SmartShunt never syncs to 100 % – BMS cuts off while charge current is above 10 % tail current

Had exactly this issue in my shepherd's hut build last summer. Running a Fogstar Drift 100Ah with its internal BMS, and my Victron SmartShunt just sat there refusing to hit 100% — stuck at 97% every single time the BMS cut the charge.

The problem as I understand it: the SmartShunt needs to see the tail current drop below your set threshold before it'll call the battery full and sync. But with a lot of lithium packs (especially the self-contained ones with no separate BMS port), the BMS just yanks the charge dead — no gradual taper, no absorption phase, nothing. So the shunt never gets that gentle tail-off it's looking for.

Few things I've tried or seen suggested:

• Lower the tail current % in VictronConnect — drop it right down, maybe 1–2%. Doesn't always help if there's genuinely no taper at all
• Set a "charged voltage" trigger — if your pack hits, say, 54.8 V (on 48 V), you can use that alongside a short time delay as the sync condition
• Rely on a periodic full sync manually — bit rubbish but some folk just reset it occasionally

Honestly the cleanest fix I found was setting the charged voltage just below where the BMS kicks in, combined with a 5-minute hold time. Got me reliably syncing within a few cycles.

Anyone else running internal-BMS packs (the budget all-in-one type especially) hit this? Curious whether different brands behave differently — wondering if something like a Jakiper or EG4 has the same hard cutoff behaviour.

@MikeCross your BMS is pulling the rug out from under the SmartShunt before it ever sees the tail current drop low enough to trigger a sync — classic catch-22 where the Fogstar's internal BMS terminates charge whilst the shunt is still thinking "nah, we're not done here mate."

Set your charged voltage in the SmartShunt to maybe 0.1–0.2V below whatever your charger's absorption voltage is, so it registers the voltage condition before the BMS panics and disconnects. Had the exact same nonsense on my shepherd's hut Drift cells until I tweaked that one parameter — been syncing clean ever since.

Also worth double-checking your tail current percentage in VictronConnect isn't set too aggressively low for a 100Ah cell.

@MikeCross had the same maddening dance on my narrowboat last winter. The Fogstar BMS is doing its job — it hits full voltage and quietly disconnects before the SmartShunt ever sees that tail current taper below your threshold.

What worked for me: drop your charged voltage in the SmartShunt settings to about 50mV below where the BMS actually trips. That way the shunt declares "full" fractionally before the BMS pulls the plug.

Also worth checking your tail current percentage — I had mine at 4% which sounds fine on paper, but with a 100Ah bank that's 4A, and the BMS was cutting at 6A. Bumped it to 8% and suddenly the SmartShunt started syncing properly.

It's a timing race more than a settings problem, really. You're just trying to win by a whisker.

@MikeCross the fix nobody mentions: drop your SmartShunt charged voltage threshold to about 0.2V below whatever your BMS terminates at. That way the shunt registers "charged" on the voltage signal before the BMS pulls the plug, instead of chasing a tail current that never arrives.

Also worth checking your charge detection time setting — default 3 minutes is often too short when solar input is variable. Bump it to 5 or 6.

Done this on my own setup after weeks of the same nonsense. Sorted it completely. The SmartShunt isn't broken, it's just fighting a battle it can't win against the BMS with default settings.

@MikeCross worth adding to what @VitoProject said — also check your tail current setting in the SmartShunt. Default is 4% but with a BMS cutting off early you might need to bump it up to 8–10% so the shunt declares "full" before the BMS pulls the plug. It's a bit of a fudge, I know, but it gets your SOC syncing properly.

Also, are you running any load during absorption? Even a small parasitic draw can keep current artificially high and delay the tail current dropping. On my shepherd's hut setup I found the fridge compressor cycling was masking the true tail current completely. Switching to monitoring only during quiet periods made a big difference. Once you've got a clean sync the shunt tracks brilliantly from there.