Solar array voltage collapse.

[Sunkins]

For a degree project I've built a small model solar array (extremely small) with the goal of demonstrating the response curve of solar panels, and if I can, also demonstrate maximum power point tracking. The open-circuit voltage is 30V, the short-circuit current is 58mA.

I used a potentiometer as a load and took readings as I adjusted it which gave a perfect response curve. I then tried using an N-channel MOSFET as a variable load but as I slowly increased the gate voltage the solar arrays voltage would suddenly collapses to 0V after a certain point. I can't get the solar array's output to stay in the middle of the curve, i.e. anywhere near the maximum power point.

I've also tried using a DC-DC buck-boost converter to achieve the same thing but the it has the same problem. As I increase the converters output, it's input impedance increases, causing the solar arrays voltage to drop as expected. Once the voltage gets down to around 25V the voltage then collapses to 0V.

I have searched but can't find an answer to why this happens. Does anyone know why voltage doesn't remain stable all the way along the full response curve?

[ejeffrey]

Once you get past the power point, the current falloff is pretty sharp.  Your MOSFET is acting as a switch which it basically "wants" to do anyway.  You should build a constant current load with it instead, see any of the threads or videos on the topic here. You need a source resistor to sense the current and an opamp to adjust the gate to maintain a constant current.

It will still be quite sharp but you will have better control to trace out the curve. Than directly driving the gate.

The DCDC converter is the worst case load.  It looks like a constant power load.  Once you get to the power point the available power goes down  with increasing current draw.  So if you try to request more power it wont be able to do so and will instead just drop to zero volts as you see.

[David Hess]

The MOSFET is heating up causing its transconductance and threshold voltage to change in a way that it turns fully on.  To prevent this, add some resistance in series with the source terminal.

The switching regulator has a negative differential resistance characteristic at its input which causes the current to increase as the input voltage decreases.  To fix this, disable the feedback and run the switching regulator with a constant but adjustable duty cycle.

[Sunkins]

Thanks for your replys guys. Just wanted to update with the solution.
I did eventually find that with both the converter and the MOSFET it was caused by the devices heating up as you said David. Once the a certain amount of current was allowed to flow it would cause a runaway effect where the MOSFET would heat up, allowing more current to flow, causing more heat, etc.... until the MOSFET fully opened.
I strapped the MOSFET to a heatsink and the voltage is now much more stable,

Thanks guys.

[Zada Moore]

When the voltage output drops significantly due to factors like shading or dirt on the panels. Preventing this collapse is vital for system efficiency. Regular maintenance, monitoring, and high-quality components are key to mitigating this risk. Do you agree with me?