would normally not let enough current to pass to trip the RCD, so I suspect your meter is perhaps a 1000 ohms per volt analog meter?

Experimentation is a great way to learn anything. However, mains power line voltages everywhere in the world can be lethal. It is clear to me that you don't know what you are doing. You cannot even clearly describe what you think you are doing. Be safe!

Sorry, but which part have I not described?

Allow me to make it easier for you by asking this question: which parts are wired in parallel and which are wired in series?

If you can't give a helpful response, then don't bother!

You said “The variac is wired in line with a soft-start relay and a 20A mcb.”

In line… like a potentiometer…in line?

with a soft start relay… which relay terminals are connected to … what?

Can you provide a quick schematic sketch of what you have, and what you have done? That should help quite a bit.

On the live side there is an RCD (the neutral also goes through the RCD), then an MCB (live only), then a soft-start relay (live only), then the live side of the variac. On the other side of the variac is the neutral, which also goes to the load. The "Brush" is connected to the other side of the load.

Well, this sounds like nonsense to me

I guess I shouldn’t respond further. Best of luck in your future endeavors.

Your load must take more than 20 amperes, or your MCB is defective. What is your load?

What voltage do you measure at the input hot to the variac? (to variac neutral)? That voltage measurement should also trip the MCB.

We have a minimum of 5 pieces, and I am imaging these strewn across a work surface connected together in some unconfirmed fashion. The jumpers may well be 18 Ga copper covered alum speaker wires with alligator clips squeezed onto the cables. This Variac is stated as 15A, is it not common that this device has an on board 15A fuse or breaker? However, a downstream (I guess) 20A breaker opens first - this does offer validity to your observation we may have a bad MCB here. In addition, this assembly presumably has an electric plug (NEMA 1-15P, 5-20P?) on one end of a power supply cable (a retired and repurposed lamp cord?) this plug pokes into an electric hole is the wall that is itself protected from overload by a MCB of unknown type or rating.

So, yeah… I am out of ideas.

A part of your post regarding the inrush current puzzled me, and I would ask you to clarify this.

In pursuit of my toy O scale model train hobby I have come across a similar issue. The older transformers (before the mid 1970’s) were really a triac, limited to 20vac output. The described trouble was a fairly common issue until the transformer manufacturers started to provide a fixed tap at 14vac, this marketed as a source for accessories… lights for buildings, animated bits around your train layout. This tap was independent of the throttle, and as long as the modeler had some current being drawn from the 14vac tap, the nuisance breaker trips went away.