BMS for DIY motorcycle starter battery.

Been down this rabbit hole for my cabin setup and honestly the same principles apply — you want something that can handle those brutal cranking amps without browning out and leaving you stranded in a layby at 11pm in the rain (very British experience, 10/10 would not recommend).

Few thoughts from the trenches:

Cell choice matters more than the BMS here. LiFePO4 is the obvious call — safer chemistry, handles the abuse, and won't torch your garage like certain other lithium options. Grab some quality 26650 or 32700 cells rather than bodging together laptop pulls.

BMS spec to look at:

• Peak discharge rating needs to be well above your starter motor's cranking draw — I'd say 3-4x your calculated peak
• Look at the pulse rating specifically, not just continuous
• JBD/Overkill Solar units are decent entry points, Daly if you're budget-conscious (though you get what you pay for)

The real gotcha with starter batteries is that BMS latency — some cheaper units cut out mid-crank because the spike trips the overcurrent protection before it recognises it as a legitimate load. Absolute nightmare.

Anyone running a Victron SmartShunt alongside their DIY starter pack? Curious whether the SOC accuracy holds up with those weird spike-and-recover discharge patterns — my Fogstar cells on the cabin EV charger behave completely differently to what a starter battery sees.

What's your cell count target and engine displacement? That changes everything about the BMS selection really.

@SussexSolar raises a good point about the cranking current demands — this is where a lot of DIYers come unstuck.

For a motorcycle starter application specifically, the BMS needs to handle instantaneous discharge rates that can hit 10-15C on a small pack. Most budget BMS units will latch-trip under that load, leaving you with a dead bike.

Worth looking at the Daly Smart BMS range — they handle instantaneous overcurrent far more gracefully than their price suggests, and the configurable delay timers mean a genuine crank pulse won't trigger a nuisance trip.

Cell matching is equally critical here. Poorly matched cells cause one cell to sag under crank load, triggering undervoltage protection mid-start.

I run a similar high-discharge BMS on my narrowboat bow thruster circuit — same brutal instantaneous demand, same solution. Get your internal resistance figures matched within 5mΩ and you'll be sorted.

@SussexSolar @DevonDweller — worth noting that most cheap BMS units spec their peak discharge current very optimistically. The figure that actually matters for cranking is the pulse duration they can sustain, typically quoted at 30 seconds, and many budget units derate catastrophically below even that.

For a motorcycle starter application I'd be looking at a BMS with a genuine low-temperature cutoff as well — lithium cells will struggle to deliver cranking amps in a cold garage regardless of what the BMS allows through.

I run a Victron SmartShunt alongside my static caravan bank and the data logging has shown me just how brutal even "small" surge loads are on a system. Starter loads are an order of magnitude worse.

If budget allows, JBD or Daly units have reasonably honest datasheets at least. Don't trust anything without documented test methodology.

@Compo hit the nail on the head about dodgy peak current specs — learned that the hard way with my tiny house build when a supposedly "100A continuous" unit folded like wet cardboard the moment I put any real load through it.

For starter applications specifically, I'd look at units that publish their sustained discharge rating rather than the headline peak figure. JK BMS have been reasonably honest about their numbers in my experience, and the active balancing versions handle sudden high-draw events noticeably better than passive alternatives.

Worth also considering that a motorcycle starter pulse is extremely brief — we're talking milliseconds — so your cell choice matters as much as the BMS. LiFePO4 cells with low internal resistance will do half the work for you before the BMS even has to respond.

@Brummie @Compo — the other thing worth flagging specifically for starter applications is the BMS recovery time after a high-current trip. Some cheaper units will cut out under cranking load and take 30–60 seconds to reset. Fine for a solar setup, absolutely not fine when you're trying to restart a bike on the side of a wet A-road.

For my shepherd's hut build I ended up going with a Daly Smart BMS and it's been solid, but I'd want something with proper pre-charge and instantaneous recovery for a starter battery. JBD/Overkill units have a decent reputation for this but do your research on the specific model rather than just the brand — quality control varies across their range.

Also check whether your chosen BMS will even tolerate the voltage spike from the alternator when the engine fires. That catches people out.