[derefr]

I recently bought a foundue set; the hotplate it comes with has a magsafe regular 110VAC wall cord (no DC adaptation) that's magsafe on the device end. I wondered why I had never seen one of these before. Then I realized that that flat device end is actually incredibly dangerous to the touch when plugged into the wall. It's just slightly less dangerous than potentially knocking over a pot of boiling cheese. For regular appliances that aren't going to give you third-degree burns, it'd probably be preferable to just let the device get yanked off the table, than to risk exposure to live current.

I wonder if, in the future, we could have household wall sockets and extension cords with "sense pins" ala USB PD; where the mains-voltage live "rail" isn't energized unless a device hops on the logic-level control rail and negotiates for it. Then we could truly live in a MagSafe-everything world. (Then again, to block that kind of current, they'd probably need to use relays and other non-solid-state parts, so they might not be the most durable things...)

[dylan604]

This is how every single Apple charging cords with the green/amber light in them has behaved. There's a bit of communication between device and cable before juice flows.

Did the end of the cable not have recessed "pins" specifically to avoid accidental shorting? Seems like it would never be U/L certified without some safety.

[derefr]

> This is how every single Apple charging cords with the green/amber light in them has behaved. There's a bit of communication between device and cable before juice flows.

Yes and no. MagSafe is safe because of the sense current and the handshake; but it's able to do that because it's just not driving very much wattage through the cable; it's low enough that it can direct the power with a simple transistor.

To make mains-voltage AC cords do that, you need a lot more than a transistor. A 1500W current (from e.g. a microwave, or a kettle) can arc a gap much wider than most transistors are printed at. Which is why, in even the most modern smart-home remote light-switch doodads, mains-voltage gets toggled using relays, rather than anything solid-state.

It's lovely to say "just do what MagSafe does", but with electricity, "quantity has a quality all its own" — i.e. very different engineering challenges to overcome.

> Did the end of the cable not have recessed "pins" specifically to avoid accidental shorting?

It has recessed pins (pads, actually), but only barely (by about 5mm); basically to the point that a round metal table leg could make contact with the live pad within the recess.

And, as far as I can tell, there's no sense logic in the cable, either. No click of a relay coming from the cable when it gets connected; and no place for a transformer to live (not that it needs one—it's a hotplate, i.e. a big thick piece of iron you run mains-voltage AC current through.)

There is a relay inside the hotplate itself, which roughly acts as a thermostat (rather than a rheostat) to toggle the coil on and off to keep the cheese at temperature. You can hear it ticking on and off, and an indicator light goes on and off along with it.

But for that logic to work, the cable has to be drawing power to power it. So, AFIACT, the cable itself is always live.