EEVblog Electronics Community Forum
Electronics => Projects, Designs, and Technical Stuff => Topic started by: ElectricPower on November 26, 2023, 09:14:58 pm
-
I have an TP4056 board with protection i want to use with an 6V 10W solar panel for lighting up one spotlight like one of theese in this package:
https://www.aliexpress.com/item/1005005361773301.html?spm=a2g0o.productlist.main.5.1ed257c4BxAaUp&algo_pvid=12c7e79e-df01-4514-973b-2fcbf168e8a8&algo_exp_id=12c7e79e-df01-4514-973b-2fcbf168e8a8-2&pdp_npi=4%40dis%21NOK%21382.90%21237.35%21%21%2135.70%21%21%402101f01817010329199405632ef6c1%2112000035203870291%21sea%21NO%21901862616%21&curPageLogUid=dNMyH79d40uK (https://www.aliexpress.com/item/1005005361773301.html?spm=a2g0o.productlist.main.5.1ed257c4BxAaUp&algo_pvid=12c7e79e-df01-4514-973b-2fcbf168e8a8&algo_exp_id=12c7e79e-df01-4514-973b-2fcbf168e8a8-2&pdp_npi=4%40dis%21NOK%21382.90%21237.35%21%21%2135.70%21%21%402101f01817010329199405632ef6c1%2112000035203870291%21sea%21NO%21901862616%21&curPageLogUid=dNMyH79d40uK)
I want to add an bigger solar panel because winters her in Norway is long and dark and the days is short.
I want the solar light off when it's daylight when battery charges, and on when night and running on the battery. I have testet the cirquit with an 5mm 0.1w diode in ky worshop and it works regards to the light turning on/off, but i have not testet the charging part.
Can i simply use an p-channel mosfet like the chematic under?
-
I think your circuit will work. You might need a high-value pulldown resistor on the gate so it won't float, but that probably is not necessary if there is some current flow from Vin through the TP4056 to ground when its not charging, which there probably is.
This may not switch cleanly. There will be a period as the sun is going down or coming up when there could be a mixture of charging and spotlight activity, or even a period when neither is taking place, but I assume that doesn't really matter since it will only be for a short time.
If you have a series of cloudy days, you may not get much charging done during the day, and the lights will only stay on for a while at night. I don't know that there's much solution for that. Anything you can do to minimize the load current will help greatly with that.
-
I think your circuit will work. You might need a high-value pulldown resistor on the gate so it won't float, but that probably is not necessary if there is some current flow from Vin through the TP4056 to ground when its not charging, which there probably is.
This may not switch cleanly. There will be a period as the sun is going down or coming up when there could be a mixture of charging and spotlight activity, or even a period when neither is taking place, but I assume that doesn't really matter since it will only be for a short time.
If you have a series of cloudy days, you may not get much charging done during the day, and the lights will only stay on for a while at night. I don't know that there's much solution for that. Anything you can do to minimize the load current will help greatly with that.
Thanks for fast reply :)
I can consider to use to 18650 in parallel and a bigger 6V panel lat delivers enough amps easier.
But the lights i linked to uses only one 2.200mAh 18650 for each lamp and the solar panel is 6V and around 3W. That will work fine in sunny areas and in Norway when it's summer. But maybe not perfect when it's winter with 12 hour darkness and 12 hours light and when it's light it is only 50% chance for at the sun is shining.
So i need to concider worst case scenario witch is let's say two weeks with light but cloudy days. I want the 18650 cell be fully charged from 0-100% with 12 hour cloudy but light days. I do not know if a 6V 10W solar can do that it real life.
-
If you're in Norway, your latitude is something near 60 degrees North. You won't get anything close to 12 hours of daytime in the winter. You'll need to make the lights as dim as possible while still working.
The NPD6020P mosfet in your drawing has been discontinued. But you might still find them online. I don't know of a good substitute in a TO-220 package.
-
If you're in Norway, your latitude is something near 60 degrees North. You won't get anything close to 12 hours of daytime in the winter. You'll need to make the lights as dim as possible while still working.
The NPD6020P mosfet in your drawing has been discontinued. But you might still find them online. I don't know of a good substitute in a TO-220 package.
You are right. My fault. Its daytime between 10:00ish to around 15:00ish. Så around five hours of daytime. I think the lamps i linked to in my forst post draws about 150mAh. With two 18650 in parallel (fully charged) thats no problem, but i wil need an powerfull 6V panel that can give full power (about 1 amp or more) in daylight without sun to charge the batteries to 100% in 5 hours.
I do not know if the NDP6020 is a good choice here, but i have a lot of them. If you know a better alternative please give me a link to it :)
-
I think the NPD6020P is an excellent choice - so long as you already have them.
Winter time in the high latitudes limits the usefulness of solar power. Do you have any alternative for powering the lights, or charging the batteries?
By the way, you need to read up on low temperature limitations for charging lithium batteries. I think below 0C you have problems. So you could have bright sunshine, but still be unable to charge unless you warm up the batteries.
-
Okay, i have concluded that i'm going to use NI-MH batteries instead and leds that draws less power. 18650's is not a good choice for norwegian winter. I live in the hottest part of Norway but still -10 to -20 degrees celcius is pretty normal on the coldest days. But normally it stays hotter than -10 celcius.
-
With NiMH, the circuit can be extremely simple. I just don't know how much light you will get, or for how long.
-
I ordered some PCB's from JLCPCB for mounting of that P-channel mosfet that i soldered to TP4056 controller.
But when i connected the battery and a 6V input from my bench PSU and a small load (500mA) the load stayed on all the time when i turned on and off the imput voltage which should represent a solar cell input.
Can anybody see any rookie mistakes here?
Under is my shematic and some pictures
-
I don't understand your schematic. Where is the solar panel connected? What does the mosfet do? And I don't think a TP4056 will work with NiMH cells. So I don't really understand what's going on.
For NiMH circuits, look at this site:
https://www.homemade-circuits.com/simplest-automatic-led-solar-light/ (https://www.homemade-circuits.com/simplest-automatic-led-solar-light/)
Edit: Ok, so the mosfet is to switch off the load when the solar panel is providing power. If you apply 6V to the gate, and 4.2V to the source, the mosfet should be off. Is it possible you have the mosfet pinout wrong? If you have the source and drain reversed, the body diode would conduct even if the mosfet is off.
Edit2: No, the mosfet looks right on the PC board, unless you physically mounted it backwards. But if you didn't do that, I think it should work.
-
I use 18650 here.
Yes, i think the mosfet is mounted correctly.
Solar is connected to TP4056 input via the PCB i made.
I do not understand why the load is on when i apply power (6V) to the solar input.
When i tur of the solar input, i can se an 0.5V voltage drop to the output (load) but it's not turned off....
-
Can you measure the voltage on the gate, drain and source when input is off, and when it's on?
Your mosfet is NDP6020P, not NDP6020 (which is N-channel)?
A small voltage drop like that can mean current is flowing through the body diode when the mosfet is turned off. That would mean the drain and source are reversed, but I don't see that in your PC board.
-
Ok, I found a TP4056 module and an NDP6020P in my junque box, and breadboarded your circuit. It worked fine. When I turned on 5V input, the mosfet turned off.
-
Thanks for ansvers :)
I reaaly do not know whats wrong. I will measure the voltage on the tree pins lter today. The description on the mosfet is "NDP6020P".
I think i have to breadboard it so i can test a little easier.
-
I think the tab on the mosfet is connected to one of its pins. If the back of the mosfet is touching the board, maybe some unwanted connection is being made. Well, just grasping at straws.