I have a Particle Argon connected to the standard battery and a solar panel.
I charged the battery fully and then connected it to the solar panel.
It has been running for 2 days now, steadily losing power, solar doesn’t do enough on cloudy Belgian days.
(No sleep)
Going from 4.1 V to 3.6 V now and reporting it all the while.
(Reading Voltage & Percentage
double voltage = analogRead(BATT) * 0.0011224;
double percentage = ((voltage-3.1)/1.1)*100;
Hope this is correct?)
Now however it is in a restart loop (i guess?), it blinks green 4 times and then a white blink and reset.
When I disconnect the solar panel it works normally again.
Any idea what is going wrong? How can I avoid this behavior?
Can you provide more details about the solar panel setup? What voltage is it supplying? Do you have some kind of charge controller hooked up, or is the panel wired directly across the battery, and therefore directly to the Argon?
The specs for that panel state a nominal output voltage of 5.5VDC, which is below the absolute maximum supply voltage rating for the Argon, but well above the normal operating range. I’m not sure how the Argon behaves if you over-volt the USB input, but it can’t be good. Of course with the battery in the circuit we don’t really know what the resulting voltage is without putting a meter on it.
Trying to charge a lithium battery up to 5.5V or higher is decidedly not safe!
If you want to use that panel you’re going to have to add some kind of solar charge controller to the system. Hook the battery to the Argon using the battery connections, not the USB input. Feed the power from the solar panel through the charge controller and then into the battery. Your Argon should be happier. Your battery will last a LOT longer (and not explode), and you may find that you have more power available than you thought.
Currently the battery is connected to the battery port and the solar panel is linked through usb to the argon. So I think that’s like you said it should be. I have no problem with overcharging a battery as my solar panel doesn’t manage to give enough for a wifi connected argon without sleep modes.
The problem is when the battery gets below a certain voltage 3.4V? it gets in a reset loop with the solar panel connected, when disconnecting the solar panel at that time it starts and runs fine on the same 3.4V (30%?) battery.
Sidequestion: Min voltage for the Argon is 3.1V, right?
When the Argon detects USB voltage it attempts to run from the USB power. It also has an on-board LiPo charge controller that will attempt to charge the battery if the battery voltage is low. This is key because it means the Argon will attempt to draw more power than it needs just to run from the USB supply - but only when the battery needs a charge.
I’m guessing that what’s happening here is that the solar panel can’t supply the needed current. The argon is looking for a minimum of 500mA from the USB supply. A 3W 5.5V panel can produce this much current, but only in full sunlight. So the argon is seeing a USB supply, but it’s not one that the chip can run from - which causes the loop. It’s flip-flopping between running from the battery and trying to run (and charge the battery) from the USB supply, with enough of a voltage drop in the process to cause a reset.
The solution is still the same though. The argon needs to run exclusively from the battery. The solar panel can be used to charge the battery if you add a solar charge controller to the circuit. The charge controller on board the argon isn’t suited for this. It’s designed around a nice constant supply of USB current - not the vagaries of a solar panel.
Thanks for the explanation!
So it means I need a solar charge controller which connects to both the battery & the solar panel?
Where do I get such a thing? And how do I hook it up?
There is the cheap/easy way and there is the “correct” way to handle the solar panel. A real solar charge controller will monitor the battery state and supply only the current needed to properly charge the battery and run the Argon. Problem is that these tend to be expensive, and most of the ones I can find are designed around larger panels and battery packs.
The fast and cheap way to do this is to use a DC-to-DC buck converter. Use the solar panel as the input voltage and set the output to 4.1V or so. The converter will supply a constant 4.1V so long as there’s enough solar to do so. This is called “float charging” and it’s decidedly hard on the lifespan of a lithium battery, but it works and won’t cost you more than 10 USD.
The sunbuddy is the best option in my opinnion and will perform better than the Adafruit model due to the fact that you can set the exact Maximum Power voltage via a resistor or by using the Potentiometer they built into the board.
In the voltaic test they did not set the Maximum Power voltage for the panels they were using which will affect the results of the test dramatically. The TI chip used in that test does the same thing the Sunbuddy does but can charge at 8 amps instead of 2 amp max on the Sunbuddy which is all you should ever need for a Particle Setup.
These boards are not true MPPT tracking designs where the Maximum Power point of the connected solar panel is constantly adjusted and tracked. These designs lock the panel from operating blow the set Maximim Power Point voltage which is the Solar Panels Vmp rating. These charge controllers start to limit charging current to keep the solar panel from operating below their Vmp voltage under all lighting conditions.
If you get the Sunbuddy you just need to be sure you set the MPP voltage to the Vmp of the solar panel your using.
The Electron and Boron already have this MPP feature built in and it works great with the 6v Voltaic panels.
Agreed that the TI24650 is overkill for the Boron (unless you’re doing something odd to drive total power consumption up to the point where you need a large panel).
MPP in the tests was set at 5.9V with Voltaic panels. Buried a bit at the bottom of thepost.
It is a bit of a judgement call to pick the MPP as a) the cell manufacturer’s spec doesn’t match the actual MPP of the panel and b) the MPP voltage changes with temperature (a fair bit) and irradiance (less). You don’t want to pick an MPP voltage that is so high that the panel crashes trying to reach that voltage, but go too low and you’re leaving power on the table.
Just brainstorming here:
In a typical Li-Po Battery +Solar application, reaching 3.4V is a problem by itself.
That's too low, and very little capacity remains.
It's already time to put the MCU to Sleep and wait for re-charge (I realize we still have to wait a week or two for Sleep Modes on Gen3).
The Li-Po voltage Dropping that low points to needing to address the Power Budget by:
Harvest more Solar Energy (larger Panel)
Store more Energy (larger Li-Po)
Use less Energy (Sleep)
Since the Argon & Xenon already have a Li-Po Charger On-board, addressing the Project's Power Budget might be more beneficial than adding an additional charge controller ?
I’m also running an argon when this thread caught my eye. I have a 9W Solar panel that has a usb plug in it and I thought was totally safe to plug into the argon and just let it deal with charging the 2000mWh lipo I have plugged in. It seemed to work so well I set up a xenon with exactly the same setup and found that (I guess with the additional wifi running) the argon doesn’t last quite so long if the days get too cloudy, too many times in a row. The solar panel has a red light that indicates it’s spitting out juice and while it does activate on cloudy days it’s not quite as effective as sunshine hitting it directly at 7am.
Because I’m in canada and winter is cold and the days are short I was thinking about buying a bigger panel of the same variety but bigger so it would be capable of catching a few more rays on those cloudy, short days. Is this not a wise idea?
I'm ready to start playing with the sleep modes on the argon. First attempts weren't so succesfull as it wouldn't reconnect to wifi again. But that's going to be required if I want to survive the cloudy days.
(I realize we still have to wait a week or two for Sleep Modes on Gen3).
What do you mean with Gen3? is this hardware ? How can I tell which gen I've got? And what's about to change for the sleep modes then? Link?
Love that this is similar to what opportunity's engineers were also working on 15 years ago :). They had it a bit harder with powering radiators and recovering from complete power outage and all..
