|
Post by lbendlin on Jan 11, 2019 9:11:52 GMT -8
I had fried my first SunControl board, and have now the following precautions in place
- SwitchDoc Solar Cell switchboard (to prevent current cross-leakage if solar panels have different insolation
- Zener diode 6.2V at the end of the solar panel feed
- Thermistor to control battery charging (probably not relevant here)
- larger battery (10000mAh) - probably not relevant for the issue
Here's what I see for the SunControl
battery: 4.11V 0.0mA 0.000W
output: 5.9V 96.4mA 0.569W
solar cell: 6.29V -109.6mA -0.689W
servo: 6.53V 0.0mA 0.000W
capacity: 101% contribution: 121%
I see that the battery is full and isn't charging. I see that the solar cell is limited to 6.29V (so the Zener is definitely working) and to 690mW (ie throwing away around 6W through the Zener, nicely heating the inside of the weatherbox)
BUT what I also see is an output voltage of 5.9V ! I thought that should not happen with all the precautions I described above. The Raspberry Pi doesn't seem to mind it, showing about the same average wattage (550 mW) as if it would run on 5.2V (I see about 105mA in that scenario)
How can I get the output voltage down to a safer level? I could attach another voltage regulator board but that defeats the purpose of having the SunControl do all the work.
(another interesting tidbit is that the SunControl board seems to consume/throw away about 120 mW (20%) of the power that comes through. That value is higher when charging , and much lower when there's no input from the panels ( more around 4% loss))
|
|
|
Post by lbendlin on Jan 13, 2019 5:46:44 GMT -8
Here's a sample chart for yesterday's data. When the battery is fully charged the output load creeps up from the usual 5.2V to up to 5.9V depending on the sun pressure (I got a couple of trees that provide periods of relief - and get the output load back to 5.2V Attachments:
|
|
|
Post by SDL on Jan 13, 2019 11:47:51 GMT -8
Lbendlin,
The power should be no higher than 5.25 V. That's the specification. Could it be you have blown another board? Measure the voltage with the raspberry Pi unplugged.
Not surprised about the loss when charging. I'm pleased with the 4% loss when discharging.
BP
|
|
|
Post by lbendlin on Jan 13, 2019 15:44:45 GMT -8
When I blew the other board the load would switch off completely under similar conditions. That's not happening here. The load is working just fine (if a bit toasty). I wonder what will happen closer to summer with more sun pressure. Maybe I have to unplug one of the solar panels if one turns out to be good enough to fully charge the battery.
Did you never have a situation in Curacao where the battery would be fully charged (or too warm for a charge)? I am asking that because everything works really well as long as the battery is still accepting a charge. Things only go crazy with a full battery. (In other words maybe - what is your expectation of how a SunControl board should behave that has only a (large enough) solar panel attached but no battery?)
It's a kind of catch-22. My ultimate goal is to run the setup 24/7, regardless of the weather. This means I need to oversize both the battery (currently 10000mAh) and the solar panels (2x6W). With full sun one panel might be more than enough, but with our usual New England weather I would rather want to have a 300W panel to pull the photons out of the cloud cover. If course that's out of my price range.
I feel there has to be a better way of system capacity management.
|
|
|
Post by SDL on Jan 13, 2019 17:06:27 GMT -8
Lbendin,
I'll PM shortly.
BP
|
|
|
Post by lbendlin on Jan 14, 2019 6:48:16 GMT -8
Picture as per PM. Attachments:
|
|
|
Post by lbendlin on Jan 14, 2019 7:33:17 GMT -8
On a possibly unrelated note - I was just thinking since I have the ability to rotate the solar panels (usually for optimal insolation) I could actually use that to my advantage and rotate the panels AWAY from the sun in such a scenario. Similar to what they do to wind turbine blades when the wind becomes too strong. I'll give that some more thought.
|
|
|
Post by SDL on Jan 15, 2019 5:46:45 GMT -8
lbendlin,
While the tracking is just addressing the symptoms and not the cause, what a clever way of adjusting the solar panels to mitigate the problem.
BP
|
|
|
Post by SDL on Jan 15, 2019 13:09:32 GMT -8
Lbendlin,
I just read the solar cell specifications that you are using. Totally violates the 6.5V max voltage specification of both SunControl and SunAir. That's what killed your board and maybe what damaged (I don't know if you used it before the zener with these cells). And I think it explains the drop off as the battery gets charged (the voltage goes over the Zener diode voltage 6.2V.
Output Open Circuit Voltage: 7.7V Peak Voltage: 6.5V Peak Current: 930mA Peak Power: 6.0W Power Tolerance: +/-10% For maximum power output, orient the panel towards the sun
John
|
|
|
Post by lbendlin on Jan 15, 2019 17:33:39 GMT -8
John, it may well be that that was what killed the first SunControl board. But now with the Zener in place (and working reliably to keep the input voltage at 6.2V (well, 6.29V actually) this part of the issue is taken care of. Still doesn't explain the upcreep of the load voltage when no charging is taking place, i think.
|
|
|
Post by SDL on Jan 15, 2019 18:37:57 GMT -8
Correct. It doesn't. I'll check that when it comes back.
John
|
|
|
Post by lbendlin on Jan 17, 2019 15:00:14 GMT -8
lbendlin, While the tracking is just addressing the symptoms and not the cause, what a clever way of adjusting the solar panels to mitigate the problem. BP Would the QPM be a better option (even considering that it would require additional wiring) ? At least for the solar panels. Probably not so much for the batteries (thinking of running two of the 10000mAh ones, but it might be hard to keep them in sync)
|
|
|
Post by SDL on Jan 18, 2019 9:34:55 GMT -8
Solar Panels, yes the QPM would work. John has done that in SunRover. Not so much in the batteries. You would have to carefully check the current limitations. And you could then start a fire in your LiPos (or blow your board) with a programming error!
BP
|
|
|
Post by lbendlin on Jan 27, 2019 6:57:59 GMT -8
Ready to incorporate the QPM into my setup, and am debating varying configurations
- Connect each solar panel to the QPM individually, then to the multi panel board and the Zener - Connect all solar panels to the multi panel board, then the combined output to the QPM and the Zener - Solar panels to multi panel board, to Zener, and then to QPM
One of the considerations is what to do with the excess energy. In Winter I would like to use it to heat the box (ideally without the need for any active electronics) but in summer I'd rather disconnect the solar cells to avoid extra heat buildup. (Also wondering if any of the excess energy could realistically be used to heat the solar panels themselves, to melt away the remaining snow that's covering them)
Would you have some guidance on the most appropriate approach?
|
|
|
Post by SDL on Jan 27, 2019 16:27:53 GMT -8
I like the idea of using the excess energy to keep the box warm using peltier junctions. Cool it down in the summer too. It would be an interesting project. BP
|
|