How MPPT Works: Algorithms and PWM Comparison

What the maximum power point actually is

Plot module current against module voltage and the result is a characteristic curve. At zero volts the module delivers its short-circuit current with no power. At open-circuit voltage it delivers no current and again no power. Between those extremes the product of current and voltage rises to a single peak, and that peak is the maximum power point. For a typical crystalline module it sits at roughly seventy-five to eighty percent of open-circuit voltage and around ninety-five percent of short-circuit current.

The point is not fixed. Rising irradiance lifts current almost proportionally while barely changing voltage, so the peak moves upward. Rising cell temperature reduces voltage by roughly 0.3 percent per degree, so the peak moves leftward on a hot afternoon. Over a clear day the maximum power point traces a continuously moving target, which is precisely why active tracking is required rather than a fixed operating voltage.

Technical diagram shown at a readable responsive scale.

How an MPPT controller tracks it

A tracking controller is a direct-current converter with a control loop. The converter can present an adjustable effective load to the array, which sets the operating voltage, and it delivers the resulting power to the battery or the inverter stage at whatever voltage that destination requires. Because input and output voltages are decoupled by the converter, the array can be held at its optimum while the output follows the battery.

The control loop measures input voltage and current, computes power, makes a small change to the operating point and observes whether power increased or decreased. It then continues in the direction that improved power and reverses when power falls. Typical loops execute several times per second, so the operating point follows changing irradiance closely, including during broken cloud conditions.

Common tracking algorithms

Several algorithms are in general use, and they differ in speed, steady-state accuracy and behaviour under partial shading. Perturb and observe is the most widely implemented because it is simple and robust. Incremental conductance responds faster to rapid irradiance change. Constant voltage approximation is the crudest, holding the array at a fixed fraction of open-circuit voltage, and is found only in very low-cost equipment.

More elaborate schemes add a periodic global sweep of the whole curve to escape local maxima, which matters when partial shading creates multiple power peaks. Others use temperature or irradiance measurement to predict the operating point rather than search for it. For practical purposes the differences between good implementations amount to one or two percent of annual harvest, so tracking efficiency figures on a datasheet are worth comparing but rarely decisive.

• Perturb and observe: simple, robust, small steady-state oscillation around the peak.
• Incremental conductance: faster response to rapidly changing irradiance.
• Constant voltage approximation: low cost, noticeably lower harvest.
• Global sweep variants: periodically scan the full curve to avoid settling on a shaded local peak.

Technical diagram shown at a readable responsive scale.

MPPT versus PWM: the 10 to 30 percent harvest gap

A pulse width modulation controller is fundamentally different. It has no converter stage, so it simply connects the array to the battery through a switch that is modulated to regulate charging. The consequence is that the array is pulled down to battery voltage, whatever its own optimum happens to be. A module whose maximum power voltage is 32 V connected to a 13 V battery operates at 13 V, and the difference in voltage is lost as unrecovered potential.

Worked example: a 400 W module with a maximum power point at 32 V and 12.5 A charging a 12 V nominal battery. Under pulse width modulation it operates near 13.5 V and about 13 A, delivering roughly 175 W. Under tracking control the module operates near its own optimum and the converter delivers roughly 380 W to the battery after conversion losses. That is more than double, which is why the gap is largest when module voltage and battery voltage are badly matched.

• Well-matched low-voltage module on a matching battery: gap of roughly ten to fifteen percent.
• Modern high-voltage module on a low-voltage battery: gap of thirty percent or considerably more.
• Cold and bright conditions widen the gap, because module voltage rises while battery voltage does not.
• Tracking control also allows higher string voltage, which reduces conductor cross-section and cable loss.

When PWM is still acceptable

Pulse width modulation remains defensible in narrow circumstances: very small systems below roughly two hundred watts, where the absolute energy difference is small and controller cost is a large share of the budget; installations using purpose-made low-voltage modules whose maximum power voltage is close to battery charging voltage; and applications where simplicity and minimal electronics are valued above harvest, such as basic gate or sign supplies.

Outside those cases the economics favour tracking control decisively. Where the array is more than a few hundred watts, the additional energy recovered in the first season commonly exceeds the price difference between the two controller types, and the ability to run higher string voltage saves conductor cost as well.

MPPT input limits and string configuration

Every tracker input carries three hard limits: an absolute maximum input voltage that must never be exceeded, an operating voltage range within which tracking functions, and a maximum input current or array power. String design must satisfy all three simultaneously at the temperature extremes of the site, not merely at standard test conditions.

The upper voltage limit is set by open-circuit voltage at the coldest expected temperature, since voltage rises roughly 0.25 to 0.30 percent per degree below 25 degrees Celsius. The lower limit is set by maximum power voltage at the hottest expected cell temperature. Between them lies the permissible series count. The current limit then bounds how many strings may be paralleled onto that input, and paralleling is what creates the need for individual string protection.

• Calculate string open-circuit voltage at record low ambient temperature and keep a margin below the maximum.
• Calculate string maximum power voltage at high cell temperature and stay above the lower tracking limit.
• Respect the maximum input current and maximum array power for each tracker input separately.
• Never mix different series counts on a single tracker input.

DC protection upstream of the charge controller

The tracker protects itself, not the array. Between the modules and the controller the installation needs per-string overcurrent protection using gPV type fuses rated for photovoltaic duty, a load-break isolator so the array can be disconnected safely for service, and a direct-current surge protective device to divert induced transients away from the controller input stage.

Surge protection deserves particular emphasis here because the tracker input stage is the most electrically sensitive and most expensive component in a small system, and array wiring forms a large loop that couples atmospheric transients efficiently. A surge protective device with a continuous operating voltage above the maximum string voltage, installed close to the controller, is the least costly insurance available.

• One gPV fuse per string, at not less than 1.25 times module short-circuit current.
• A load-break DC isolator rated for full string voltage and current.
• A DC surge protective device installed close to the controller input.
• An enclosure with the correct ingress and corrosion rating for the site.

Sizing the balance of system around MPPT inputs

The cleanest way to organise a multi-tracker system is to mirror the tracker structure in the combining equipment. Each tracker input receives its own group of protected string inputs, its own monitoring and its own isolation, so a fault or an underperforming string is immediately attributable to one input rather than to the array as a whole.

NEUTRON configures PV combiner equipment and DC control and protection assemblies against that structure: the number of tracker inputs, the strings per input, the per-string short-circuit current and the maximum string voltage at low temperature. Supplying those four figures allows the enclosure to arrive with the correct fuse ratings, busbar capacity, isolator rating and surge protection class, matched to the third-party tracking equipment it feeds.