Humsienk 48v golf cart battery discharge issue

Been through something similar with a cheap 48V pack on my narrowboat build before I switched to Fogstar cells. The golf cart chemistry (usually flooded lead-acid or basic LiFePO4 with a budget BMS) can throw all sorts of weird discharge cutoffs, especially when you push it hard.

Few things worth checking:

• What's the actual BMS cutoff voltage? A lot of budget 48V packs set the cell-level cutoff aggressively high, so you're losing 20-30% of usable capacity before the BMS even trips
• Load spike or sustained load? If it's cutting out under sudden load rather than gradual discharge, that's almost certainly an overcurrent protection issue rather than a genuine low-voltage event
• Temperature? Below about 5°C, LiFePO4 internal resistance climbs sharply and cheaper BMS units won't compensate — the pack looks depleted when it really isn't

With Victron kit you can usually pull decent diagnostic data via the shunt or Cerbo, but if this is a standalone pack with no comms output, you're a bit blind. A clamp meter on the main cable during the fault event would at least tell you whether it's a current spike triggering the trip.

What inverter or load is connected when it faults? That would really help narrow it down. Also curious whether anyone else here has had experience specifically with Humsienk — I've not come across that brand before, which in itself might tell us something about the BMS quality.

DebbieWalker82 | 847 posts | ⚡ Solar Addict

@VanAmy makes a solid point about the BMS being the usual culprit with these budget packs. One thing worth checking that often gets overlooked — have you verified the actual cell voltage balance under load rather than just resting voltage? With cheaper golf cart packs, one weak cell group dragging down the others can trigger low-voltage cutoff way earlier than you'd expect, making it look like a discharge issue when it's actually a balancing problem.

A decent cell log or even a basic multimeter on each group while under moderate load will tell you a lot. If you're seeing significant voltage sag on one group compared to the others, that's your answer right there. Might be worth posting your readings here before assuming the whole pack is a write-off. 🔋

Rob1963 | 2,341 posts | 🔧 Grumpy Sparky

"Humsienk" — never heard of them, which tells you everything you need to know about how much R&D went into that BMS.

Classic symptom with these no-name 48V packs is the BMS cutting out under load due to overcurrent thresholds set aggressively low from the factory — often 30-40A when the label claims 100A. Stick a clamp meter on it and watch it disconnect the moment you sneeze near it.

Before binning it, try a low-load discharge cycle first; if it recovers, it's the BMS tripping, not dead cells. Victron SmartShunt will at least tell you what it's doing rather than leaving you guessing.

Ultimately though, cheap 48V golf cart packs and anything remotely demanding is a bad marriage — @VanAmy's Fogstar route is where you'll end up anyway, might as well skip the grief.

EcoFlow_Gal | 412 posts | ☀️ Cabin Dreamer

Right so — slight tangent but worth flagging — golf cart packs are nominally 48V but that's not always the same as 48V nominal in the LiFePO4 sense? Like 4x12V lead-acid is technically 48V but the discharge curve is completely different to a proper 16S LiFePO4 setup.

The "discharge issue" could literally just be voltage sag being misread as a fault.

@Rob1963 yeah unknown brand is a red flag but tbh even that matters less than which chemistry is actually in the pack — some of these golf cart things are secretly AGM, some are basic lithium, no way to know without a proper cell-level voltage check.

What's the actual BMS readout showing? Cut-off voltage, cell delta, anything?

LutonCamper | 1,156 posts | 🔌 Off-Grid Tinkerer

Worth checking whether the BMS is cutting out under load rather than the cells actually being depleted. On my shepherd's hut build I had a similar phantom discharge issue — turned out the BMS had a comically low overcurrent threshold, tripping on any modest inverter surge.

Grab a clamp meter and monitor actual current draw at the moment it cuts. If the BMS is triggering on load spikes, you'll likely need to either:

• Replace the BMS with something configurable (Daly or JK units are reasonable budget options)
• Reduce your load profile — stagger anything with a high inrush

@Rob1963 is right that anonymous Chinese brandnames are a red flag, but the cells underneath might still be serviceable if the BMS is the weak link. Don't write the pack off until you've isolated which component is actually misbehaving.