TP4056 doesn't have any MPPT kind of logic, it's not the right choice for solar.
There are many solar charger modules for Li-ion around though.
it's ok to link
As i am totally clueless about solar and it is my first project with this tech, would you mind giving me some recommendations of possibles Li-Ion solutions which could fit here maybe, working with my 12v 25w panel ?
(charger+battery)
(post above edited with link)
Since you already have this setup, before you rebuild all from scratch, calculate/test if that 16mA discharge can be acceptable or not. For 7Ah battery, daily 0.3Ah "autoconsumption" might still be ok (depending on the weather you have there).
That's a calculation from a LLM assistant:
Battery Capacity
Total energy: 12 V × 7 Ah = 84 Wh
Daily load consumption: 0.24 W × 24 h = 5.76 Wh/day
Autonomy with no solar input:
84 Wh ÷ 5.76 Wh/day ≈ 14.5 days
So without any solar charging, the system can run for 2 weeks straight.
Winter Solar Yield (France, Dec–Jan)
Typical winter sun hours: ~1 to 2.5 effective hours/day
Estimated daily solar production:
25 W × 1.5 h ≈ 37.5 Wh/day (optimistic but achievable)
Required per day by the system: 5.76 Wh
That means even during winter, solar input can theoretically cover the load 6–8× over, assuming:
Clean panel (no snow/dirt)
Good angle (facing south)
Some sun every day
Worst-Case: Cloudy Days
Let’s say you get 5 days of full clouds (no charge):
5 × 5.76 Wh = 28.8 Wh needed
Battery = 84 Wh, so still enough—but you're using ~34% of your capacity.
Margin is tight but manageable.
Conclusion
Yes, the setup works for winter, but you're close to the edge.
1 Like
To answer that, you need to give specs of the panel. "12V panel" can mean "good for 12V battery" (normally 18V panel) or "nominal voltage is 12V"
The only specs i can find are these:
Specification:
Power under normal conditions:20w/H
Tolerance: +/-5%
SLA Battery Voltage: 12V
Type of Solar Cell: Polycrystalline
Open Circuit Voltage(Voc): 12V
Maximum System Voltage: 12V
I wouldn't trust any of those numbers.
But you can easily measure open circuit voltage of your panel.
That DFRobot charger requires 18V panel.
I expect space is not problem. Some old car / motorcycle battery could have many times 7Ah capacity even if old and abandoned.
This is probably the most easy solution yes !
From the AE listing image you linked (which might not be real), I would estimate that your panel has ~20V open circuit voltage.
AE "specs" give 12V though...
Nicolas, sorry for the delay, I was away. You're very welcome.
I used some DC power connectors, barrel-style. I got them from Amazon and there were 10 male, 10 female (both with pigtails), and 10 male and 10 female with screw terminals. When you say "banana connectors" I think of the style of connector commonly used on multimeters. I've had some really shoddy ones on a cheap, Chinese multimeter many years in the past...I'm not a huge fan. I cannot say definitively whether you should or should not use them, but I can say I would look at them very carefully. The ones I encountered and hated had slits in the male pin like a + sign and the only "springiness" was minimal. The "springiness" was not really good at making constant connection and the pins seemed fragile in terms of being able to bend the 4 petals ever so slightly, reducing their contact with the female port. I'm not describing it well so I'll say this: they were sketchy. Regular screw terminal blocks might be a good choice, too.
Be sure to check the datasheet on the 7805 for input and output capacitors. Also, bulk capacitance may be beneficial to add but I cannot say how much because I don't know enough...I'm new, too. 
Have you implemented anything to handle the back EMF/spike from the motor when it is turned off?
--HC
When you say that "it" is drawing 16mA, are you meaning the solar controller, the solar panel, or ...?
If you're measuring 16mA being drawn from the battery to the solar controller (reverse current from a charging state) that would seem concerning. However, I've not had occasion to test mine to see if it does that. I will be at that property tomorrow and I will try to remember to disconnect the solar panel and test for current from the battery to the solar controller. One big reason to buy and use a solar battery controller/tender is to eliminate the reverse flow of current from the battery to the solar panel when the panel isn't producing current. Some panels have built-in diodes for that purpose but I don't know enough to say if they "all" do.
20mA = 0.02A. Theoretically, a 5Ah battery should be able to power that load for 5 / 0.02 = 250 hours...theoretically. And that's without running the actuator/motor. But, it should get through a few long, dark nights/cloudy days okay. The real question would be the recharge rate during the day. Maybe this: 20mA draw for 24 hours = 0.48Ah drawn. 0.48Ah / 5 hours = 0.096A/hour charge rate to the battery to recover. Maybe. Throw in some losses and margin for error: 120mA from the solar charger to the battery each hour for 5 hours per day should cover that 20mA drain throughout the entire day. That may be garbage and still doesn't take into account the draw of the motor/actuator. But there should be a way to calculate how much current you need into the battery for this many hours of the day to offset 24 hours drawn at some amperage (20mA in your case).
It won't really matter so much how many volts the panel generates (as long as it's more than the minimum voltage to charge the battery and cover any system losses) as it will how much current it produces; some panels will have similar voltage output but more or less current (here, usually they're just labeled as how many Watts they (supposedly) produce). I would check the current from the panel to the solar controller while it's charging to get an idea of how much you're getting.
--HC
1 Like
Does the solar controller always have its display active, or will it go into its own "sleep mode" and blank the display after some amount of time if there is no active charging or interaction with the controls?
Nicholas, I cannot speak to the various ICs: I don't know squat.
I'm using one of these to handle charging my batteries (2x 6v SLA in series):
https://www.batterytender.com/products/battery-tender%C2%AE-5-45w-automatic-solar-controller
In Los Estados Unidos, hunting is quite the thing and in my area, deer hunting in particular. It's common to use dispensers to throw corn for the deer and those dispensers commonly run on 6 volts (and some on 12v). Small, 6V, 5Ah (or similar) SLA (sealed lead acid) batteries are easy to get from many stores, not to mention online. I have two of these in series from a southern-engineered (bricoler) project last year. These batteries cost me about $10 each as I recall. You mention 18650 batteries: a lot of this is subjective, there aren't "right" or "wrong" answers but my take on that is this: solar panels charging 12v SLA (and similar) batteries is a totally common thing (also think: car battery tenders with solar panels). I would stick with that solution if for no other reason that there are tons of materials and users so it should be easy to source well-tested solutions.
--HC
Hey! So just based off what you said, here's a list of possible issues that might be occurring:
-
Grounding issue or ground loop – Causes erratic behavior; ensure all grounds are tied at a single point.
-
Voltage dips or unstable 5V rail – Use decoupling caps (100nF + 10µF) near ICs, especially RTC and Arduino.
-
Back-EMF from motor – Add flyback diodes across motor terminals even with L298N.
-
L298N’s 5V regulator overheating – Bypass it; power logic separately.
-
Relay switching noise – Causes microcontroller brownout/reset; add flyback diode and snubber across relay.
-
Inrush current or poor solar charging – Capacitor bank may help; verify stable power under load.
-
Bad solder joint or cold joint – Thermal cycling reveals it later.
-
RTC losing time – May be underpowered or I2C interference; decouple and isolate SDA/SCL with resistors.
-
USB smoking – Likely 5V backfeed from step-down into USB; Schottky diode fix was correct.
Check those, especially power and grounding.
1 Like
There is more than one poor solder joint here...
Especially in the middle column.
1 Like
Cheap solar charge controllers don't have any boost converter circuit. If your panel has 12Voc, it will never charge the battery.
1 Like
Hi,I'm a layman who have great passion on Arduino project,can you tell me how to judge a solder joint is good or bad?
See post #49
The joint should shine.
In your case the wire is surrounded by tin. The tin should flow onto the wire...
Not if you are using lead free solder, it will often look dull.