Can I mix solar panels with different voltages
It is possible to mix solar panels of different voltages, but you must pay attention to matching the controller and the wiring method.
For example, wiring an 18V panel in parallel with a 36V panel will cause the high-voltage panel to drop to around 18V operation, resulting in an efficiency decrease of about 20-40%.
It is recommended to use an MPPT controller or adjust via a DC-DC buck-boost module to keep the input voltage range within the controller's allowed limits, such as 12-48V.
Lose Power
How Much Power is Lost
When you are preparing to expand a 3,000W single-phase inverter system and mix panels with a 36V operating voltage together with 18V panels, the physical power loss rate usually fluctuates between 15% and 35%.
The open-circuit voltage of a standard 400W monocrystalline silicon panel is around 45V, with a short-circuit current of about 11.5A. If it is forced into an old array originally consisting of multiple 200W panels with a 20V operating voltage, the internal resistance of the entire circuit will climb by approximately 2.3 ohms. The increase in internal resistance will cause the circuit to dissipate nearly 45W of extra heat out of thin air every hour it converts electrical energy, and the surface temperature of the panel's backsheet will subsequently rise by 4 to 6 degrees Celsius.
For every 1 degree Celsius increase in the temperature of a single panel, the output power will drop according to an attenuation coefficient of 0.35% to 0.4%. Cumulatively, the total power generation for a day could drop by 1.2 kilowatt-hours. Assuming the system is connected to three panels of different specifications, as long as a shadow from leaves covers an area of just 5% of the total locally, the short-circuit current of the entire mixed array will produce a drastic fluctuation of more than 2.5 A, and the bypass diodes will be forced to turn on within 0.1 seconds.
Calculating Series Losses
If you wire a panel with a rated voltage of 36V and nominal current of 8A in series with another panel of 18V and 10A current on a 4 sq mm (12 AWG) cable, the total system voltage will stack up to 54V, but the total current will be forcibly capped at the lowest extreme of 8A. According to the power calculation formula, the 18V panel that could originally run at 10A will have an actual output of only 18V multiplied by 8A, equaling 144W. Compared to its 180W rated load, a full 36W of production capacity is wiped out, resulting in a single-panel loss rate as high as 20%.
If the cable length of the entire series system exceeds 15 meters, the voltage drop of the direct current during transmission will reach about 0.8 V, and line loss will consume an additional 6 W of useful electrical energy. For a 3,000W inverter operating within a broad MPPT voltage range of 120V to 450V, the 20% current missing from the input end will push the machine's startup time back by about 45 minutes during weak light periods in the morning and evening.
When the system faces a high-irradiance environment at 12 PM noon, with sunlight hitting the glass at a normal rated power of 1,000 W per square meter, the 18V panel limited to an 8A current actually absorbs only about 65% of the photon energy. The remaining 35% of the spectral bandwidth is completely idle, causing at least 150 W of solar radiation per square meter to be reflected back into the atmosphere in vain.
Parallel Inefficiency
Suppose you use a Y-branch connector in the combiner box to parallel a 30V, 10A 300W panel with a 15V, 10A 150W panel. The total output current of the circuit can indeed reach a peak data point close to 20A. However, the laws of physics will forcefully pin the operating voltage of the entire parallel circuit down to the baseline value of 15V, causing the actual operating voltage of that large 300W panel to shrink by as much as 50%.
The high-power panel originally purchased for $150 can now only output 15V multiplied by 10A, which is 150W. To handle the ultra-large combined current of up to 20A, you must upgrade the specification of the main trunk line from 4 sq mm to 6 sq mm or even 10 sq mm, and the procurement cost per meter of cable will increase by an extra $1.5 to $2.
The high-frequency large current shocks will also leave the 15A specification anti-reverse diodes in a long-term high-temperature overload state of 85 degrees Celsius, and the average service life of a single diode will drastically drop from an expected 10 years to less than 3 years.
Tracking Inaccuracies
Current single-phase photovoltaic inverters are generally equipped with 1 to 2 MPPT algorithm modules that can reach a peak tracking efficiency of 99.9%, used to constantly seek that product point of 2 variables that can output maximum power. When 36V and 18V panels are mixed together, two obvious power peaks will appear on the entire P-V characteristic curve: one in the higher voltage section and the other in the lower voltage section. The drop between the two can reach as much as 40W to 60W.
Under this bimodal distribution, the microprocessor inside the inverter, calculating thousands of times per second, has a 30% probability of getting stuck at the lower local extreme point, unable to climb to the true peak. This algorithmic misjudgment will inexplicably reduce the daily average power generation of the entire panel array by 8% to 12%. If calculated at a local electricity price of $0.15 per kWh, a 5,000W rooftop array will lose approximately $350 in economic revenue per year.
Clear Accounting
l To force two panels with a voltage difference of over 10 V together, you may need to separately purchase a DC-DC buck-boost module supporting a 60 A wide voltage input, with hardware unit prices fluctuating between $45 and $65.
l After installing the conversion module, the current will generate inherent conversion losses when passing through the inductor and capacitor modules. Even for advanced models using silicon carbide power devices, the conversion efficiency can at best only reach 96%, with 4% of the electrical energy forever turning into waste heat.
l For a standard 280W module with 60 cells in series, the FF (Fill Factor) value in its parameters can usually reach above 0.75. However, in a severely mismatched mixed system, the actual FF value measured by testing instruments will plummet to 0.62 or even lower.
l Looking at the life cycle, for silicon wafers working long-term under non-rated voltage parameters, the probability of PID (Potential Induced Degradation) effects occurring will increase by about 15%. A panel that originally could last 25 years might fall below the 80% power warranty baseline in just its 18th year.
l If panels of different voltages are completely decoupled by adding a $120 microinverter, the overall Levelized Cost of Energy (LCOE) of the entire system will rise by about $0.02 per kWh in the first 5 years, but over a long cycle of more than 10 years, it can yield an extra return of about 15% in total power generation.
Is it Safe
Will It Catch Fire
Forcibly paralleling a panel with an open-circuit voltage of 44V and a 22V panel on the same busbar will create a physical potential difference that forces current from the high-voltage side to backfeed into the silicon wafers on the low-voltage side. A 22V panel nominally bearing a short-circuit current of 10A will have to carry an additional 8A of reverse current from the 44V panel, turning the entire panel from a power generation device into a heating element that consumes 176W of electrical energy. After absorbing 176 W of heat, the EVA encapsulant film and PET backsheet on the back of the panel will see their surface temperature climb from a normal 45 degrees Celsius to over 85 degrees Celsius within 15 minutes.
The bypass diodes of crystalline silicon cells can usually only withstand an extreme reverse current of 15A to 20A. Once the junction temperature exceeds the physical upper limit of 150 degrees Celsius, the PN junction of the diode will be broken down, causing the entire cell string to permanently open-circuit, burning a scorched hole 3 to 5 centimeters in diameter on the backsheet.
Can the Cables Withstand It
For conventional 12 AWG (4 sq mm) photovoltaic dedicated DC cables, the nominal maximum allowable ampacity in the air is 30 A, and continuous operation in a 60 degrees Celsius working environment complies with specifications. In a mixed system with mismatched voltages, paralleling 3 panels of different specifications will stack the total current to 32A or 35A at the peak light intensity at 12 PM noon.
Running an out-of-spec 35A current continuously through a 4 sq mm copper core for 4 hours will cause the ohmic heating effect of the copper wire to push the temperature of the cross-linked polyethylene (XLPE) insulation layer on the outer sheath beyond the softening critical point of 120 degrees Celsius. Once the insulation layer softens and ruptures, and the physical distance between the positive and negative cables is less than 5 millimeters, a DC voltage as high as 80 V can break down the air, drawing a DC arc with temperatures reaching up to 3000 degrees Celsius.
A DC arc lacks the zero-crossing points of 50 or 60 times per second, making it very difficult to self-extinguish once struck. The high temperature of 3,000 degrees Celsius will melt the plastic junction box into a liquid state within 0.5 seconds, igniting surrounding dry wooden roof structures or leaves.
Will It Burn the Inverter
The design upper limit for input voltage (Voc) of most domestic light-duty MPPT controllers remains in the 100V to 150V range. Wiring one 44V board in series with two 22V boards yields a theoretical total open-circuit voltage of 88V, leaving a 12V margin from the 100V red line. When the temperature drops to minus 15 degrees Celsius on a winter morning, polycrystalline silicon panels have a temperature coefficient of negative 0.3% per degree Celsius. In extremely cold weather, the actual open-circuit voltage of this panel string will surge by 15%, climbing to 101.2 V.
The instant high voltage of 101.2 V surges into the input end of the MPPT along the cable. The filter capacitor inside the controller, rated to withstand only 100 V, will burst, and the gate of the silicon carbide field-effect transistor (MOSFET) will be broken down. Repair costs for the motherboard range between $150 and $250.
When designing a mixed series system, you must reserve an extra 20% to 25% bandwidth for the inverter's maximum input voltage. For a device with a nominal upper limit of 150V, the theoretical open-circuit voltage of the daily permitted mixed array must be controlled within 115V.
Anti-Reverse Devices
If physical constraints require paralleling 36V and 18V operating voltage panels in a single system, installing an anti-reverse diode or a series fuse on every positive branch of the line is the only means of physical isolation. Purchasing an MC4 interface anti-reverse diode with a rated current of 20A and a withstand voltage of 1000V costs around $12 to $18. Plug the $12 module into the positive output end of the 18V low-voltage panel. When the high-voltage current from the 36V panel attempts to backfeed, the unidirectional conductive property of the diode acts like a one-way valve, keeping out reverse currents as high as 15A.
The anti-reverse diode itself will produce a forward voltage drop of 0.5V to 0.7V while operating. Calculated at a normal running current of 10A, the diode itself will continuously consume 5W to 7W of electrical energy and accumulate around 60 degrees Celsius of heat on its surface. It needs to be placed in a well-ventilated shaded area of the mounting bracket.
What About Leakage
Mixing panels of different ages and packaging processes will significantly increase the probability of electrical leakage across the entire array due to differences in insulation performance. For a 200W panel used for 8 years, the water permeability of the surface EVA encapsulant may have risen by 15%. During continuous rainy weather, the insulation resistance between the panel frame and the internal cells will plummet from 50 megohms at the factory to less than 2 megohms.
Once an aging panel is wired in series with a brand new 400W panel, the total system voltage is elevated to above 80V. The 80 V direct current will leak along the wet rainwater and aluminum alloy frame, forming a leakage current of 30 mA to 50 mA. A 30 mA DC current constantly touching the human body for more than 0.5 seconds will cause muscle spasms. By using a yellow-green dual-color grounding copper wire with a cross-sectional area of no less than 6 sq mm to wire all the aluminum alloy frames of the panels in series and driving it into a galvanized ground rod 1.5 meters deep underground, you can completely conduct 50 mA of leakage current into the earth within 0.1 seconds. The material cost for this entire grounding construction is around $45.
Best Solutions
Optimal Solution
If you absolutely must mix solar panels of different voltages, the most brute-force but effective method is to directly separate the physical circuits. Mainstream hybrid inverters currently on the market are usually equipped with 2 independent MPPT (Maximum Power Point Tracking) channels. You can wire those few new 400W panels with a 36V operating voltage and 11A current in series to form one route and connect them to MPPT channel A; then wire the old 100W panels with 18V and 5A in series to form another route and connect them to MPPT channel B.
These two channels are completely physically isolated inside the machine. MPPT A will independently find the maximum power point within a voltage range of 36V to 150V, while MPPT B will operate independently through a DC-DC circuit within a range of 18V to 80V. This "divide and conquer" strategy can zero out the mixed connection loss that might otherwise be as high as 35%, letting every panel run at its theoretical efficiency of 99.5%.
A 3,000W inverter supporting dual-channel input usually costs between $600 and $800, which is only about $150 more expensive than a regular model with single-channel input. However, this $150 price difference can be entirely recouped in the first year through the extra 500 kWh of electricity generated (calculated at $0.20 per kWh).
Use Microinverters
Whether you have a 36V 72-cell panel or an 18V 36-cell panel, the wide voltage input end of a microinverter (usually supporting 16V to 60V) can take it all. Once converted into alternating current, all panels are essentially connected in parallel to the same AC bus, with the voltage frequency automatically synchronized to 50 Hz or 60 Hz. The unit price of a 300W microinverter is around $120. Although the initial hardware cost will be 30% to 40% higher than a centralized series solution, it completely eliminates the "weakest link effect."
If one of the old panels has its voltage drop to 16V due to aging, the adjacent new 45V panel can still output 400W at full power, without interfering with each other. This solution is particularly suitable for complex roof orientations, or situations where you need to slowly add solar panels in batches. Expanding the system is as simple as building blocks—just plug and play.
Parameter Matching
You should first wire those two small 12V panels in series using a 2.5 sq mm jumper wire to turn them into a "virtual" 24V module. The open-circuit voltage of this group of panels wired in series will stack up to 36V, exactly matching the voltage of that large 24V panel (the error is usually within 0.5V). At this time, you then wire this "virtual module" in parallel with that real 24V large panel.
As long as the output current difference between these two sets of circuits is within 3A—for example, one set is 5A and the other is 8A—the total current after paralleling can smoothly reach 13A, with the total voltage stabilizing at 36V, and the overall power loss can be controlled within 2%. This makeshift "2 series, 1 parallel" method only requires you to buy an extra pair of MC4 Y-branch connectors (costing about $8), saving you the cost of buying controllers worth hundreds of dollars. It is the most commonly used low-cost remedy by veteran players.
Add Optimizers
When you forcibly wire a panel with a voltage of 30 V in series into a circuit with a current of 10 A, and this panel can only pass an 8 A current, the optimizer will automatically turn on the "buck-boost" mode. It forcibly drops the voltage of the panel from 30V down to 24V, while utilizing the principle of inductor energy storage to boost the current from 8A to 10A, forcing it to integrate into the main circuit.
Although this process generates about 1.5 W to 3 W of its own heat loss, with a conversion efficiency of around 98.5%, it saves the power generation capability of the entire string of panels. An optimizer from a mainstream brand (like Tigo or a SolarEdge compatible model) is priced between $50 and $80. You usually only need to equip it to those "troublesome" panels, not the whole house, making the retrofit cost relatively controllable.
Be sure to note: When purchasing an optimizer, pay close attention to its maximum input voltage (Voc) and maximum current (Isc) parameters. The voltage upper limit of ordinary household optimizers is usually 80V or 90V. If you wire two high-power panels in series and then connect an optimizer, it is highly likely that when the open-circuit voltage spikes during low winter temperatures, the MOS tube inside the optimizer will be directly broken down and burned out.
Solution Comparison
To allow you to choose more intuitively, the core data of the four mixed connection solutions are pulled out here for a horizontal evaluation. The data is based on a standard 3 kW household system model:
Solution Type | Hardware Cost (USD) | Power Recovery Rate | Construction Difficulty | System Scalability | Core Risk Point |
Dual MPPT | $600 - $800 (Inverter) | 99% - 100% | Low (Plug and play) | Medium (Limited by port count) | Requires replacing the whole machine, old machine wasted |
All Microinverters | $120/panel x quantity | 100% | Medium (Requires laying AC lines) | Extremely High (Infinite stacking) | Troubleshooting individual faults is cumbersome |
Add Optimizers | $50 - $80/panel | 95% - 98% | Low (In-line connection) | Low (Requires protocol matching) | Many electronic modules, increased heat generation points |
Series-Parallel Matching | $15 (Cable connectors) | 90% - 95% | High (Requires precise voltage calculation) | Extremely Low (Parameters must be locked in) | Excessive voltage difference causes reverse current |
How to Choose Cables
No matter which solution you choose, when handling a mixed system, the wire gauge standard for the DC cables must be determined by the group with the "highest current." If the current of one group of panels in your parallel system is 10A and the other group is 20A, the total current after convergence is 30A. You absolutely cannot continue to use the original 4 sq mm (12 AWG) cable, but must replace the entire line with 6 sq mm (10 AWG) or even 10 sq mm (8 AWG) thick copper wire.
The safe ampacity of a 10 AWG cable in a 60-degree Celsius environment is 40 A, which ensures that when a large current passes through, the voltage drop of the cable itself is controlled within 2%. If you are reluctant to change the wires, squeezing a 30A current onto thin wires will result in a 1.5V voltage drop for every 10 meters of distance. This 1.5 V multiplied by 30 A equals 45 W of thermal power consumption, equivalent to lighting a light bulb on the wires that can never be turned off. This not only wastes electricity but also accelerates the aging of the insulation layer, which will inevitably crack and leak electricity within three years.