Victron BMV-712 showing different SOC to my BMS — which one do I trust?

by Birch Runner · 1 month ago 41 views 4 replies
Birch Runner
Birch Runner
Active Member
21 posts
thumb_up 5 likes
Joined Jan 2024
1 month ago
#26617

Ran into exactly this last winter at the cabin. My Fogstar 280Ah LiFePO4 cells were showing 87% on the BMS app but the BMV-712 was sat at 71%. Drove me a bit mad for a week.

Short version of what I learned: trust neither blindly, but understand what each one is actually measuring.

The BMV does coulomb counting — it tracks every amp in and out from a known reference point (your last "full" sync). If that sync point is off, or if there's any drift over time, the SOC figure wanders. Mine was set to sync at 99% charged rather than waiting for the full absorption cycle to complete, so it was never getting a clean reset.

The BMS, on the other hand, is usually reading cell voltages and estimating SOC from that. LiFePO4's flat discharge curve makes this notoriously unreliable in the middle range — it'll look like 80% whether you're at 60% or 85%.

What I'd suggest:

  • Check your BMV's charge efficiency factor and Peukert exponent settings — defaults aren't always right for LiFePO4
  • Make sure the BMV is syncing at a genuine 100% (tail current properly configured)
  • Use the BMS reading mainly as a sanity check at the voltage extremes

Once I sorted the tail current threshold on the BMV, both readings came within 4–5% of each other pretty consistently. Not perfect, but liveable.

Curious whether anyone else has found the Victron integration with a smart BMS (JK or Daly via Bluetooth) helps reconcile the two figures?

DontPanic
DontPanic
Active Member
19 posts
thumb_up 2 likes
Joined Nov 2024
1 month ago
#26627

@BirchRunner the core issue is that they're measuring fundamentally different things. The BMV-712 is coulomb-counting from a fixed reference point — it's integrating current over time, so any drift in the shunt calibration or an unsynced full-charge event will compound errors over weeks.

Your BMS, meanwhile, is almost certainly using cell voltage curves to estimate SOC, which on LiFePO4 is notoriously unreliable across the flat middle section of the discharge curve (roughly 20–90%).

Which to trust? Neither unconditionally. What I do with my own setup:

  • Force a proper full-charge cycle to let the BMV-712 re-sync its 100% reference
  • Check that your shunt value in the BMV settings exactly matches your actual shunt resistance
  • The BMS voltage-based reading becomes more accurate at the extremes (near full or near empty)

After a sync, the BMV-712 is generally more reliable day-to-day on LiFePO4.

Wayne Knight
Wayne Knight
Active Member
18 posts
Joined Sep 2024
1 month ago
#26652

@BirchRunner my BMV-712 and Fogstar BMS had a similar domestic dispute — turns out the BMV just needed a proper sync at a genuine 100% full charge and suddenly they were best mates again 🤝

WhatsAFuse65
WhatsAFuse65
Active Member
42 posts
thumb_up 20 likes
Joined May 2023
1 month ago
#26667

@BirchRunner worth checking your BMV's Peukert exponent and charge efficiency factor settings too — defaults are often wrong for LiFePO4. Victron ships it configured more for lead-acid behaviour out of the box.

Also, LiFePO4 has that notoriously flat voltage curve, which makes voltage-based SOC estimation (what most cheap BMS units fall back on) essentially useless across 80% of the charge range. The BMV's coulomb counting is more reliable if it's been properly calibrated and you're giving it clean sync points.

My setup on the static van — 400Ah Fogstar Drift cells — had a 12% discrepancy until I set charge efficiency to 99% and dialled the Peukert down to 1.05. Sorted it within a few cycles.

Don't trust either reading blindly. Understand why each one says what it says.

Tango
Tango
Active Member
31 posts
thumb_up 8 likes
Joined Jul 2024
1 month ago
#26696

@BirchRunner had this exact drama on the narrowboat last spring. Both are technically "right" but measuring differently — key thing nobody's mentioned yet is parasitic loads.

If you've got anything drawing current that the BMV's shunt isn't seeing (dodgy wiring bypass, something upstream of the shunt), the coulomb counting drifts fast. Took me ages to track down a bilge pump wired wrong side of my shunt causing mine to go haywire.

Also worth doing a forced full sync — charge to 100%, hold absorption for the full duration, let the BMV hit its tail current threshold. Once it's properly synced they usually agree within a few percent and stay close.

The Victron app makes this easier if you've got the Bluetooth dongle. Can watch it sync in real time.

Log in to join the discussion.

Log In to Reply