Solar panels produce direct current, but the grid and most loads run on alternating current, so every photovoltaic system needs an inverter to bridge the two. The string inverter is the workhorse that does this job for a group of panels wired in series, and it sits at a middle scale between the tiny inverter under a single module and the giant inverter serving a whole array. This guide defines the string inverter, contrasts it with central and micro-inverters, and focuses on the register map a monitoring platform reads over Modbus, and how the string-level data it exposes pinpoints an underperforming or offline string.
String inverter in one line: A string inverter is a photovoltaic inverter that converts the direct-current output of one or more strings of solar panels, panels wired in series, into grid-compatible alternating current. It typically serves several strings through a handful of MPPT inputs and produces single or three-phase AC output. Positioned between the per-module micro-inverter and the array-scale central inverter, the string inverter is the most common topology on commercial and utility solar because it balances cost, efficiency, and the ability to isolate faults at the string level.
A string is a set of solar panels wired in series, so their voltages add up to a working DC voltage. A string inverter takes the DC from one or several such strings and inverts it to AC, and it usually contains multiple MPPT inputs so that a few strings can each be tracked at their own maximum power point. Physically it is a wall- or rack-mounted box, sized for anything from a rooftop commercial system to a block of a utility solar farm, and a large project is built from many string inverters distributed across the site rather than one enormous unit.
The central inverter is the opposite end of the scale. Instead of many modest inverters, a central-inverter design gathers the DC from a large section of the array, often through combiner boxes, into one very large inverter that handles a big block of capacity at once. Central inverters can be efficient and cost-effective per watt at large scale, but they concentrate risk: when a central inverter goes down, a large swath of the array stops producing, and their coarse MPPT means shading or mismatch across the block is handled less finely than distributed string inverters manage.
The micro-inverter is the other extreme, a tiny inverter mounted at each individual panel, so every module inverts its own DC to AC and is tracked on its own. Micro-inverters give the finest granularity and the best tolerance of per-module shading, but at the cost of many more units to buy and maintain. The string inverter sits deliberately in the middle: coarser than a micro-inverter but far more granular than a central inverter, giving good MPPT at the string level and, crucially, isolating faults to a single inverter or string so that one problem does not take down a large block. That balance is why string inverters dominate commercial and much of utility-scale solar.
On the DC side, a string inverter is characterized by how many strings it accepts and how those strings map onto its MPPT inputs. Each MPPT input tracks its own maximum power point, so grouping strings thoughtfully across the inputs lets strings in different conditions, some shaded or soiled, be tracked separately rather than dragging each other down. The inverter also enforces DC limits, a maximum voltage and current it can accept, which set how many panels can go in a string and how many strings share an input. On the AC side, the inverter synchronizes to the grid and outputs single-phase or, on larger units, three-phase AC at grid voltage and frequency, along with the protective functions that disconnect it if the grid goes out of bounds.
The defining operational advantage of distributing many string inverters across a solar farm is fault isolation. Because each inverter serves only its own slice of the array, a failure is contained: one inverter tripping or one string faulting takes out only that inverter's production, while the rest of the farm keeps generating. Compared with a central-inverter design, where a single failure can idle a large block, the distributed string-inverter layout degrades gracefully and localizes problems to a small, findable area. That containment is worth real money in availability, because most faults become a minor, quickly located loss rather than a major outage.
This granularity also shapes how a site is maintained. When production dips, the question is not whether the whole array is down but which inverter or string is responsible, and the distributed architecture means the answer points to a specific box among many. A technician can be sent to one inverter with a known problem rather than diagnosing a monolith, and the healthy inverters keep earning while the repair happens. The trade-off is more units to monitor, which is exactly why string-level monitoring matters so much for this topology.
String inverters expose their internal state as data, and the common language for reading it is Modbus, very often following the SunSpec model that standardizes how inverters lay out their registers. Because SunSpec defines a consistent map, a monitoring platform can read inverters from different manufacturers in largely the same way instead of learning a bespoke register layout for each brand. Through this map the platform polls the values that describe what the inverter is doing, cyclically reading each device across the site into one dataset.
The registers a monitoring system reads fall into a few families. There are AC-side quantities: the AC power the inverter is exporting, along with AC voltage, current, and frequency, which show what is actually reaching the grid. There are DC-side quantities, including per-MPPT or per-string DC voltage and current, which reveal how much each group of panels is contributing. There are environmental and health values such as internal or heatsink temperature, and there are status and fault codes that report whether the inverter is running normally, curtailed, or in a fault state, and if faulted, why. Together these give a full picture of each inverter's production and condition.
Where this becomes powerful is string-level diagnosis. On a cloud SCADA platform such as Merobix, polling per-string DC current from every inverter and comparing strings against each other under the same sunlight makes an underperforming or offline string jump out. A string reading well below its neighbours points to shading, soiling, a bad connection, or a fault in that string, while a string reading zero points to it being disconnected or its protection tripped. Because the data is per-string and standardized across brands, an operator can trend the whole farm, spot the one string that is off, quantify its lost energy, and send a technician to the exact inverter and string, rather than inferring a problem only from a sagging site-wide total that hides which string is at fault.
A string inverter is a moderately sized unit serving a few strings of panels, and a large project uses many of them distributed across the site. A central inverter is one very large unit that inverts the DC from a big block of the array at once, usually fed through combiner boxes. The string approach isolates faults to a small area and tracks strings more finely, while the central approach can be cost-effective at scale but concentrates a large amount of production behind a single point of failure.
Most string inverters expose their measurements over Modbus, very often following the SunSpec model that standardizes the register layout across manufacturers. A monitoring platform polls those registers to read AC power, voltage, current and frequency, per-string DC voltage and current, internal temperature, and status and fault codes. Because SunSpec makes the map consistent, one platform can read inverters from many brands in largely the same way and pull them all into a single dataset.
By comparing the per-string DC current from every inverter against each other under the same sunlight. A string producing well below its neighbours stands out as shaded, soiled, poorly connected, or faulted, and a string reading zero is disconnected or has tripped its protection. Because the data is per-string, a monitoring platform can point directly to the affected inverter and string and quantify the lost energy, instead of leaving the problem hidden inside a lower site-wide total.
This page references the protocol specifications published by the organizations below. Editions, product capabilities, and documentation change over time - confirm current requirements and specifications directly with the source.
Last reviewed: July 27, 2026. Merobix is not affiliated with, endorsed by, or sponsored by these organizations; their names are used only to identify the standards and products discussed.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.