Lug insulation meltdown

[dmsims]

I have had my Victron 1200W inverter / DIY 24V battery build in a shed for just over a year. I had a 3D printed mount made for the Overkill BMS and wanted to install and tidy up the wiring a bit.

I had some 25mm2 / 3 AWG cables premade

All seemed! well with the install but after a couple of days VRM was reporting strange voltage drops:



This was enought to make me investigate and openning the shed door was greeted with the smell of burning plastic

The positive (Victron) busbar had melted where I made the new connection



Looking at the old cable and a spare I had made the mistake of not removing excess heatshrink on the contact side meaning a small contact area



This was duly removed with a Stanley knife installed a new bus bar and no issues since.

Battery fuse did not blow, heat/smoke alarm did not go off - I am thinking of putting some Sonoff temperature probes there

 

[Sean Very North]

Bin that bus bar, you'll never get a reliable connection when you have non metallic components under compression.

 

[HRTKD]

I thought for sure that was a knockoff Victron bus bar, but it matches the picture on the Victron website.

Lack of surface contact is a killer!

 

[Tomthumb62]

Thanks for posting this. Just about to reassemble my system and this is the kind of rookie mistake I could see myself making.

Glad you discovered this before it turned into something worse!

 

[Hedges]

Lack of surface contact is a killer!

I've come to the (educated but not experimentally substantiated) belief that gross contact area is not what matters, rather compressive force is. Because actual contact patch is far smaller than overlap area.

If heat shrink took some of the force, that would reduce force between metal surfaces. And heatshrink would creep, relieving force.

Bin that bus bar, you'll never get a reliable connection when you have non metallic components under compression.

Agreed, if that is what these have. If there is a nut directly under busbar, and plastic is held between bolt head and nut, it could work.
Can't tell how it is done in Victron photo or OP's first photo. Second photo with fiber washer between plastic holder and busbar, I don't like.
But beyond end stud, the others lifted out without melting. Could be some or all do have direct metal on metal compression. The design may actually be fine, can't tell for sure.

I don't like the second jam nut too well. It could serve as a lock, but underlying nut is the one that provides compression.

 

[Bluedog225]

Second nut can take tension off the first nut’s threads. Seems unwise and unhelpful.

 

[HRTKD]

I've come to the (educated but not experimentally substantiated) belief that gross contact area is not what matters, rather compressive force is. Because actual contact patch is far smaller than overlap area.

While I don't disagree that compressive force is critical, more surface contact is better than less surface contact. A lug with lots of surface contact that isn't tight, it just as worthless as a tight lug with very little surface contact.