As an Amazon Associate, Akermin earns from qualifying purchases. Product links on this page are affiliate links.
Open circuit voltage rises on a cold morning mainly because the cells are cold, and only marginally because the light is bright. Voltage depends on the logarithm of photocurrent, so doubling the irradiance adds only a small number of millivolts per cell, while it depends almost linearly on temperature through the diode saturation current, which falls steeply as the junction cools. A clear winter dawn delivers both conditions at once, which is why the highest voltage a string will ever present is measured in the first part of a cold bright morning rather than at noon in summer.
Cold cells raise the voltage, bright sun barely does
Two things change between a cold dawn and a warm afternoon: how much light reaches the cells and how hot they are. Both affect open circuit voltage, and they do so by amounts that are not remotely comparable. A typical crystalline module might gain around a volt across its whole series string of cells when irradiance doubles. The same module gains several volts when its cells cool by thirty degrees. The temperature term dominates by a wide margin, and any explanation that leads with the brightness has the two effects the wrong way round.
This is the opposite of the current story, where irradiance is everything and temperature is almost nothing. The two halves of the curve respond to different variables, which is the practical reason open circuit voltage and short circuit current are measured and interpreted separately rather than as one quantity. The distinction between that open circuit figure and the voltage a module actually runs at is drawn in open circuit voltage against operating voltage.
What happens inside the junction when the cells are cold
A solar cell is a large area diode with a light driven current source across it. With nothing connected, the photocurrent has nowhere to go except back through the diode, and the voltage settles at the point where the forward diode current exactly cancels the photocurrent. That balance point is the open circuit voltage, and it is set by two quantities: the photocurrent, which the light supplies, and the diode saturation current, which the semiconductor supplies.
The saturation current is the temperature sensitive one. It depends exponentially on the ratio of the bandgap energy to the thermal energy, so cooling the junction reduces it by a large factor, not a small one. A smaller saturation current means the diode needs a higher forward voltage before it will conduct enough to cancel the same photocurrent, and so the open circuit voltage rises. A secondary effect pushes the same way: silicon’s bandgap widens slightly as it cools, raising the ceiling the voltage is working against.
Working against both of those is the thermal voltage term, which shrinks as temperature falls and would on its own lower the open circuit voltage. It loses. The exponential dependence of the saturation current is far stronger than the linear thermal voltage factor, and the net result is the negative temperature coefficient every datasheet publishes.
Why irradiance only nudges the open circuit figure
The balance described above puts the photocurrent inside a logarithm. Doubling the light doubles the photocurrent and adds the thermal voltage multiplied by the natural logarithm of two to each cell, which at ordinary temperatures is somewhere around eighteen millivolts per cell for an ideal junction. Across a module of sixty cells in series that is a little over a volt for a doubling of irradiance, and real modules with non ideal behaviour sit in the same neighbourhood.
Compare that with the temperature side. A module with an open circuit voltage near 41 V and a coefficient of minus 0.27 percent per degree gains about 0.11 V for every degree it cools, so a thirty degree drop is worth more than three volts. The brightness of a cold morning is therefore a real contributor and a minor one, and the practical consequence is that a bright cold day and an overcast cold day produce string voltages much closer to each other than a cold day and a warm day do.
| Change in conditions | Approximate effect on module open circuit voltage | Why |
|---|---|---|
| Irradiance doubles | Roughly one volt on a sixty cell module | Voltage follows the logarithm of photocurrent |
| Cells cool by 10 degrees | Roughly one volt at a coefficient near minus 0.27 percent | Saturation current falls steeply as the junction cools |
| Cells cool by 35 degrees | Close to four volts on the same module | The same effect, accumulated |
| Irradiance halves at the same temperature | Roughly one volt lost | The logarithm works in both directions |
The morning reading is the one that sizes the string
String length is decided by the worst case voltage, not the typical one, because exceeding the input limit of an inverter or the rating of the cabling is a single event rather than a gradual deterioration. The calculation takes the module open circuit voltage at standard conditions, applies the voltage coefficient over the difference between the record low ambient temperature at the site and the 25 degree reference, and multiplies by the number of modules.
Work an example. A module states 41.0 V open circuit and a coefficient of minus 0.27 percent per degree. The site has a record low of minus 10 degrees Celsius, which is 35 degrees below reference. The voltage rises by 0.27 multiplied by 35, which is 9.45 percent, giving 44.87 V per module. Twenty two modules in series reach 987 V, which clears a 1000 V system limit. Twenty three reach 1032 V, which does not. One module decides whether the design is compliant, and it is decided by a condition that occurs on a handful of mornings a year.
Measuring it rather than calculating it needs the right moment and the right instrument. The procedure is set out in how to measure solar panel open circuit voltage, and the general technique on the meter side in measuring direct current voltage with a multimeter.
Taking the reading safely at the coldest hour
Two instrument details govern whether a reading is even possible. The first is the input ceiling. The ZIBOO FT-2000W panel tester states an open circuit voltage range of 12 to 150 V, a short circuit current capability of 35 A, maximum power up to 2000 W, an accuracy figure of plus or minus 0.8 percent, a CAT III 1000 V and CAT IV 600 V rating with double insulated probes to IEC and EN 61010, plus data hold and a backlit display. That 150 V ceiling is a module level or short string instrument by design, and connecting it across a full string at nine hundred volts is outside what it publishes.
The second is whether the instrument publishes an accuracy figure at all. The Fabater FT 2000W panel tester states a 2000 W capacity, maximum power point tracking, a curve testing display on a 3.2 inch screen, data hold and a backlight. Its bullets give no accuracy percentage and no measurement category, so for both of those the honest entry is: not stated in the listing bullets. A tester without a published category is one whose connection point has to be chosen conservatively, and instruments whose ratings are stated plainly are gathered in voltage testers.
Finally, treat the cold bright morning as a working hazard rather than a measurement opportunity. The array is at its annual maximum voltage, direct current does not cross zero to help extinguish an arc, and gloves make it easy to misjudge a connector. Measure with the string isolated, use the data hold rather than reading in awkward light, and record the ambient temperature next to the voltage so the figure can be compared with the design calculation afterwards.
Frequently Asked Questions
How much does open circuit voltage change per degree?
Crystalline silicon modules commonly state a voltage coefficient somewhere between a quarter and a third of a percent per degree, and the exact figure is on the datasheet. Applied to a module with an open circuit voltage in the low forties, that works out to roughly a tenth of a volt for every degree the cells cool, which accumulates quickly over a thirty five degree swing.
Does the panel have to be producing power for this to matter?
No, and that is the hazard. Open circuit voltage is present whenever light falls on the module, regardless of whether anything is connected. A string left open on a cold bright morning sits at its highest voltage of the year while delivering no useful output at all, which is precisely the condition in which somebody is most likely to be working on it.
Why is the record low temperature used instead of the average?
Because the equipment limit is not an average. Exceeding the maximum input voltage of an inverter or the system voltage rating of the cabling is a single event failure, so the design case is the coldest condition the site can produce rather than a typical one. Design guidance normally points to the lowest recorded ambient temperature for the location.
Does the voltage keep rising all morning?
No. It peaks early and then falls as the cells warm, even though irradiance continues to climb towards noon. The temperature effect is much the stronger of the two, so as soon as the module starts absorbing significant energy the voltage begins to drop. The peak sits in the first part of the morning on a clear cold day.
Is a snow covered ground relevant?
It can be, because snow is highly reflective and adds to the light reaching the module while doing nothing to warm it. A clear morning after snowfall combines unusually high in plane irradiance with unusually cold cells, which is the worst case combination for string voltage even though the brightness contribution is the smaller of the two.
Can I measure a full string with a handheld panel tester?
Only if the string voltage is inside the input range the instrument publishes, and many panel testers top out well below a full string. Check the stated voltage ceiling before connecting anything, because exceeding it is a safety matter rather than an inaccuracy. Measuring module by module, or using an instrument rated for the whole string, are the two honest options.