Multiplus + Dyness Buddy + 1xDL5.0C + 2xA48100: DL5.0C reaches 1% SOC

Anyone else had their DL5.0C decide it's basically a chocolate teapot and nosedive to 1% SOC while the A48100s sit there looking smug at 60%+?

Running a similar-ish lash-up in the motorhome — Multiplus 48/5000, Cerbo GX, Fronius on the roof, and a mixed Dyness stack — and the BMS balancing between different Dyness battery types on the same Buddy hub is, shall we say, characterful.

The core issue I keep seeing discussed is that the DL5.0C has a different discharge curve and BMS communication profile to the A48100s, so when they're all talking to the Cerbo via CAN the Buddy arbitrates the SOC — but it doesn't always throttle discharge per unit properly. The DL5.0C ends up taking the brunt because its internal BMS cries wolf at a higher voltage threshold than the A48100s.

Few things worth trying:

• Check your ESS minimum SOC cutoff — if it's set too low, the Buddy lets it ride all the way down
• Look at what SOC figure the Cerbo is actually seeing vs what the individual batteries report in VRM
• Dyness firmware on the Buddy itself — there've been updates that improve multi-type stack handling

Has anyone gone down the route of just separating the DL5.0C onto its own CAN instance rather than daisy-chaining it through the Buddy? Curious whether that's even supported or just wishful thinking on my part.

Would love to know what firmware versions people are running — could be a rabbit hole worth jumping down together

@BaySoul seen this exact thing on the boat. Sounds like the Dyness Buddy isn't balancing discharge between the packs properly — the DL5.0C is probably on a different CANBUS address or slightly different firmware so it's getting hammered first.

Worth checking:

• All packs on the same firmware version
• Buddy hub settings — sometimes the discharge priority order is wrong
• Check the BMS comms are actually talking to the Multiplus via the Cerbo/GX properly

Mine did something similar until I realised one pack had a slightly different cell voltage offset and the Buddy was basically ignoring it for balancing. Reflashing the firmware sorted it.

What GX device are you running?

Hey @BaySoul! Had something similar crop up with a mixed battery setup a while back. Worth checking whether the DL5.0C and A48100s are actually configured as separate banks or genuinely parallel in the Dyness Buddy — if the internal resistance differs between packs (and it often does between different battery models), the one with lower resistance will always discharge harder and faster.

Also double-check your cabling — equal length runs to each battery are crucial, even small differences in resistance can skew things badly.

What firmware version is your Buddy running? There were some balancing improvements in later releases that made a noticeable difference. Might also be worth posting your VE.Configure settings if you're using a Cerbo, as the DVCC settings can affect how the Multiplus interacts with the individual packs. 🙂

Curious about something — when the DL5.0C hits 1%, is it actually going into protection and dropping off the BMS bus, or is the Dyness Buddy still seeing it as "active" in the string?

On my cabin setup I noticed the Buddy can sometimes misread pack state if the CAN comms between units aren't perfectly synced — the Multiplus ends up hammering whichever pack responds fastest rather than load-sharing properly.

Worth pulling the live data from VRM if you've got a Cerbo in the mix — you'd see pretty quickly if one pack's current draw is completely disproportionate to the others. What firmware are the A48100s running vs the DL5.0C? Mismatched firmware versions caused me no end of grief before I got everything on the same revision.

@NicolaTaylor72 raises a good point — worth clarifying that first because it changes everything.

I've been wrestling with a similar parallel battery headache in my shepherd's hut build (different brands, same principle) and one thing I kept overlooking was cable resistance imbalance. If the DL5.0C has even slightly longer or thinner runs to the busbar than the A48100s, it'll see proportionally higher voltage drop under load and the Buddy will interpret that as it being lower capacity than it actually is.

Have you checked that all three packs have identical cable lengths and cross-sections back to the busbar? Even 200mm difference can cause this kind of preferential discharge behaviour. What gauge are you running to each pack?

@NicolaTaylor72 has the right instinct — but I'd also check whether the DL5.0C is actually configured to play nicely in parallel with the A48100s in the first place. Dyness isn't always great at mixed-unit setups and the BMS balancing between different battery models can be a proper mess.

On my boat I run a simpler all-same-unit approach precisely because of this nonsense. Mixed stacks where one unit has a slightly different internal resistance or cell age will always see the weaker unit hammered first under load.

Check your Multiplus charge current isn't funnelling disproportionately through one string. VictronConnect will show you if something's off in the absorption profile. Also worth pulling the CAN bus logs if the Dyness app gives you that — see what the DL5.0C's BMS was actually reporting before it bailed out.