*Strictly speaking, this heater already has a half power setting, but it achieves it by turning off one of the two elements, letting the other run full blast. This results in:
1) A likely shorter element life -- bummer, as this heater is ceiling mounted, facing down, and a pain to take down if need be. Plus, elements are sometimes hard to find and disassembly is "significant".
Best yet, this setup makes it possible for me to run at 1/4 power - great in cool but not cold conditions, and easier on the thermostat contacts for sure.
2) The switch to change power is literally a pull switch -- who knows how long that thing is going to last.
3) Illumination via the elements is NOT particularly desired -- operating both elements at half power is rather than one at full power is less obtrusive.***
**I may add a SMALL 600v capacitor to kill any RF that might be generated.
***I've built a small version of this before: In that case I'm operating a 750 watt 230/240 v quartz heater off 120v, and then doing the rectification bit for the "low" setting. That one heats a chicken house and is so much better than brooder lamps, it isn't funny! The visible light emitted is nil, basically infrared only, so the chickens sleep. Plus, the element will probably outlive me. :-)
Your beat estimate of potential demand plus 50%.
I've also modified to switched series vs. (original) parallel element wiring to run elements "easy" but that takes a bit more of a switch, and I've definitely never seen a pull switch like that...
If you have a few in your parts box, then build it, run it and see if it blows up. It probably wont, but hey, this is the excitement of tinkering of electronics.
Temperature coefficient of nichrome wire is pretty low - not a problem for current changes here. The 6.5 amp theoretical would go to 7 amps - maybe.
Parallel a couple of rectifiers. Put a .2 or .5 ohm resistor in series with each to keep the current division reasonably equal. Watch the wattage on the resistors.
Isn’t the element/elements accessible by removing a back plate and allowing access to remove them?
Silicone makes lousy semiconductors. Pretty good to seal gutters. Look to Mr. Schottky.
Diode voltage drop is only .7 volts which will be pretty much insignificant to the formula power(P) = voltage(E) times current(I). P=E^2/R
I don’t believe that running 2 diodes in parallel is a great idea since the slightest change in characteristics is going to cause one diode to do most of the work.
These days higher amp diodes are not that expensive. I see some 15 A for under 2 dollars.
Since you don’t know the full characteristics of the element then the easiest way to get the actual current would be to set up the circuit and measure the current with a meter. Otherwise just over design and go for worst case and then some.
When in doubt, build it stout. Diodes, capacitors and resistors are cheap. Make it three times bigger than you need.
Just get a Variac then you can get all the control you want.
You could always adapt or build a PWM controller driving an H-bridge, to give you an infinitely variable temperature control.
Maybe I should have called this the “Fireworks Night” thread. And I was off by a day - geez!
BTW you basically will have no CEMF from a heating element. Therefore your reverse voltage only has to be at the peak of the AC line voltage. So for 120 volts RMS you have a Peak of around 170 volts. So a Diode with a 200 volt reverse voltage would handle it.
Even if you plan to use this only when you are running with one element, you still have to make sure you can handle both, unless you build this inside the heater after the switch.