Configuring a Hybrid Inverter: The Settings, in the Order They Matter

One line: A hybrid inverter almost never fails because of the hardware. It fails because one of five settings was left at a factory default that belonged to a different battery, a different tariff or a different country — and the unit then does exactly what it was told, every day, for years.

Commissioning a hybrid inverter takes about forty minutes. Most of that time is not spent entering values — it is spent deciding what the values should be, and the decision depends on things the unit cannot know: which chemistry is on the DC side, how much current that pack can accept, whether the site is paid for exporting or only for avoiding import, and how stable the grid is.

This guide is the sequence, in the order the parameters have to be set, with the failure each one produces when it is wrong. It is written for the HE200 and HE300 families whose parameter numbers appear throughout, but the logic holds on any hybrid unit: declare the battery, cap the current, choose the source, choose the priority, then set the thresholds that protect the pack.

Grenkago is a supplier and exporter of solar inverters and LiFePO4 energy-storage systems based in China. We are not a manufacturer and make no claim about production facilities. The parameter numbers below come from the manuals supplied with the units; wiring and cable sizing are covered separately in the installation guide, and this page deliberately does not repeat them.

Why the sequence is not arbitrary

Each of these settings is read by the unit in the context of the one before it.

The battery type declares a voltage window — what a full pack looks like, what an empty one looks like. The charge current limit declares how fast the unit may move the pack between those two states. The low-battery thresholds are then compared against voltages that only mean something once the chemistry is known: 22.4 V is a healthy resting voltage on a 24 V lithium pack and a deeply discharged lead-acid bank of the same nominal size.

Set the thresholds first, on a unit still configured for AGM, and every number you enter afterwards is measured against the wrong baseline. This is why the factory default is the most common root cause in this category: a unit shipped configured for lead-acid, installed on lithium, with the thresholds never revisited, will disconnect at a voltage the lithium BMS considers comfortable — and the customer reports "the battery only lasts two hours".

Before the first value: three checks that cost nothing

1. Measure the battery voltage with a meter, not from the display. The display shows what the unit believes; the meter shows what is there. A 48 V pack reading 51.2 V is full, one reading 46 V is nearly empty, and if the meter and the display disagree by more than a volt, resolve that before configuring anything.
2. Confirm the pack's capacity in amp-hours. Every current limit below is a multiple of it. A 200 Ah pack and a 100 Ah pack take the same C-rate and half the current.
3. Confirm whether the battery has a communication port, and which protocol it speaks. This decides whether you are on path A or path B later, and it is much easier to check with the terminals accessible than after the cover is back on.

Do this with the inverter off-load and the PV covered. A DC string in daylight is live, and the arc from a 450 V string is not a small one.

Setting 1 — battery type

Gel and flooded lead-acid need nothing beyond the correct selection — the factory default is AGM, which is also the right choice for most sealed lead-acid. Lithium needs the selection and the communication step further down; selecting LIB without configuring communication gives you a lithium charge profile with no idea of the pack's actual state.

USE is the escape hatch: it lets you type absorption and float voltages yourself. It exists for packs whose data sheet does not match any preset, and it is the correct choice on path B — no communication, voltages entered by hand. It is also the setting most often left in place by mistake after a battery swap, with the old pack's voltages still in it.

Setting 2 — charge current, and the arithmetic that sets it

The limit is a C-rate, not a number in amps, and the conversion is trivial once the capacity is known:

Our own recommendation for lead-acid charging is 0.3C as a working target, comfortably inside the ceiling and gentler on cycle life. Above 0.2C on a lead-acid bank the pack heats, gasses and loses water it cannot get back; the capacity loss shows up six months later, not on the day.

There is a second reason this setting matters, and it is electrical rather than chemical. Charge current is what sizes the battery cable. A 6,500 W output at 51.2 V draws around 127 A; the same output at 25.6 V draws about 254 A, which is why the larger units in a range move to a higher pack voltage instead of adding parallel strings. If the charge limit is set near the top of what the cable can carry, there is no margin left for the moment the load and the charger act together.

With communication, set it to maximum and let the BMS ask. Without communication, set it by the arithmetic above and write the number on the commissioning sheet — the next person to service the unit will not redo the calculation.

Setting 3 — charge source

Two configurations account for most of this parameter's use.

PV only is for a site with a reliable grid and a time-of-use tariff, where charging from the grid at peak rate defeats the point of the battery. The consequence is honest: in a week of heavy cloud the pack arrives at the evening partly charged, and the backup duration is shorter than the brochure figure.

PV plus grid is for a site where the battery's job is availability rather than arbitrage — a shop that cannot lose its freezers, a clinic, a house in a market where outages are unpredictable. The consequence is equally honest: the unit will buy electricity to keep the pack full, and the bill reflects it.

Neither is right in general. The question to ask is what the battery is for, and to answer it before touching the menu.

Setting 4 — output priority

This is the setting that decides where the load's energy comes from when more than one source is available, and it is the one most often left wrong because the wrong value still works — it just costs money quietly.

The common modes are utility first, solar first and solar-battery-utility (SBU). Utility first keeps the battery full and uses it only during an outage: maximum availability, minimum saving. SBU runs the load from solar and battery and calls the grid only when the pack reaches its low threshold: maximum saving, and the pack is part-cycled every day whether or not there is an outage.

On a site with a generator as the third source, the priority also decides how often the generator starts — which is usually the largest line item in the running cost. Choose the mode against the cost of the fuel, not against the efficiency figure on the data sheet.

Setting 5 — the thresholds that protect the pack

These are the numbers that turn a working installation into either a reliable one or a warranty claim.

Two rules govern the low-battery figure, and both are about who acts first.

The inverter must stop before the BMS does. If the inverter's cut-off sits below the battery's own protection point, the BMS will disconnect a live load under current — which on a communicating pack means the inverter loses its data source at the worst possible moment, and on any pack means an abrupt loss of supply instead of a managed one. Set the inverter's threshold above the BMS point, always.

The transfer and return thresholds must not overlap. If the unit switches back to battery at a voltage higher than the one it switched away at, it will oscillate: grid, battery, grid, battery, every few minutes, and the customer hears the relay clicking at night. A gap of fifteen index points between transfer and return is the usual fix, which is what the 20 / 35 pair above is for.

Lithium communication: one of two paths, decided once

Path A — communication works. This is the configuration to aim for.

Physically: power cables first, then the BMS or RS485 port on the battery to the BMS port on the inverter, pin to pin, over RS485 or CAN. Pin-to-pin is not a formality — a crossed pair produces a unit that looks configured and reports nothing.

Once communication is up, the inverter follows the BMS for charge and discharge limits. That is the point of the exercise: the pack's own protection logic becomes the authority, and the settings above stop being guesses.

Path B — communication fails. Then the voltages are yours to enter, and the numbers in the table above are the starting point. On HE200 the sequence is: Programme 38 communication off, then 19 / 20 / 21 for the voltage window, 07 and 06 for the current, 15 / 16 for the SBU transfer pair. On HE300: Programme 05 USE, then 29 / 26 / 27, 02 and 11, 12 / 13.

Path B is not a failure of the installation. It is a legitimate configuration for packs without a compatible BMS — but it must be recorded, because the next technician will otherwise assume the unit is following a BMS it is not talking to.

The symptom table: what a wrong setting looks like a month later

Three things that decide the outcome on a 230 V site

Most of our French-language traffic comes from France, Belgium and West Africa, and the same three issues come up in all three.

Single-phase 230 V against three-phase 400 V. A 5 kW unit on a single-phase 230 V supply is normal in France and Belgium; the same load spread across three phases needs either three units or a three-phase model. Choosing a single-phase unit for a three-phase house is a purchasing error that no amount of configuration fixes, and it is worth confirming at the survey stage by looking at the meter, not at the breaker panel.

Voltage sag, especially in West Africa. Where the supply regularly sits well below nominal, the unit's transfer logic sees a sag as a reason to switch to battery, and the pack cycles for a grid fault that was never a real outage. If the unit exposes an AC input range setting, the wider option tolerates more sag; where it does not, the practical answer is a stabiliser ahead of the inverter rather than a threshold set so low that the unit stops protecting itself.

The RCD question. The AC input of the inverter should not pass through a residual current device: the unit's harmonic filtering can be read as a leakage current and trip it. That is an input-side rule, not an output-side one — circuits fed from the inverter's output still need their protection, and the type of device is set by local rules. In France the installation standard your electrician works to is NF C 15-100; ask them to confirm the device type rather than choosing it from a blog, this one included.

What to write down before leaving site

A commissioning sheet is five lines, and it saves the next visit:

The last column is the one nobody writes and everybody needs: why the value was chosen. Six months later the reason is gone and the number looks arbitrary.

Matching the configuration to a unit

The parameters above appear on the HE200 and HE300 hybrid families, from the 2.5 kW single-unit class up to the 12 kW three-phase-ready models, and on the H3 and H4 all-in-one storage units, which ship with the battery integrated and therefore arrive with the chemistry already declared. Where a pack is supplied separately, the configuration work is the same but the responsibility for the correct charge profile is shared with the battery supplier — which is why the communication path matters more, not less.

If you are specifying rather than installing, the useful question is not "which unit is biggest" but "which unit lets me set all five of these without a service laptop". Every model in our range is configured from the front panel.

Frequently Asked Questions

Why does the order of the settings matter?

Because later settings are interpreted using earlier ones. The battery type tells the unit which voltage window to expect, the charge current limit tells it how hard it may push, and the low-battery thresholds are compared against a voltage that only makes sense once the chemistry is declared. Set the thresholds first and the unit will protect against the wrong voltage for the whole life of the installation.

How do I know whether battery communication is working?

The unit stops raising a battery-communication fault and starts reporting a state of charge that moves with the load rather than jumping in steps. If the display shows a fixed percentage that does not change over a day of charging and discharging, the inverter is estimating from voltage, not reading the BMS — which is path B, not path A. Confirm against the parameter the manual names for communication status: programme 38 on the HE200 series, programme 37 on the HE300.

Is it safe to leave the charge current at maximum?

With a communicating lithium pack, yes in most cases, because the BMS requests what it needs and the inverter follows it. Without communication, no — the inverter has no idea of the pack's capacity and will push current the cells were not rated for. In that case the limit is arithmetic: 0.2C for lead-acid, 0.5C for lithium, where C is the capacity in amp-hours.

What should the low-battery cut-off be set to?

For a 24 V lithium pack the working figure used in our own commissioning sheets is 22.4 V, doubling to 44.8 V on a 48 V system. The exact number should come from the battery supplier's data sheet, because it depends on the cell and the BMS protection point — but it must always sit above the BMS cut-off, so that the inverter stops first and the battery's own protection never has to act.

Can I configure the unit before the battery is connected?

You can enter the menus, but the settings that matter are the ones compared against a live battery voltage, so the readings you see will be meaningless. More importantly, changing battery type with the pack connected and the unit running is the fastest way to provoke a fault. Configure at first power-on with the battery connected and the inverter off-load, then bring loads on one at a time.

Conclusion

Five settings, in one order, and the installation is either right or wrong before the first load is connected. The sequence is the same on every hybrid unit we supply: declare the battery, cap the current against its capacity, decide what the battery is for, then set the thresholds so that the inverter always acts before the battery's own protection does.

The failures in the table above are all configuration failures. None of them is a hardware fault, and all of them survive a warranty inspection because the unit is doing exactly what it was told. That is the argument for spending the forty minutes properly, and for writing down what you chose.