Charging Batteries Using Solar Panel

[ecburns33]

Why are we getting a "" when calling cubesat.charge_current to monitor the charge current provided by the solar panel? The solar panel is wired to the PyCubed board correctly via the solar panel inputs. The high side and low side switches are not jumped with wires and the VCHRG and GND jumpers are installed.

[maholli]

Hi @ecburns33,

Your RBF configuration isn't correct. I went and checked the PyCubed Inhibits & RBF Header page and realized my inhibit matrix table could be made more clear to avoid this. I've updated the matrix, so revisit that page and let me know if that makes more sense.

Anyways, you basically want the opposite of what you have: the high side and low side switches need to be jumpered and all the RBF jumpers to be removed. The VCHRG RBF jumper is only used for charging the batteries via USB. I'll make sure this is explicitly stated in more places throughout the PyCubed documentation.

Regardless of the RBF and inhibit configuration, cubesat.charge_current() should be returning a number, not an empty string. Will you verify this by connecting the board to your computer (doesn't matter if solar panels are connected or not), accessing the REPL, and then entering from pycubed import cubesat followed by cubesat.charge_current(). The REPL should print a number. You can take it one step further and verify whether or not the solar charging circuit thinks it's charging or not by calling cubesat.solar_charging, which returns False in my case since I don't have any solar panels connected. ( FYI cubesat.charge_current() has a check for cubesat.solar_charging built in already, go check out pycubed.py library file for full details)

• This is what I get in the REPL:
  

Let me know what you get and we'll go from there. If you're still having trouble, please report back with a basic description of your solar panel design so we can sanity check that while we're at it.

[ecburns33]

Update: Removed the RBF headers for VCHRG and VGND and the solar panels do not charge still. In the REPL window when cubesat.charging_current() and cubesat.solar_charging were called, the values 0 and False were returned respectively. At least when calling these functions values are now returned. The solar panel was swapped out with a DC power supply at 12.0V and 200mA max to troubleshoot and rule out if it was a solar panel issue. The above commands were called and the same result of 0 and False were returned. Is the connection from the DC power supply/solar panels wrong on the board? I am putting GND from the power supply/solar panel into the “GND” whole and the positive wire from the power supply/solar panel into the “O” whole. The voltage of the solar panel bus measured with the voltmeter is 12.0V.

For two other boards, the prompt “[ERROR] [Power Monitor] [Ernno 5] Input/output error” was seen in the REPL after typing and entering “from pycubed import cubesat”...is this due to a failed I2C connection between the solar charger and the board itself? Thank you again.

[maholli]

@ecburns33 ok good, glad you're getting the proper values for those functions. Now we can proceed with troubleshooting your solar panel setup.

1. first, make sure you've visited the PyCubed page on energy harvesting and properly configured your board for your given solar panel design. What have you installed in the D2 spot?
2. second, with the way MPPT circuits work, you wont be able to easily hook up a power supply. You can change resistors on the board to disable MPPT (and therefore make a power supply work), but this isn't recommended if you plan on using solar cells. let me know if you need more details
3. post a schematic of your solar panel design and I'll give it a quick sanity check. Measure your panel's Voc and Isc and report back with those values as well.
4. also note that in order to test your solar charging, you need to have a battery connected (make sure it's not full! something like 7.4V is good), and the solar panel needs to be able to generate sufficient power. The charger won't charge the batteries if it thinks they're full. And depending on your solar cell efficiency, just desktop lamps over the solar cells is usually not enough. In class the student's typically need high power lamps or direct sun. (If you use lamps, keep in mind solar cell performance tanks as they get hot)

As for your last question:
That error means the board failed to initialize the power monitor (which isn't related to the solar charger). I'm willing to bet this is a grounding issue related to your solar panel set up. You can ignore this for now and focus on getting solar charging working.

[ecburns33]

1. We soldered on an equivalent diode to the one mentioned on the PyCubed energy harvesting page to the D2 spot. The forward bias voltage was measured at 0.145V.
2. For the solar panels, we used the 26.6 cm^2 SpectroLab UTJ cells. 6 of these are wired in series which results in a voltage of around 14.1V and a current of around 250mA.

Today we conducted a test in slightly overcast skies. The voltage on the solar connector bus was 14.65V and the current was 160mA. After the board was connected and the REPL opened, after typing "from pycubed import cubesat" the error "[ERROR] [Power Monitor] [Errno 19] Unsupported operation" appeared. Thank you as always.

[maholli]

Ok I'll just address things as they come to me:

1. Let's focus on one thing at a time -> as I said before, let's ignore the power monitor error while we figure out your solar panel situation. Remember that the power monitor and solar charger are completely different circuits.
2. Did you have D2 soldered during the previous testing?
3. You only need 4 (maybe 5 if you think a cell will be shaded) of the UTJ cells in series. This gives you a Vmp of 9.4V which is perfect. 6 Doesn't hurt, especially as you troubleshoot things.
4. Please clarify: "The voltage on the solar connector bus was 14.65V and the current was 160mA." I'm assuming this was before connecting the panel to PyCubed, right? Otherwise what you're telling me is that the solar charging circuit is working perfect.
5. Did you try cubesat.charge_current() and cubesat.solar_charging from the REPL with the panel plugged, in and in the sun, and a partially drained battery connected? What did it report?

[ecburns33]

Max, sorry for the delay. Here's the latest:

Let's focus on one thing at a time -> as I said before, let's ignore the power monitor error while we figure out your solar panel situation. Remember that the power monitor and solar charger are completely different circuits.
Did you have D2 soldered during the previous testing?
No, D2 not soldered on this board. (1)
Yes, D2 was soldered before testing began. (2)
You only need 4 (maybe 5 if you think a cell will be shaded) of the UTJ cells in series. This gives you a Vmp of 9.4V which is perfect. 6 Doesn't hurt, especially as you troubleshoot things.
Please clarify: "The voltage on the solar connector bus was 14.65V and the current was 160mA." I'm assuming this was before connecting the panel to PyCubed, right? Otherwise what you're telling me is that the solar charging circuit is working perfect.
This was after connecting the panel to PyCubed. It was 240mA before connecting.
Did you try cubesat.charge_current() and cubesat.solar_charging from the REPL with the panel plugged, in and in the sun, and a partially drained battery connected? What did it report?
Cubesat.charge_current() reported 0; both (1) and (2)
Cubest.solar_charging reported False; both (1) and (2)

Assuming this is a board I built, I verified solar charger functionality before shipping. Let's repeat that check:
without anything connected to the solar panel connections, power the board with a partially discharged battery pack (and make sure all your RBF jumpers are removed + inhibits jumpered)
7.3V Battery for specs, measured was 7.01V (1)
7.3V Battery for specs, measured was 7.01V (2)
using a multimeter, measure the voltage on the solar panel connectors between SOLAR and GND. What voltage do you read? It should be above ~7V
Voltage is 0V (see attached picture) (1)
Voltage is 0.39V (2)

Any way we can update the software on the board to ensure it is not a software issue?
Can we conduct component level testing in order to see which components, such as the solar charger, are damaged?

Thank you as always!