What Size Charge Controller Do I Need?

Published 17 September 2026

The charge controller sizing calculator will give you an amp rating. "What size charge controller do I need" is usually asked as an amp question, and that is the wrong first question. Controllers are not usually destroyed by amps. They are destroyed by array open-circuit voltage on a clear, freezing morning, when the datasheet Voc has climbed 10 to 15% and the string you sized at 25 °C is suddenly above the input limit. That failure is instant, permanent, and not a warranty claim. This page is that check, then the MPPT versus PWM current check.

The failure mode first

A panel's Voc is quoted at 25 °C cell temperature. Voltage rises as the cell cools, typically -0.27 to -0.32% per °C on the datasheet (the sign is negative because Voc falls as temperature rises). On a bright morning with snow on the ground the cells can still be at ambient when they start producing, which is the worst combination: high irradiance and low temperature at the same time.

Coefficient is in percent per degree, so -0.29%/°C is -0.0029 in the formula. T is the record low at the array, in Celsius.

Worked example: three 60-cell panels on a 150 V MPPT

Voc 40.5 V, coeff -0.29%/°C, record low -20 °C, 3 in series, controller max 150 V:

• Temperature gap = -20 − 25 = -45 °C
• Uplift = 0.0029 × 45 = 13.05%
• Voc per panel = 40.5 × 1.1305 = 45.8 V
• String Voc = 45.8 × 3 = 137 V (fits 150 V, 9% margin)
• Four in series = 183 V (destroys a 150 V controller on the first cold sunrise)

At 25 °C the four-panel string is only 162 V, which already looks tight, and people still install it because "the MPPT window goes to 145 V operating". Operating voltage is Vmp, not Voc. The input capacitors see Voc whenever the array is unloaded, including every sunrise before the controller wakes. Size strings on cold Voc.

UK record lows are milder, around -10 to -15 °C inland. The same 40.5 V panel at -10 °C is 44.6 V. Four in series is 178 V, which still kills the 150 V unit. Cold in Britain is enough. Do not skip the correction because "we do not get winters like Canada".

Run every candidate string through the series vs parallel calculator. It applies this formula and also shows hot Vmp, which is the opposite failure: a string that is safe when cold can fall below the MPPT's minimum on a 65 °C roof and stop charging on the brightest day of the year.

Then size the amps, and do MPPT and PWM differently

Once the string voltage fits, the current rating is a different question. The charge controller sizing calculator implements both.

MPPT: power in, current out

An MPPT is a DC converter. Array watts divided by battery volts is the output current, then × 1.25 from NEC 690.8(A).

Cabin array 3,000 W on a 48 V bank:

• 3,000 ÷ 48 = 62.5 A
• × 1.25 = 78 A → 80 A MPPT
• That 80 A unit's array ceiling at 48 V is 80 × 48 = 3,840 W. The 3 kW array fits.

The same 3 kW array on a 12 V bank is 3,000 ÷ 12 × 1.25 = 312 A. That controller does not exist at hobby prices, which is why large arrays are 48 V. Voltage of the array is not what sets MPPT current; bank voltage is.

PWM: current through, voltage wasted

A PWM controller is a switch. Size it on Isc × parallel strings × 1.25. Array wattage is the wrong input and will undersize a PWM unit attached to high-voltage grid panels.

PWM also holds the array at battery voltage, so a 36 V-nominal panel on a 12 V bank throws away most of its power. Use PWM only with true 12 V-nominal panels (Vmp around 18 V) on a 12 V bank, or the 24 V equivalent. Otherwise pay for MPPT.

1.25, 1.56, and string fuses

NEC 690.8(A) treats PV current as continuous and applies 1.25. Some designers stack it twice when sizing overcurrent protection, which is 1.56, and that second factor belongs on the fuse or breaker, not automatically on the controller. The PV fuse and breaker sizing calculator is the place for 1.56. Putting 1.56 into the controller amp field will buy a larger box than the manufacturer sized for the same array, which is harmless, just expensive.

Two or more parallel strings need string fuses so that a shorted panel cannot be fed by its neighbours. That is independent of MPPT versus PWM. A 2S4P array that "fits" a 150 V controller on cold Voc can still start a fire in the combiner if the fuses are missing. Parallel is how you stay under the voltage limit; fuses are how parallel stays safe.

Cable between array and controller sees the array current, not the battery current. On a high-voltage MPPT string that current is modest. On a PWM 12 V parallel roof it is Isc times strings, and the solar cable size calculator plus the DC voltage drop calculator are what stop you melting a 4 mm² tail on a 40 A PWM run.

A sequence that does not blow up the box

1. Choose bank voltage from the battery bank guide and the inverter, not from the leftover controller in the van.
2. Choose MPPT unless the panels are 12 V-nominal on a 12 V bank.
3. Read the controller's maximum PV voltage. That number, not the MPPT operating window, is the hard limit.
4. Pick a series count whose cold Voc at record low sits at least 10% below that limit.
5. Add parallel strings until you have the watts you need.
6. Size amps: watts ÷ bank volts × 1.25, round up to a sold size.
7. Confirm the controller's array-watt ceiling at that bank voltage. Some units allow over-panelling and clip; some void the warranty. Exceeding Voc is never allowed.

If a cheap 100 V PWM unit is already in the van and the panels are 40 V Voc, you cannot series them at all, even in Britain. Parallel only, and accept the PWM loss, or replace the controller. Do not "try it in summer".

Frequently Asked Questions