...
This circuit is very easy to create with discrete circuit components. To get higher panel voltage, we can just connect a bunch of these circuits in series, or just put a bunch of diodes in series and adjust the other values as appropriate.
Here’s a simulation for it in Falstad circuit simulator: https://tinyurl.com/23x589dx
Let’s try and come up with a reasonable design a component selection for it.
...
Diode string - 30 of these in series should be reasonable: https://www.digikey.ca/en/products/detail/vishay-general-semiconductor-diodes-division/VS-E4TU2006FP-N3/8269295
We can connect all the diodes in series, but have a movable connection so that the string length can be adjusted for different numbers of solar panels.
...
Building and Testing the Circuit
Before jumping into a purchase to make one of these, we can make a small model of one with regular diodes and test the IV curve. Just need to make sure to keep the tests quick so the diodes don’t heat up too much.
I made the circuit with the following parts/parameters:
...
All tests were done with a python script available here: https://github.com/mbA2D/Test_Equipment_Control
I used Eload_cv_sweep.py to control the e-load and do all the measurements, and used Eload_cv_sweep_graph.py to make the graphs.
Test 1
Current Source: Digital power supply set to 1V, 0.1A (low current since I didn’t want to burn up the diode. Larger currents later)
Diode: Generic 10A-rated silicon diode (1 diode only for the first test).
RSH: 10k 1/8W resistor
RS: 5A glass fuse acting as the series resistance (about 30mOhm resistance).
I tested with an e-load connected to the output, set in constant voltage mode. I swept the constant voltage from 0 to 1.2V while measuring the voltage and current and plotted the following graph:
...
Test 2
Current Source: Digital power supply set to 10V, 0.5A
Diode: Generic 10A-rated silicon diode (10 diodes in series).
RSH: 10k 1/8W resistor
RS: 5A glass fuse acting as the series resistance (about 30mOhm resistance).
...
I tested with an e-load connected to the output, set in constant voltage mode. Also used remote sense on the e-load to get more accurate voltage measurements (not really needed at these low currents, but it was easy to do). I swept the constant voltage from 0 to 10V while measuring the voltage and current and plotted the following graph:
...
Test Results:
These tests clearly show a maximum power point is reached with a smooth curve to the top of it, in contrast to the IV curve of a power supply that has a sharp drop at the MPP. This simulates solar panels very well and can be used to thoroughly test the MPPTs, removing the solar panels from the equation.
More Design Points
1 Diode is roughly 1 or 2 solar cells equivalent depending on forward voltage. We want to be able to do up to 30V output voltage even at low currents, so we need 30/0.7 = 42. We’ll go with 40 diodes since the ones we’re using have slightly higher voltage drop. 20 diodes on each heatsink.
Diode spacing: 200mm x 70mm heatsink
Diodes about 20mm tall, 10mm wide. 20mm horizontal spacing.
Heatsink can dissipate 300W - user should be responsible for obeying this. Safety features could be added in the future (Arduino-controlled MOSFET with a bunch of temperature sensors, voltage sensor, and current sensor).
Rshunt: 0.1Ohm should be good. P = I^2 * R = 10*10*0.1 = 10W
Rseries: 1k should be good. P = V^2 / R = 30*30 / 1000 = 0.9W
Power Input and output connectors (and remote sense)
Thermocouple adapter:
K-type thermocouple to dual banana jack adapter.