Battery monitoring: the one gauge that stops you guessing your amp-hours

Somewhere out past Marree a few years back, I watched a bloke in a near-new van argue with his fridge. Voltmeter on the wall said 12.4 volts, which he reckoned was "still half a tank." Twenty minutes later the fridge alarm went off and the thing shut down to protect the compressor. He wasn't out of charge because the battery lied to him. He was out of charge because he was reading the wrong instrument for the job.

That's the guts of this article. A battery monitor with a shunt is the single upgrade that stops you guessing your amp-hours, and yet it's the bit most people skip when they're speccing a new setup, because a voltmeter is free (it's built into most DC-DC chargers and solar controllers) and a proper monitor is another $150 to $400 on top of a system that already feels expensive.

I've run both. Nine years of towing has taught me the voltmeter is fine for a rough sanity check and useless for anything else.

Why voltage alone tells you almost nothing

Voltage measures pressure in the system, not the amount of charge stored. For a lead-acid or AGM battery, the voltage curve does slope down reasonably predictably as the battery discharges, which is why old-school voltmeter-based "fuel gauges" sort of worked, badly, on those chemistries. You'd see 12.6V at rest and know you were close to full, 12.2V and know you'd used a fair chunk, 11.8V and know you were in trouble.

Lithium (LiFePO4) breaks that logic completely. A 100Ah lithium battery sits at roughly 13.2 to 13.3 volts resting from about 90% down to about 20% state of charge. The curve is almost flat across the whole usable range, then it falls off a cliff near empty. That means your voltmeter can read a healthy 13.1V while you're sitting at 30% charge, then drop to 12.8V ten minutes later when you're at 8% and about to hit low-voltage cutoff. There's no gentle warning. You go from "looks fine" to "shut down" in the time it takes to boil the kettle.

I'd argue this is the main reason lithium upgrades disappoint some tourers in year one. They bolt in a drop-in lithium replacement, keep the old voltage-only display, and then can't work out why the system seems to run fine for days and then quits with no notice. The battery didn't fail. The instrumentation was never built to read it.

What a shunt actually measures

A shunt is a precision resistor wired in series with the negative lead between your battery bank and every load and every charge source. Every amp that goes in or out of the battery has to pass through it. The monitor reads the tiny voltage drop across the shunt many times a second and does the maths: amps in, amps out, running total, converted to a state-of-charge percentage against the battery's rated capacity.

It's the same principle as a fuel flow meter on a boat, rather than a fuel gauge that guesses from tank shape and float position. It's counting what actually moved, not inferring it from a proxy reading.

Good units, the Victron BMV and SmartShunt range, the Redarc Manager30's built-in monitoring, the Renogy DCC50S paired with a monitor, all do this the same fundamental way. They differ in display quality, Bluetooth app polish and how many extra sensors you can hang off them (tank levels, temperature, a second battery bank), but the core function is identical: count the electrons, keep a running percentage.

Where the shunt actually goes

This trips people up during install, so it's worth being precise. The shunt sits in the main negative cable between the battery's negative terminal and the busbar or chassis point where every other negative (fridge, lights, DC-DC charger, solar controller, inverter) connects. If any load or charge source has its negative wired directly to the battery terminal, bypassing the busbar, the monitor will never see that current and your reading will drift out over weeks.

I've seen this exact fault on a mate's van, an inverter earthed straight to a chassis stud instead of through the shared busbar. His monitor read 60% while the battery was closer to 35%. Took us an hour with a multimeter to trace it back. Worth getting right the first time, or getting an auto-electrician to do it, because re-wiring a negative bus after the annexe and the cabinetry are back in is a miserable afternoon.

Setting the battery capacity, and why "full synchronisation" matters

Every shunt monitor needs you to tell it the rated capacity of your bank in amp-hours, and most also want a "charged voltage" and a "tail current" setting, the point at which the monitor decides the battery is genuinely full and resets its count to 100%. Get this wrong and the percentage will slowly drift regardless of how good the shunt is, because the running count is only ever as accurate as the last full recalibration.

Practically, that means running your battery up to a proper full charge, sitting at absorption or float for the time the battery maker specifies, at least once every week or two of hard use. If you're purely on solar and never quite reach a true full charge on cloudy runs, the percentage figure will wander low over time even though nothing is actually wrong with the battery. I've had this happen through a stretch of overcast weather doing the Gibb River Road and had to run the genset for an hour purely to force a full recalibration.

Lithium versus AGM through the lens of the shunt

This is where the monitor earns its keep hardest. With AGM, you generally shouldn't use more than 50% of rated capacity before recharging, so a 200Ah AGM bank gives you roughly 100Ah of real, sustainable use. Push much past that regularly and you shorten the battery's life significantly. With lithium, you can comfortably use 80 to 90% of rated capacity without meaningful damage, so that same 200Ah figure gives you 160 to 180Ah of genuinely usable charge.

Without a shunt monitor, you have no reliable way to know where either battery actually sits in that range, and on lithium in particular, the voltage tells you nothing until it's nearly too late. If you're weighing up the two chemistries properly, our lithium versus AGM cost-per-usable-amp-hour breakdown goes through the actual dollar figures per usable amp-hour, and a decent shunt monitor is really the tool that lets you use those figures with any confidence out in the field.

Sizing it against what you're actually running

A monitor is only useful if you've done the sums on the load side too. If you haven't worked out what your fridge, water pump, lighting and any inverter loads actually draw over a 24-hour cycle, the percentage on the shunt display is a number without context. I've gone through this properly in sizing your 12V fridge to your battery, and it's worth doing that maths before you buy the monitor, not after, so you know roughly what daily percentage swing to expect and can spot a genuine fault (a stuck compressor, a parasitic draw you didn't account for) rather than assuming the unit itself is wrong.

Same goes if you're running a dual-battery setup off a single alternator feed. A DC-DC charger changes the whole charging picture compared to a straight alternator connection, and if you haven't read up on why that matters, our piece on DC-DC chargers and why the alternator alone won't cut it covers the charging side that feeds directly into what your shunt monitor will show you.

What I actually run, and the honest downside

My own rig has a Victron SmartShunt talking to a small Bluetooth display and the VictronConnect app on my phone. It cost more than the basic voltmeter that came wired into the van from the factory, and it took an extra hour during the lithium conversion to mount the shunt somewhere sensible and re-run the main negative through it properly.

The honest downside is that it's another thing that can fail, another Bluetooth connection to troubleshoot, another setting (the charged voltage and tail current figures) that can be wrong from the factory and needs checking. I've had the app lose its pairing on a corrugated stretch and refuse to reconnect until I power-cycled the whole system, which is a minor annoyance at a servo, and a genuinely bad time if it happens the same afternoon your fridge alarm goes off with no other way to check state of charge.

But weighed against the alternative, a fridge full of meat shutting down on the Oodnadatta because a voltmeter read "fine" right up until it wasn't, I'll take the extra complexity every time. If you're running lithium especially, I'd call this close to compulsory rather than optional, and I don't say that about much in this game.