Fast blow or Time delay fuses?

[R.G.]

Looks like an Omron G3NE-220TL-US DC5 SSR might do the switching. Non-zero crossing (I'd like to do the timing myself, not have it always go at 0), SCR based, 250A inrush for 1 AC cycle, about $29, Mouser.

[R.G.]

Mmmmm...
General purpose relays get you to 200A inrush for under $5 each, but impose a 20mS max switching time and bounce on top of that, so you have to pull the coil almost three half-cyles early and hope it is consistent in closing-time.

[trobbins]

The paper I recall was able to model the various static and dynamic inductive influences occurring during primary winding inrush on a standalone transformer (no loads), and was able to compare them with measured results.

The hassle with doing diy experiments is to sweep through the whole phase angle range, and then to also do the basic measurements of winding resistance and inductances (at the excitation voltages being used in the tests). Doing a couple of random phase angle tests, that may include contact bounce would require quite a few to be captured to try and get enough statistics to appreciate the likely worst-case level (which is the aim for fuse assessment). Trying to then overlay some form of pre-conditioning on the PT to see how that then impacts the situation is then another difficult variable to set up.

[R.G.]

Yeah - I remember sitting at that bench for several days, accumulating a stack of polaroid scope traces.

That's why I was seriously considering a solid state relay over a hard-contact relay, to have some control of the "make time" delay from the zero crossing. Otherwise, you have to just go statistical on the results.

It's one of those fascinating problems, though.

[Mike Sulzer]

Mmmmm...
General purpose relays get you to 200A inrush for under $5 each, but impose a 20mS max switching time and bounce on top of that, so you have to pull the coil almost three half-cyles early and hope it is consistent in closing-time.

How about a big triac where you enable the switching on one zero crossing so that it fires on the next, then disable the switching to prevent a refire?

[R.G.]

Yeah, there are all kinds of things that might help. An interesting one to test the "remanent field" theory would be a variable turn OFF in a half cycle to see both which polarity makes it best/worst for the next turn on. That's kind of why I went for the SSRs. I was hoping that I could find a MOSFET SSR so I could actually turn it off at a certain point in a half cycle to set up for the next turn on. SSRs come in triac output, SCR output and MOSFET output. The triac is simpler, but a bit more fragile. The SCR is what's used for big, burly SSRs, and the MOSFETs give you on/off control at any time, at the expense of - well, added expense.