Nick SayerIn North America, power is usually provisioned to a premises as either split phase or three-phase power. In split phase, there are two hot lines and a neutral line, which is connected to an earthed ground at the service entrance. The two hot lines are each 120VAC RMS relative to the neutral, and 180 degrees out of phase with each other. This allows 120 volt circuits to be provisioned with one of the two hot lines, the neutral and the ground, or 240 volt circuits to be provisioned with both of the two hot lines and a ground (and optionally the neutral as well). Split phase service is the norm for residences and small businesses. Larger businesses get three phase power, which is exactly the same, except that there are three hot lines, each 120 degrees apart in phase. Any two hot lines will yield a 208 volt circuit. Most devices that require 240 volts will work acceptably well with 208 volts.
Sometimes you need to temporarily power a 240 volt device. We have this issue because we occasionally visit our parent's house with our electric vehicle, and desire to charge it. While we can charge with a single 120 volt outlet, it takes far longer than if we can obtain 240 volt power. This device allows us to do exactly that.
I must pause at this point and raise the very serious safety concerns that this project brings with it. In general, you cannot expect to find two outlets on different circuits (which must also happen to be from opposite phases) right next to each other on the wall. That means you're likely going to be using extension cords to do this. You must use heavy duty models for this purpose. You must not exceed the rating of the lightest-duty component in the entire system. Failure to heed this warning will very likely start an electrical fire. In addition, the concept of attempting to combine two separate circuits back into one is fraught with all sorts of dangerous traps. Every portion of this design was chosen to ameliorate a particular safety hazard. Do not attempt to construct this device until and unless you understand what the design is attempting to achieve and do not attempt to make any changes to the design without fully understanding the ramifications.
In general, doing this is actually a really bad idea. I can't recommend anyone attempt to do this without thinking it through. The only reason I am documenting this project is in the hopes that anyone who does try to achieve this won't fall into some of the very dangerous traps that are endemic to the concept.
The design
Follow along with the schematic as you read this design description. Note that the schematic does not completely represent how the circuit is actually constructed, as the high-current path must be made with large gauge wire while the relay coils and pilot lamps can use smaller wire.
There are three sections of the design, each centered around a DPST relay. Two of the relays have 120VAC coils and the third has a 208/240VAC coil. Each of the first two is connected to a 120V plug, such that the hot and neutral directly activate the coil. The hot line of each plug goes into one pole of the relay. The two relays form a sort of "X" figure. The idea is that each hot line must be switched on by both relays. If either relay turns off, then both hot lines will be disconnected from what follows. So one of the hot lines is switched by the first relay, and the output of that switch goes to the opposite relay where it is switched again. The other relay has the exact same setup - its hot line is switched by one of its own poles, and then fed through the unused pole on the opposite relay. Along the way, the two primary relays have neon indicators across their coils. This allows you to see at a glance if the device is functioning and if not, why not.
The purpose of these two relays is to insure that nothing is connected to the output unless both source circuits are energized. One of the hazards of this concept is that if you yank one of the 120 volt plugs of a naively designed version of this box, a high voltage would be present on the prong of the plug. This is because a circuit would exist from the opposite phase's hot line through the target device and out to the plug. The primary relays prevent this sort of thing from happening.
The output of the two primary relays goes into the secondary relay. This one has a 208/240V coil, and its two poles switch the power on to the output. On the output side of the relay, there is a third neon indicator that provides a visual indication that power is present.
This secondary relay's purpose is to only allow the power to flow to the output if it is at 208/240 volts. If the two input circuits are not on different phases, there will be no voltage potential between them and the relay will not close. This prevents a voltage difference between either hot line on the output and ground, which would be the case if this relay was not present. In general, the target device would not work, but its wiring would still be hazardous, as it would be 120 volts away from ground. The goal of this design, once again, is to insure that only proper power is supplied to the target, or nothing is supplied at all.
Another hazard that this relay avoids is that if one of the primary circuits is miswired such that the hot and neutral are exchanged, then the primary relays would still pass power through, but it would be 120 volts. Allowing this to flow to a device expecting 240 volts might damage it. When this happens, the secondary relay makes an angry buzzing noise, but does not engage.
One commenter has noticed that there is a potential design issue that's worth bringing up. If you switch off one of the primary hot lines without disconnecting the related neutral line, then there is a complete circuit to supply power to that relay from its own neutral, through the primary coil, through the 240V coil and to the opposite hot. If both primaries are engaged, then switching off a single hot line (rather than disconnecting the cable, which would cut off both the hot and neutral), would not disconnect all of the power until one of the primary relays is opened. The 240V coil's DC resistance is 3.8kΩ and the 120V coils are 950Ω. If you compute the voltage drop considering both coils as a voltage divider, you wind up with 24 volts across the "dead" coil, which exceeds its release voltage spec of 12 volts. I haven't seen this problem because I have AC voltage indicator LED modules across each primary coil. My speculation is that they represent enough of a load to starve the primary coil when it's hot (and only it's hot) is cut. This may be because they're LEDs, but specified to operate over 120-240VAC. This suggests there is a current regulator built in to allow them to operate over such a wide range. That current regulator may allow it to draw sufficient current in the scenario I've outlined to starve the coil and let it drop out. And, of course, once one primary drops out, then both phases are disconnected from the 240V coil (and the output).
A note about the neutral
Do not attempt to supply a neutral to the output!
This device is intended to supply a hot-hot-ground receptacle only.
In general, you cannot make any assumptions about the incoming neutral lines at all. If you try to join them together and feed them to the output, then if you encounter a hot-neutral reversed 120V outlet, you will have created a short circuit.
Operating notes
The device should be silent in operation, except for the relays clicking once when connected.
This device is utterly incompatible with GFCI protected outlets or circuits. The 240 volt relay coil will draw enough current to instantly trip a GFCI. GFCI's work by sensing when any current that leaves the outlet on the hot line doesn't then arrive on the neutral line. This device takes current from one hot line and returns it to the other phase via another hot line.
If you hear a loud buzzing noise when the second hot line is connected, it means that one of your circuits is miswired with a hot-neutral reversal. Do not leave the device connected this way for an extended period, as it's not good for the 240 volt relay coil. Disconnect the power and either use a different outlet or fix the broken one.
Do not use this device on ungrounded outlets or with "cheater plugs." Not providing 240 volt equipment with a proper earth ground is a bad idea.
While the recommended case starts out as water resistant, the construction procedure drills large holes and places non-water resistant components in them. Consequently, this device must be considered unfit for outdoor use.
If either of the two 120 volt circuits powering the device becomes de-energized (if the circuit breaker pops or the fuse blows or if it's unplugged), then the output will be shut off. Two of the pilot lights will go out - the output light and one of the two input lights. The dark input light will tell you which of the two circuits needs to be reset.
Note that the amount of current you will be able to pull from this device will be limited by a number of factors. A lot of interest has come about lately in 240 volt circuits because of electric vehicles. Remember that EVSEs (car chargers) are continuous-duty devices. That means that you must derate them by 20%. The beefiest heavy-duty extension cords you are likely to find for 120 volts will be rated at 15 amps. Derating that by 20% yields 12 amps. Keep in mind as well, that unless you're using a circuit with only this device on it, you will be sharing the available current. This also suggests that a derating may be in order, but the magnitude of it is up to you to determine.
The absolute maximum current of this device as designed is 24A. The relays are rated for 30A, but the QD terminals are not, and they represent the limiting factor. Note that as designed, with standard household 120V plugs, you would only ever be able to use the device with 20A circuits anyway, and you could only count on 20A if the circuit (actually both of them) was dedicated.