I am agitated about not being able to run the A/C from a 3500 W generator. Why doesn’t that work? Well, because all generators that have a 220 capability split the power betwen the two 110 V outlets/circuits. I’ve seen a few recent generators that have a switch with two positions — (1) keep the 220, or (2) all power to 110.
You can do this, too, if you’re really adventurous. Here’s the before (top) and after (bottom) wiring diragram for my generator:
In order to completely understand what’s going on here, the generator has two windings, each providing 110 volts. If they are wired in series and 180 degrees out of phase, you get 220 volts across them and still can have 110 across each winding, with a common neutral. That’s the top schematic.
But if you can wire the windings in parallel (and in phase), you lose the 220 capability, but you can now get full power at 110.
Forutnately, my generator neutral wire was easily accessible on top of the wire bundle on the windings. Cutting a bit of the linen and lacquer freed up the wire (be extremely cautious not to nick
the insulation on the windings (it’s just lacquer) or cut a wire by accident!!!). Cutting one winding away from the neutral and providing a fourth wire from this winding out to the sockets was all that was needed to provide full power at 110V. Here’s the business end of the generator with the rotor
support frame removed:
Here’s the two windings soldered to the neutral wire. Just cut one of the windings away and attach a fourth lead to the lose winding.
I retained the separate 110 circuits and wired the existing 220 connector so that you need a pigtail to short the two 110 circuits together.
I’m not sure that retaining the two independent 110 V circuits is of any value, but if the pigtail is not plugged in, the electrical characteristics of the generator are unchanged, except you don’t have 220 volts available any longer. The existing circuit breakers still function to independently limit the current in each winding. If you happen to screw up the phasing of the windings, the circuit breakers will tell you right away!
Yay! Now I can run my A/C no problem. I never did use the 220 capability anyway. (took about 3 hours, including missteps…)
I thought I’d better insert a a remark about paralleling generators. Honda can do it because their generators are really DC with a high power inverter. The inverters can be kept in perfect phase by tieing the inverter clocks together, even though the generator RPM and DC power output aren’t necessarily matched.
In most generators, however, the frequency of the AC depends directly on generator RPM, which in small generators is not very stable, nor is it locked to 60 hertz. So you can’t tie two non-inverter generators together because they would not be in phase and you’d pop circuit breakers, like immediately. (BARTS says you can parallel these generators–they get in phase because of the mechanical forces in the windings, which which imparts torque that forces their RPM to match. I would be hesitant to try this.)
In my example above, not only are the windings always in perfect phase, but they are really just one winding with four wires (which can be split into two windings 180 degrees out of phase to get 220V), so the voltage output is also essentially identical. Consequently, paralleling the windings is no big
deal, electrically speaking.
The final design is a bit modified from the above description. I decided it wasn’t safe to have a standard connector that really was going to perform the shorting plug function. So a standard 110V, 30A connector is installed on the genset. Now, when you plug in you get power from just one winding or, when you close the switch, you get it from both windings. This allows the two 15A connectors to function on separate circuits with separate circuit breakers.
I tested this by running the old Armstrong AC and found less than 1 volt voltage drop. Happy, Happy. Burning man dust storms, take that. Pow. Wham. Zap.
I looked at the thread and realized I didn’t put any data in regarding how closely matched the windings were. It turns out the voltage was less than 1/2 volt and the resistance of the windings were near two ohms. So the circulating current is on the order of 1/4 Amp, or 25 Watts. That amount of heat can be easily dissipated by a small generator and is way less than what is being dissipated
when there is an external load of 3,000 Watts (500 Watts).