Solution for charging port reliability on high-vibration e-moped?

[eweichat]

Built a 72V pack for my commuter, but the charging port is driving me nuts. The standard GX16 just isn't cutting it—the vibration keeps loosening the pins, and after a rainy ride, I found signs of arcing inside the connector.
I’m looking for a "plug-and-forget" upgrade. Thinking about a bayonet-style IP68 connector (maybe an RJ45 or USB-type for the smart BMS comms and a beefier circular one for power).
Does anyone have a source for compact Inline IP68 connectors that can handle 10AWG wire? I need something that can take a beating on potholes but stays bone-dry inside. Any specific industrial series that has worked for your e-bike/e-moto builds?

 

[99t4]

Does anyone have a source for compact Inline IP68 connectors that can handle 10AWG wire?

10 AWG? How many amps are you charging at?

 

[lnanek]

Maybe just fill the back of the connector with potting compound? I take apart every controller/battery/battery combiner whatever I get just to see what's inside and half of them are solid rubber/epoxy inside covering all the PCB, components, and connectors.

 

[Apartunis]

Reliable charge connectors only use two wires. Anderson Powerpoles, XT whatevers, SAE connectors, some faraway country's mains plug. The basic problem has been solved many times. If you have preconceptions about how it must look, or whether it really needs to transmit unnecessary data? Well.

 

[lnanek]

It is kind of weird to need more than 2 from a pack. Most people who aren't trying for proprietary lock in just get a Bluetooth BMS if they really want more data out.

That said, I never had issues with the old mixed D-sub 7W2 connector on the Bionx packs. Maybe you could even get some with screws on the sides to hold them together. I think that tops out at 40A for the power pins, though. I saw an XT60i with a third pin recently too.

Arcing on the charge port is odd. On discharge it's common if the connection is being made physically without an on/off switch actuated after and capacitors are being filled immediately. There's anti-spark connectors that engage a resistor first when pressed in halfway, and only full contact when pushed in the rest of the way for that, though. Weird it could happen on charge pins. My charger sends some test voltages a few ms every few seconds until an actual battery is connected and it can start CC/CV, but doesn't dump a bunch in blindly.

Maybe in addition to potting the inside you could make a dummy plug for the outside. Just a male connector with nothing but rubber in the back that plugs in to the female and supports it while riding, then gets unplugged for charging.