Testing the Nomura MPPTs
In MSXII, the MPPTs were never fully validated to source the error for inconsistent charging of the battery. If the below block diagram is to be implemented, extensive testing must be conducted to verify the MPPTs are not the source of the issues. The following will be the testing procedure:
Unit testing of MPPTs
Each MPPT will be tested individually. They will be isolated, then attached to an E-load to draw constant current . A 240 Watt LED bulb will be shone to imitate sunlight on a cloudy day or in the nighttime testing.
The following table should outline results from varying the E-load and intensity of the sunlight. The first cell in series was isolated, and attached to an E-load.
| Light Source | V(in)oc | E-Load Setting | Measured Voltage Output | Measured Current | extra notes |
|---|---|---|---|---|---|
| Sunlight | 25.131 | ||||
| Clouds | 16.2 | 7.0kohm | 10.03 | 0.0014 | |
| Clouds | 16.2 | 10ohm | 0.155 | 0.0152 | |
| Clouds | 16.2 | 70W | basically 0 | 0.015 | |
| 240W light | 9V | 5ohm | 0.227 | 0.0422 | reading the vout from mppt was 0.56V while output from solarsenseslave was 0.056V |
| 240W light | .866 (dmm) 25V(eload off) | 5ohm | .285 | 0.0566 | |
Note: when the Mppts are lit by the light outside on a fully cloudy day, the mppt goes into short circuit mode where it provides basically no output voltage increase for a normal resistance on the line. However, using the 240W flashlights provided by Micah, we are able to see a huge increase in the voltage input and the voltage out on the line (from 15V(out)oc when cloudy to 25V(out)oc). On a sunny day outside the voltage output from the mppt is still 25V which is pretty good.
Another note: the voltage on the input of the mppt changes drastically when the outside environment does not even change. For example, the voltage output(oc) was 15V, which changed to 25V after several minutes and changed to 8.8V after several more minutes of doing nothing (same light source and intensity).
Due to the variance in the measuremntes, each mppt was taken out of the holders and tested with the bench supply and DC E-Load. The following table summarizes the results. Note that the input open circuit voltage was set to 27V by the output voltage setting potentiometer. The input open circuit voltage (voltage from power supply) should also always start at 20V (unless modified in the notes).
| MPPT# | DC E-Load Settings | V(in)oc | V(out)oc | Vin | Vout | Iin | Iout | max temp (deg C) | Notes |
|---|---|---|---|---|---|---|---|---|---|
| 1 | 6ohm | 15.98 | 25.245 | 10.01 | 18.516 | 6.264 | 3.087 | 44 | 24-25V has audible frequency oscillations (can hear) |
| 1 | 7ohm | 15.98 | 25.245 | 33.01 | 32.339 | 4.608 | 4.62 | 31.3 | When in audible range, heat increases really quickly. With an increase in resistance on the E-load is a increase in the voltage output with a significant decrease in the output current for a total less amount of output power (of course its following the inverse exponential curve until open circuit voltage in this case → infinite impedance). |
| 1 | 6ohm | - | - | - | 25.622 | 5.768 | 4.27 | 49 | during the time of this testing I also wanted to test the efficiency loss from running the mppt with an input voltage of 25-27V, in which the DCDC operates at an audible frequency of oscillations. Before the testing, I noticed that there was an increase in temperature at this frequency, so I wanted to find out the loss in efficiency. It was concluded that the loss in about 6 watts at this point, which was about the same power loss as at 20V input. It seems that the audible noise comes from the fact that the input voltage matches the upper limit of the output voltage determined by the output potentiometer. |
| 2 | 6ohm | - | - | - | 25.721 | 5.87 | 4.289 | 50 | preformed similarily to mppt #1 except that running it at 27V was 1W less efficient than under 27V. |
| 3 | 5ohm | - | - | - | 25.606 | 6.938 | 5.1206 | 43 | The power drop for this testing (as outlined) was 7W (about 1W more than normal) |
| 4 | 6ohm | - | - | - | 25.549 | 5.743 | 4.256 | 49 | About a 6.5W drop in efficiency. |
| 5 | 6ohm | - | - | - | 25.696 | 5.793 | 4.283 | 55 | Preforms similarly to the others, except gets to a higher temperature even though just 6W drop |
| 6 | 6ohm | - | - | - | 25.706 | 5.816 | 4.284 | 55 | Preforms similarly to the others, (6W drop) |
V(in)oc & V(out)oc were measured without an electronic load connected.
Note, the maximum output power for any mppt is determined by decreasing the resistance on the DC load until the power supply goes into Constant current mode. This limits the maximum input power knowing the mppts are limited to 7A on input
what is the minimum power out that we expect from each mppt?
what is the maximum power out that we expect form each mppt?
were each of them meeting this requirement?
String Testing of MPPTs
Each MPPT once unit tested, will be tested in series for the total array output power, and continuity. Like unit testing, this will be run for at least 10 minutes for full validation.
I will start by connecting 2 mppts in series, with an output voltage predicted to be about 50 volts with a current limited by the power supply.
The first part of the process involved combining both mppts in series, and based on the resources in the bay, testing was limited to the 1A output from one of the power supplies. Both mppts were set for a V(out)oc of about 27V, and limited to a current out of 1A. The dc load was set to constant resistance mode, and decreased from 100ohms until the resistance before either power supply would go into constant current mode. Both power supplies were set to 20V input.
The following table outlines the results:
| Current input MPPT1 | Current Input MPPT2 | DC Load Voltage | DC Load Current | DC Load Resistance | Max Temperature | Notes |
|---|---|---|---|---|---|---|
| 0.906 | 0.985 | 50.543 | 0.7218 | 70 | negligible | When the resistance is decreased so that one of the mppts goes into constant current mode, the mppt cannot stabilize itself at a specific voltage, and instead fluctuates like crazy. This may be an issue if a mppt needs to maintain a certain current output, and one mppt lags behind others. Also in any case, one (or both) of the mppts will have fluctuating current (before current limit is met) in order to keep a constant current from the constant resistance. |
| 2.819 | 2.595 | 49.907 | 2.0787 | 24 | 32.5 | was using two power supplies limited to 3A output (still decreased CR value until one of them went into constant current mode) 20V input for this and above on power supplies |
| 2.818 | 2.605 | 57.759 | 2.595 | 20 | 30 | Input on mppt 1 is 25V and on 2 is 32V (limits of the power supplies) |
Array testing
| MPPT # | Resistance (Load) | Voltage | Current |
|---|---|---|---|
| 1 | 1000 | 15.918 | 0.01533 |
| 1 | 10 | 0.671 | 0.0667 |
| 2 | 1000 | 15.937 | 0.1553 |
| 2 | 10 | 0.731 | 0.073 |
| 3 | 1000 | 16.332 | 0.0161 |
| 3 | 10 | 0.745 | 0.0742 |
| 4 | 1000 | 16.437 | 0.0161 |
| 4 | 10 | 0.866 | 0.0863 |
| 5 | 1000 | 16.591 | 0.0159 |
| 5 | 10 | 0.833 | 0.0831 |
| 6 | 1000 | 16.563 | 0.0159 |
| 6 | 10 | 0.931 | 0.0929 |
To conclude this testing, the only way that the cells are the point of failure is if in direct sunlight they still do not have a sufficient power output. I will try to test this, however it is winter so getting the car outside may be a little tricky.
SPI Communication Testing of MPPTs
After obtaining a SPI logic analyzer, we will record the SPI messages given by the SPV1020 MPPT IC.
The team is using 11 Nomura mppts (same model as used on MS12). Unlike in MS12 in which solar sense was broken up into two boards (a solar sense slave that was placed beneath each mppt and a solar slave master that brought the power from each solar sense slave in series for each array → two masters for each array of solar cells in series were paralleled to provide a power path to the batteries), for MS14 it will be combined into one board for the following advantages:
- Easier to assemble and debug two boards as opposed to 11 different solar slaves
- Easier wiring as we will need to only harness the thermistors going to each solar panel section, isolated spi communication, and each power line going to each mppt from solar sense board
Other features of solar sense:
- current, voltage and temperature sense of the outputs of each mppt
- isolated spi communication between the board and the mppts
Block Diagram:
