I decided to build a high powered variant of the Hydra for no other reason than to prove the contactor adapter as a means to make it work.

The contactors are Dayton 6GNV0 40A Definite Purpose contactors. They are rated for 40A, but with resistive loads, they can go up to 50A. EMotorWerks sells J1772 cables that are rated for 50A. With 8 gauge wire with 75°C insulation (which is what you get inside 8 gauge SOOW cable), the power wiring is good for 50A as well. The Cooper-Bussmann 16220-2 power distribution module is rated for well over 100A. With all of those components, the Mega-Hydra can charge two cars at 25A or a single car at 50A.

To use it at its full rated capacity, you'll need a circuit breaker and dedicated circuit rated at more than 60 amps. The NEC requires continuous duty applications (like EVSEs) to derate their circuits by 20%, meaning you can only use 50A with a 62.5 amp breaker. However, a better plan would be to use a 60A breaker and set the hydra to use up to 48A. You also will need to hard-wire your Hydra, as the NEMA 6-50 or 14-50 are only rated for 50 amps. That would yield a usable capacity of 40A (so your Hydra would be able to charge two cars at 20 amps or a single car at 40).

In principle, you can make a 40/80 Hydra if you can find an 80A rated contactor and 80A rated J1772 cables and plugs. 6 gauge chassis power cable with 75°C insulation and a 100A rated circuit and breaker.

Of course, all of this equally applies to the OpenEVSE Plus v2 and DIY boards - you would just be using either two single pole contactors or a single two-pole one (personally, I doubt that the benefit of two independent relays is worth the expense - not to mention the weight and volume).

Assembly is virtually identical to the reference design of the Hydra EVSE. One change is in the addition of the contactor adapter boards and the wiring of the contactor coils. That's described in the ContactorAdapter page of this wiki. To mount the contactor adapters, I used 3/8" nylon standoffs and 1/2" 4-40 nuts and bolts to secure them to the mounting plate just above the contactors. Remember that when using the contactor adapter, the inputs are polarized, unlike the relays that they replace. The positive relay terminal for the Hydra is the further terminal away from the pilot terminals. The negative one is the closer of the two (when looking at the display in the correct orientation, positive is the "top" and negative is the "bottom" one). I used solid 22 AWG wire and crimp-on .25" QD terminals. The Dayton contactors have 4 QD terminals below each box-screw terminal, and two QD terminals on each coil input. I made a wire with a QD at both ends to connect one coil terminal to one of the AC line terminals. I made two wires with a QD at only one end for the contactor adapter - one from the other AC line terminal, and one from the other coil terminal. The free ends of each went into the AC side of the contactor adapter. I used stranded red and black 22 AWG wire twisted in a loose braid to go from the DC side to the Hydra logic board.

The physical arrangement of the parts was also slightly different. Since I used the larger case and mounting panel, I had enough horizontal space to mount the distribution module between the two contactors. Because of that, the short #8 wires linking the contactor line sides to the distribution module passed conveniently close to the logic board, so the ammeter coils were placed there rather than on the actual J1772 cables as is traditional. That's ok, though, since the contactor doesn't alter the load characteristics in a meaningful way.

Since this hydra has a higher current limit, the burden resistors on the logic board needed to be reduced. 30 ohm resistors were used, resulting in a CURRENT_SCALE_FACTOR value of 165. The Mega-Hydra also proved to me that the burden resistors of the reference design need to be reduced as well. The speculation is that most EVs use switching supplies in their battery chargers, and their current consumption waveforms are therefore not pure sine waves, but rather something far more "peaky," but with the same RMS consumption. At the high end of the current scale, the reference design with 56 ohm burden resistors reads low because of this (the peaks are being clipped by the 89 mA per count range). Also, the MAXIMUM_OUTLET_CURRENT needed to be increased to 50000. Other than those two changes, the firmware remained identical to that for the reference design, and the operation and feature sets are the same.