Why would battery connector melt?

My new build has been running smoothly for about 3 months now, but on the way into work today the bike intermittently lost power, the voltage indicator showing a temporary (2 second) drop in voltage.

I made it into work and had a quick look around the bike and it appears the problem is from a dodgy connection - my layout is as below:

BATTERY---BREAKER---SHUNT---CONTROLLER

There is an XT90 connector between the breaker and the shunt, the positive bullet within the connector has actually melted the yellow plastic and separated from it (the yellow has turned black!). The negative is completely fine.

Is this just down to a poor connection/solder do you think?

Thanks
Jon

Probably. If there's high enough resistance, and current flow thru it is high enough, there'll be a lot of heat.

If the connection point is not in open air, all that heat is trapped inside the heat-generating point, and can quickly or eventually cause a damaged shell. Depending on the connector design (like with Anderson PP's), ;the damage can actually make the problem worse (sometimes rapidly) because it makes the connection even worse, even higher resistance--sometimes even causing disconnect and/or arcing (which coudl cause a fire)


FWIW, I just recently went thru SB Cruiser's wiring, and removed all connectors except thsoe on the battery itself. Everything else is now directly connected. This greatly decreased voltage sag under load.

(exception is presently phase wires on mtoor/cotnroller because that's still an experiment in progress).

I understand direct connection from a performance and reliability point of view, but wouldn't this create a nightmare regarding maintenance?

I might give that a go... I don't fancy breaking down half way home... 13 miles with only one gear and a 40kg bike doesn't sound too good. I might hard wire it on my lunch break.

Just get a better solder joint on the xtc. When connecting main battery wires to controllers, there is usually a spark. So that could have caused the connector itself to start making poor contact.

You can do the anti spark connector thing. But for max reliability, you might consider dual plugs. a y and two plugs. Make one take the spark all the time, leaving the other cherry. And at half the load each, not near as likely to ever get warm. You'll always have one that works.

For field repair, carry some wire nuts and wire. If you must, cut out the bad, and splice in new to get home.

jonnybump said:

There is an XT90 connector between the breaker and the shunt, the positive bullet within the connector has actually melted the yellow plastic and separated from it (the yellow has turned black!). The negative is completely fine.

Is this just down to a poor connection/solder do you think?

Click to expand...

Unless you are pulling more that 90amps (which the XT90 stands for) it means the connector was dodgy.
You could go the next level which is a 150Amp bullet connector, the XT150 nothing beats bullet connectors.
https://hobbyking.com/en_us/xt150-connectors-w-6mm-gold-connectors-red-black-5pairs-bag.html

It was more than likely just a bad connection from the start, which led to minor arcing across it, which burned some metal away, and made the connection even worse, leading to more arcing, etc. You probably could have discovered it before it failed by feeling each plug after a hard run, I'd bet it was running noticeably hot the whole time. This is a simple test I do everytime I make/break any of my major connections, though I usually use an infrared thermometer. A visual inspection of the pins/sockets will also usually show at least some discoloration and/or pitting in the hot connection.

The connectors have an amperage rating, but that's based on a clean, tight fit. If it's dirty or loose that rating no longer applies. Even a bit of dirt can prevent the contacts from mating properly, leaving a small gap, causing a small arc that eventually becomes bigger, etc.

Turns out it wasn't the solder joint after all - that part was relatively clean. The actual male bullet connector was rattling around inside the female part - it wasn't a good connection at all. There were broken bits of metal in there and everything. It doesn't surprise me it failed.

It's strange because there is no connecting and disconnecting of these wires - my new build controller was permanently connected to the battery and power was controlled via a key switch. Granted, I took these wires/connectors off my first build and that probably had a good 1000 connect/disconnects which caused large sparks - plus one big short when the charger positive and negative shorted together. Maybe it's just been bad since then, and I just hadn't noticed.

jonnybump said:

I understand direct connection from a performance and reliability point of view, but wouldn't this create a nightmare regarding maintenance?.

Click to expand...

Maintenance of what? If there are no connectors there's no maintenance to do on them.

If you have a connector for the battery and need to unplug, just do that.

If you need to disconnect the system without unplugging battery, flip the breaker. I have a breaker on CrazyBike2 that's been cycled at least several hundreds of times by now, *and* is exposed to the elements (has gotten wet quite a few times in rain). So you probably won't wear it out (unless its rated specifically not to be used as a switch).


If you mean maintenance on other things, what would need disconnection other than removing the battery from it?

Hubmotor wheels you can have connectors on if you need really short wires, but I just use long enough wires to pull the wheel out if I have to.

Anything else that might need maintenance I just ensure there's a long enough wire loop at that point to move it for that.