A low-voltage disconnect protects a battery by cutting the load off when the battery voltage falls too low, preventing an over-discharge that would damage the cells. But the disconnect voltage is only half the design. The other half is the reconnect voltage, the higher level the battery must recover to before the load is switched back on, and the gap between the two is called hysteresis. If that gap is too narrow, the system falls into a destructive loop of disconnecting and reconnecting many times a minute, and the RTU it powers browns out repeatedly. Getting the width of that gap right is a small setting with an outsized effect on whether a remote site rides out a low-battery event gracefully or thrashes itself into an outage.
LVD Reconnect Hysteresis in one line: Low-voltage disconnect reconnect hysteresis is the deliberate gap between the voltage at which an LVD cuts off the load and the higher voltage at which it reconnects the load. The gap is necessary because a battery's voltage rises as soon as the load is removed, so a narrow gap makes the LVD reconnect immediately, reload the battery, sag, and disconnect again in a rapid cycle. A wide enough gap lets the battery genuinely recover before the load returns, preventing the chattering and the repeated RTU brownouts that rapid cycling causes.
The root of the problem is that a battery's terminal voltage is not a fixed measure of its charge; it depends heavily on whether a load is drawing current at that instant. Under load, the battery's internal resistance drops the terminal voltage below its true resting level, and the moment the load is removed the terminal voltage springs back up. This recovery is immediate and can be substantial, especially on a tired or cold battery with higher internal resistance. So when an LVD trips at its disconnect voltage and drops the load, the battery voltage instantly rebounds, often by a meaningful amount, purely because the load is gone rather than because the battery has actually gained charge.
If the reconnect voltage is set only slightly above the disconnect voltage, that instant rebound is enough to satisfy it. The LVD sees the recovered voltage, decides the battery is fine, and reconnects the load. The load immediately pulls the voltage back down, the LVD sees it cross the disconnect threshold again, and it cuts off once more, whereupon the voltage rebounds and the cycle repeats. This is chattering, and it can happen several times a second. The battery never actually recharges; it just bounces around the disconnect point while the load is switched on and off relentlessly, which is bad for the switching contacts, useless for the battery, and destructive for the equipment being powered.
The remedy is hysteresis: setting the reconnect voltage high enough above the disconnect voltage that the load-off rebound alone cannot reach it. With a wide gap, the LVD trips, the load drops, and the voltage rebounds partway but not all the way to the reconnect level, so the load stays off. The load only comes back when the battery has genuinely recharged enough, from the solar array or another source, to lift the resting voltage all the way up to the reconnect setpoint. That guarantees the reconnection reflects real recovered charge rather than the momentary artifact of removing the load, and it breaks the chatter loop.
The disconnect threshold is chosen to protect the battery from over-discharge, so it is set at the lowest voltage the battery should be taken to under load before the cells are at risk of damage or excessive cycle-life loss. It has to account for the fact that the voltage is measured under load, so it corresponds to a deeper state of discharge than the same voltage would at rest. Setting it too high wastes usable capacity by cutting the load off while the battery still has energy to give; setting it too low risks over-discharging and sulfating or otherwise harming the battery. The manufacturer's recommended cutoff for the battery chemistry is the anchor for this number.
The reconnect threshold is then set high enough above the disconnect to clear the load-off rebound with margin, and high enough that when the load returns the battery has recovered a genuinely useful amount of charge, not just barely crossed back over. In practice this means a gap of a full volt or more on a nominal twelve volt system, wider than beginners expect, because the goal is not to reconnect the instant the battery has any recovered charge but to reconnect only once it can actually carry the load again for a meaningful time. A generous reconnect setpoint also gives the array a chance to build real charge during the daylight before the load is reimposed, which is exactly what you want after a deep discharge.
The two thresholds have to be considered together as a pair rather than tuned in isolation, because the width between them is the whole point. A common mistake is to lower the disconnect voltage to squeeze more runtime out of the battery without raising the reconnect voltage to match, which narrows the gap and reintroduces chatter. Another is to set both thresholds close together on the assumption that a fast reconnect keeps the site up more of the time, when in reality the fast reconnect just thrashes the load and keeps the battery pinned near empty. The right mental model is to pick a safe disconnect point for the battery, then set the reconnect deliberately far above it so recovery is real before the load returns.
The equipment that pays the price for a badly set hysteresis is the RTU and its radio. When the LVD chatters, the load, which is the RTU, is switched on and off many times in quick succession, and each switch-on hits a still-depleted battery whose voltage immediately sags again. The RTU experiences this as repeated brownouts: its supply voltage dips below the level it needs to run, so it resets, reboots, drops its communication link, and may corrupt logged data or miss alarms in the process. A wide hysteresis prevents this by ensuring that when the load does come back, the battery can actually hold it above the RTU's minimum voltage rather than immediately collapsing again.
This is where the interaction between the LVD hysteresis and the RTU's own brownout threshold matters. The reconnect voltage has to be comfortably above the point where the RTU browns out under load, or the site reconnects the RTU straight into a brownout. A well-designed remote power system therefore treats the RTU's minimum operating voltage, the LVD disconnect, and the LVD reconnect as a coordinated set: the RTU brownout at the bottom, the disconnect above it to protect both the battery and the RTU, and the reconnect well above that to guarantee a stable restart. Hysteresis is the tool that keeps these levels from collapsing into each other.
Because the whole event happens at a remote site nobody is watching in person, a monitoring platform is what makes a chattering LVD visible before it becomes a mystery outage. A cloud SCADA system such as Merobix that trends battery voltage at a fine enough resolution reveals the telltale sawtooth of an LVD cycling on and off, and it captures the correlated pattern of RTU resets and dropped communications. Seeing that signature tells an operator immediately that the hysteresis is too narrow or the battery is failing, rather than leaving them to guess why a site keeps dropping offline for a few minutes at a time. The trend turns a fast, invisible electrical thrash into a diagnosable pattern that points straight at the reconnect setting.
Because a battery's voltage jumps up the instant the load is removed, purely from the load being gone rather than from any real recharge. If the reconnect voltage is only slightly above the disconnect voltage, that rebound alone reconnects the load, which sags the voltage back down and trips the disconnect again, producing rapid on-off chattering. A wide hysteresis gap sets the reconnect high enough that the load-off rebound cannot reach it, so the load only returns after the battery has genuinely recharged.
Wide enough that the voltage rebound from simply removing the load cannot reach the reconnect setpoint, and high enough that when the load returns the battery can actually carry it for a meaningful time. On a nominal twelve volt system this often means a full volt or more between the two thresholds. The disconnect is anchored to the battery's safe cutoff voltage, and the reconnect is set deliberately far above it so recovery is real, not just a momentary artifact.
The load, which is the RTU, gets switched on and off many times in quick succession as the LVD chatters, and each reconnect hits a still-depleted battery that immediately sags again. The RTU sees this as repeated brownouts, so it resets, reboots, drops its communication link, and can lose logged data or miss alarms. Setting the reconnect voltage comfortably above the RTU's brownout point ensures the RTU restarts into a stable supply rather than straight back into a collapse.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.