It is a clone, USB-C
I ordered a fresh batch to see if it was a defective one, or maybe got damaged in some tests.
You need to be very careful when supplying 5V to the 5V pin.
With a clone I do not recomment having anything connected via USB or Vin either on purpose or accidentally when powered by 5V.
Good to know!
Would it be safer to supply a 7-8v voltage to VIN and then still be able to use 5V pin to power the DS3231 and the relay, and use USB at the same time?
I feel it would really help to be able to use the console with the arduino connected to troubleshoot.
For now i added a 1n5817 between my +5v source and the 5v pin of the arduino, but not sure it is enough protection.
Likely not.
You really don't know how that crappy buck behaves in some specific situations, like for example when you disconnect the battery and that capacitor discharges.
That's questonable with a real nano for a clone who knows.
but not sure it is enough protection.
Protection against what?
On the Uno they have a fairly complicated scheme to make sur the board is only powered by source. The nano does not have any of that circuitry.
I'd go with the LM7805. At least you know it will provide a well regulated 5V output even with a poorly regulated 12V input and can even reduce some noise if present.
1 Like
It make sense yes.
So the next steps, in order of preference, are:
-
LM7805 instead of the buck converter to provide a stable 5v output.
test a few days like that -
replace the LM298N + Relay by a TB6612FNG.
Thanks again for all the advices.
I will post an update in a few days after more testing.
Look carefully at the nano, it may be a clone that runs at 8MHz with Vcc of 3.3v, they are a bit rare but do exist.
You're right about the possible cold joints, i will resolder them to be sure.
The other side is an intentional connection, same, i will resolder this part properly.
thanks for checking !
I have done a similar project. I cannot address all you have going on but I will add some of my experiences which may help.
I used a solar panel/battery setup. I used a Battery Tender brand solar controller to handle the interface from the solar panel to the battery. The Nano (and the clones I use) have a voltage regulator but I'm powering a DS3231, a 5V relay, and a 4x20 LCD from 5V and I wasn't going to ask the Nano's vreg for that loading. I used a 7805 with a heatsink to get my 5V for everything that needed 5V. I cannot speak to the buck converter, I chose the 7805 because I had experience with using them.
My setup was working fine on the bench with no motor connected. Worked for weeks, zero problems. Until I connected the motor and then it went south. I was using a 12v motor which drew about 1A loaded or unloaded (I didn't have much load on it). It had about 24" of wire to it which may or may not have made any difference. I could activate the motor twice before the data on the LCD backpack got corrupted and Line 1 turned into hieroglyphics. Another 2 cycles and the LCD went blank. Resetting the Nano (or power disconnected/reconnect) made it fine again. Like you, I was reversing the motor so a straight diode across the motor was out. I forget his username but a guy named Leo on here posted a reply to such a question back in 2019 and I found it: use a full-wave bridge rectifier. I cobbled one together with 4 diodes (1N4007 equivalent) and connected the Pos to the Pos source, GND to GND, and the motor wires to the AC points of the rectifier. That solved my problem and is what is in place in my design in use.
I'm re-evaluating some of that project including how I drive that motor. I chose a DPDT relay to reverse the current and a MOSFET to disable all motor power when not actually running. It works well but it takes a lot of PCB space. I'm considering using TI DRV8251, FWIW. This thread has offered a couple of other options which I will explore.
Best of luck w your project. Having an automatic door opener is super handy and the commercial units seem to suck.
--HC
Thanks for your input HC, very interesting to see you had similar issues and how you solved it. The point that the problems happens only after some time make indeed the troubleshooting process more complicated.
On my side, i already replaced the Nano, and now measure the correct +5V as it should be.
I tried two options:
- use the USB from the solar controller
- added a 7805 to provide stable +5v from the load output.
Both seems to work fine for now, after 3 days of testing
What i realized as well, which was probably the source of a lot of problems, i was connecting the power using banana connectors, and they had a mechanical problems, some gaps in the metal terminals, making the connection unstable. I strongly suspect the arduino reboots and other problems coming from this.
I replaced this part, and will continue the testing a bit longer to be 100% sure.
1 Like
I have another question now 
For now i am using one of these classic cheap blue 12v solar controller between thee panel and the battery. (cf. photos in my other posts)
I am just realizing that they are using 16mA constantly.
My circuit (Arduino+L298+Relay+RTC) is using ~4mA in sleep mode, which i am ok with. I was wondering what would be the options to reduce the 16mA from the solar controller.
~20mA in total, with a 25V Panel can be a bit short for winter maybe? (center-east of france)
Or do you think this consommation is acceptable?
Disconnect it when it's dark out.
but then when disconnected, arduino has no power, so cannot control anymore any relay to conenct/disconnect the solar controller circuit ?
The voltage of the panel is not as important as the amount of current it can supply, and for how much time.
Are you needing to water plants in the winter as frequently as in the summer? If the pump usage decreases then that might more than compensate for the decreased charge time. Pardon my stupidity, as @kmin points out below, I got confused with another discussion.
Chickens care about that? 
1 Like
Oops, got this confused with another discussion I was reading about a plant watering system.
1 Like
It is for an automated chicken door 
So it needs to stay powered all the time.
The circuit itself has an acceptable low consumption (4mA)
the problem now is the solar controller itself (constant ~16mA)
The option could be maybe to switch to a TP4056 and 18650 Batteries?
And then use a step up to bring the current back to 12v for the motor part?
(schematic from this amazing project: RitonDuino: Porte Motorisée de Poulailler (1ère Partie) )
I found this solar charger for lead battery as well, with Quiescent Current <4 mA
Solar Power Manager for 12V Lead-Acid Battery from DFROBOT