Reading a Battery Monitor Before You Blame the Inverter

A shunt based monitor works by measuring the tiny voltage developed across a precision resistor placed in the battery negative path. Every ampere that goes into or out of the bank has to pass through that resistor, and the monitor integrates the current over time to count amp hours. That is a genuinely accurate method and it is far better than reading voltage alone.

It has one hard requirement, which is that every load and every charge source connects on the far side of the shunt. If a single circuit is grounded to the chassis or to the battery post directly instead of through the shunt, that current is invisible. The monitor then reports a state of charge that drifts further from reality every week, and it will do so forever without any indication that anything is wrong.

• The shunt sits in the battery negative path and counts every ampere
• Any load grounded around it is invisible to the monitor
• A drifting state of charge with no obvious cause suggests a bypassed load
• Check that the chassis ground for the alternator path also routes correctly

On flooded lead acid, resting voltage after several hours with no load or charge is a rough state of charge indicator and it is the number most people quote. It is also the number that most often misleads, because a bank with high internal resistance can rest at a perfectly respectable voltage and then collapse the moment current is asked of it.

On lithium iron phosphate the situation is worse for voltage as a proxy. The discharge curve is deliberately flat, so a bank at seventy percent and a bank at thirty percent can read within a couple of tenths of a volt of each other. Voltage on a lithium bank tells you almost nothing about remaining capacity, which is exactly why counted amp hours from a shunt exist.

Read the voltage at the battery terminals with a meter, not from a display. Then switch on a known load: a resistive heater element, a microwave through the inverter, or several halogen lamps, anything that draws a substantial and steady current. Watch the terminal voltage as the load comes on and hold it for thirty seconds.

A healthy bank sags a little and stabilises. A bank at the end of its life drops sharply and keeps drifting down. On one van we measured 12.7 volts at rest and 11.6 volts under about forty amperes on a bank that looked fine and was five years old on flooded cells. The inverter had been cutting out at exactly the point it was designed to, and the owner had already replaced it once.

With the same load running, put your meter probes on either side of each individual connection in turn: battery post to cable lug, lug to bus bar, bus bar to inverter terminal, and every crimp you can reach. You are measuring the voltage dropped across that one joint. A good connection reads a few hundredths of a volt. A bad one reads two tenths or more and often reads much worse.

This is where a surprising share of these faults live. A corroded lug or an under torqued terminal behaves exactly like a tired battery from the inverter's point of view, because the inverter cannot tell whether the voltage went missing inside the bank or on the way to it. Finding a quarter volt across one crimp is a twenty dollar repair that was about to be a bank replacement.

There is a signature. The bank holds voltage under a comparable direct current load, the connections all read clean, and the unit still faults. Then look at the inverter's own environment: a blocked cooling intake, an ambient temperature in a closed cabinet that exceeds its rating, or a fan that has stopped. Thermal shutdown reads very much like a low voltage shutdown from the outside.

The other real inverter fault is on the alternating current side rather than the direct current side. A transfer switch that has not fully changed over, a tripped ground fault device downstream, or a neutral bonding arrangement that the inverter objects to will all produce no output with a healthy bank. Those are diagnosis by elimination, and they bill at the diagnostic rate because the elimination takes time.

Run these in order and stop when you find the answer. Most faults resolve at step three or four, and everything before step five costs nothing but a multimeter and forty minutes. Owners who arrive with these numbers written down shorten the diagnostic time we bill, which is the whole reason we publish the method rather than keeping it.

1. Read resting terminal voltage with a meter at the posts
2. Apply a known steady load and record voltage after thirty seconds
3. Measure the drop across every connection with the load running
4. Confirm every load and charge source routes through the shunt
5. Check the inverter's cooling path and ambient cabinet temperature
6. Test the alternating current side, transfer switch and ground fault devices