Confusion making solar charger

Thread Starter

Dheorl

Joined May 24, 2017
12
I'll start with an apology as I'm sure threads like this could get annoying. I've tried to educate myself as well as I can, and will present what my assumptions are, but still have some confusion with my project and would like clarification.

Basically I'm doing what I'm sure many have done before, and making a battery boosted solar charger for smartphones etc. I'm aiming for a 5V, 2A output and have some questions about what best to use circuitry wise.

The most basic of questions is whether to go for a higher voltage battery (7.4V) and step down, or lower voltage (3.7V) and step up. I'm leaning towards higher as from what I understand it should be less straining on the battery as less current will be drawn, and the circuitry involved is slightly more efficient.

Second is what to use as a controller. My understanding is MPPT is the most optimum, but at this size not enough to make a noticeable difference for the increase in complexity, which means I should be looking at a linear charge controller? In all honesty there's more of these than I know how to differentiate between, but the ones from microchip or linear look hopeful, any particular recommendations? The panels I'll be using are fairly open at the moment, but will likely be along the lines of 3, 6V 2W panels wired in series (which I think eliminates the microchip options due to input voltage). Will the voltage difference between source and battery lead to problems/inefficiencies or should I be ok? N.B. due to the conditions this might be used in (hot cars etc) is something with the option for a thermal sensor a wise idea?

Lastly is how to integrate the load (in this case the USB out for the phone). I'm after what I imagine is the fairly standard concept of the panel powering the charger when possible (or as much of it as it physically can) whilst being boosted by the battery when needed, and then when their is excess/no load, the panel charging the battery, basically so at the output I always get as close to 5V, 2A as possible, with as little drain on the battery until it's dead. I know to an extent this will sort of take care of itself if wired in parallel, but I've read of various downsides of this, with a couple of solutions involving p-channel MOSFETs. Is the solution of bypassing the charge controller to the load via a single p-channel suitable or is a more complex solution necessary/better. I assume a voltage limiter before the load would in this case need to be capable or stepping the full solar panel output down to 5V.

Thanks in advance for any help you can give to any of the above questions. I hope they're not to inane, I've tried my best but am a bit nervous of going ahead without clarification.
 
Last edited:

dl324

Joined Mar 30, 2015
9,155
Welcome to AAC!
Basically I'm doing what I'm sure many have done before, and making a battery boosted solar charger for smartphones etc. I'm aiming for a 5V, 2A output and have some questions about what best to use circuitry wise.

The most basic of questions is whether to go for a higher voltage battery (7.4V) and step down, or lower voltage (3.7V) and step up. I'm leaning towards higher as from what I understand it should be less straining on the battery as less current will be drawn, and the circuitry involved is slightly more efficient.
You need to provide more information. What are the panel specs? What are the battery specs?

Stepping voltage up or down is not without cost. If you got 80% efficiency with switching regulators, two conversions would cost you almost half the power. Using a linear regulator to lower voltage will cost you about the same.
 

Thread Starter

Dheorl

Joined May 24, 2017
12
Welcome to AAC!
You need to provide more information. What are the panel specs? What are the battery specs?

Stepping voltage up or down is not without cost. If you got 80% efficiency with switching regulators, two conversions would cost you almost half the power. Using a linear regulator to lower voltage will cost you about the same.
The panels I'm initially planning on going for are the 6V adafruit medium ones (330mA) with three of them being connected in series. This isn't fixed though and I'm open to using other ones. In fact they'll almost certainly be swapped out for others in the future.

Battery wise I'm open to using whatever will with best to store the energy and pass it on to a 5V 2A output.
 

dl324

Joined Mar 30, 2015
9,155
The panels I'm initially planning on going for are the 6V adafruit medium ones (330mA) with three of them being connected in series. This isn't fixed though and I'm open to using other ones. In fact they'll almost certainly be swapped out for others in the future.

Battery wise I'm open to using whatever will with best to store the energy and pass it on to a 5V 2A output.
You'll give up 10-40% of your power by using an 18V panel to charge a lower voltage battery. You want to match the panel voltage with the battery you're charging as best you can to minimize wasted power.

If your using a 3.7V battery, it's most likely lithium based and requires a specific charging profile. Not following charging specs will decrease battery life.
 

Thread Starter

Dheorl

Joined May 24, 2017
12
You'll give up 10-40% of your power by using an 18V panel to charge a lower voltage battery. You want to match the panel voltage with the battery you're charging as best you can to minimize wasted power.

If your using a 3.7V battery, it's most likely lithium based and requires a specific charging profile. Not following charging specs will decrease battery life.
As I'm hoping to most of the time be operating the 5V charger directly from the panels with a bit of a boost from the battery, would it therefore be better to go with panels matched as closely to 5V as possible, so basically wire the 6V in parallel instead?

The main issue I've had with this is finding a 3.7V battery that can, when needed, provide the whole 3A of power required for long durations. I know they are out there, but they all seem rather bulky.

I have to admit I sort of assumed the various chips such as the MCP73871 or similar for linear automatically produced a suitable output for battery charging. Is there something else I need to add to the circuit to provide the charging profile required.
 

dl324

Joined Mar 30, 2015
9,155
We're stuck in a loop here. You want to design a system, but don't seem to know how to go about it. We can't help you until you have something for us to evaluate.

A key point to keep in mind is that few will design it for you. That's the equivalent of working for free and most of us don't do that.
 

Thread Starter

Dheorl

Joined May 24, 2017
12
We're stuck in a loop here. You want to design a system, but don't seem to know how to go about it. We can't help you until you have something for us to evaluate.

A key point to keep in mind is that few will design it for you. That's the equivalent of working for free and most of us don't do that.
Maybe I'm overcomplicating questions, but I'm not wanting anyone to design anything for me, I'm merely wondering about people opinions on:

1) Whether a higher voltage stepped down or a lower voltage stepped up is, in general circumstances, better
2) Whether MPPT is worth it for a system this small of if I should just go for a linear controller
3) If a single p-channel MOSFET is enough to balance a input/battery/load system

I guess perhaps these are just questions people struggle to answer more than I assumed they would.
 

dl324

Joined Mar 30, 2015
9,155
1) Whether a higher voltage stepped down or a lower voltage stepped up is, in general circumstances, better
Each step-up/down of voltage will cost you 10-40% of the input power. Whether this is acceptable depends on the application.

Matching the solar panel output voltage to battery voltage and matching battery voltage to the phone requirements will minimize power loss.
2) Whether MPPT is worth it for a system this small of if I should just go for a linear controller
It depends on efficiency vs simplicity. Will it make a difference for your system?
3) If a single p-channel MOSFET is enough to balance a input/battery/load system
You want a solar battery charger to charge a battery. You want the battery to charge smartphones. Why do you need transistors to connect/disconnect things?

Charge the battery whenever there's sufficient sunlight. Prevent the battery from discharging through the solar panel whenever there isn't. Connect and disconnect smartphones manually as needed.

Personally, I'd use a 6V lead acid battery and a diode to drop the voltage to something close to 5V for the phone. The phones must accept 5V +/10% because that's what a USB port provides.
 

Thread Starter

Dheorl

Joined May 24, 2017
12
Each step-up/down of voltage will cost you 10-40% of the input power. Whether this is acceptable depends on the application.

Matching the solar panel output voltage to battery voltage and matching battery voltage to the phone requirements will minimize power loss.
It depends on efficiency vs simplicity. Will it make a difference for your system?
You want a solar battery charger to charge a battery. You want a battery to charge smartphones. Why do you need transistors to connect/disconnect things?

Charge the battery whenever there's sufficient sunlight. Prevent the battery from discharging through the solar panel whenever there isn't. Connect and disconnect smartphones manually as needed.
Thank you for the answers. Is there a general rule of thumb power wise as to when MPPT becomes generally considered "worth it".

Unfortunately manually plugging and unplugging the phone isn't an option, and I'm hoping to bypass the battery as much as possible just having it there for a bit of extra umph when a cloud comes over, hence the thought of something extra.
 

dl324

Joined Mar 30, 2015
9,155
Is there a general rule of thumb power wise as to when MPPT becomes generally considered "worth it".
One metric would be when the cost of adding a MPPT solar charge controller gives more power gain than adding an equivalent dollar amount of additional panels.
 

Thread Starter

Dheorl

Joined May 24, 2017
12
One metric would be when the cost of adding a MPPT solar charge controller gives more power gain than adding an equivalent dollar amount of additional panels.
Thanks, suddenly seems obvious :)

I guess extending on that logic if there is a space constraint then MPPT is always worth it.
 

dendad

Joined Feb 20, 2016
3,084
I personally would go for a 12V battery. And a step down reg.
But you will need some sort of battery management to prevent overcharging and excessive discharging. Otherwise you will kill batteries quite rapidly. That could be a reason to use a MPPT controller as they usually include this protection.
That said, I made an irrigation monitor that has a 6V solar panel and a 6V 4.5AH SLA battery, and just a series diode to prevent discharge through the panel, and a power zener across the battery to prevent over charging. I did not go to 12V as this is a low power device.
There is no over discharge protection in this one so that could be a problem.
You really need to work out the total power drain over time, and the solar availability, then scale the solar panels and battery to run for maybe 3 days without charging to cover your use.
That is all rule of thumb calculations, but will be a start.
 
The first problem is the USB charge controller. The USB port has to ask for how much power is available from the 5V source. USB3 will change things. Otherwise the phone will only think 100 mA at 5V is available.
The 2A takes some hoops to go through to get there. Then there's charging only ports.
The early way was to put different resistors on the data lines.

So, you have to provide 5V and let the charger do it't thing. You also have to tell the phone at what current the 5V is available so it can charge the fastest.

So, probaby from a practical point of view, you want a wide input voltage and the ability to tell the phone how much current is available.
 
Last edited:

Thread Starter

Dheorl

Joined May 24, 2017
12
I personally would go for a 12V battery.
Thanks for your input. Why would go for a 12V battery over a 7.4V battery?

From what I understand of them charge controllers such as the MCP73871 I mentioned earlier will control a suitable profile, but this was one thing I wanted clarification on as it's not similar to anything I've used before.

As for the overall numbers, the project is very space constrained and the solar availability and power drain will both be very intermittent, so I'm sort of just going with the best I can get to get as much power into it in a time span as possible.
 

dendad

Joined Feb 20, 2016
3,084
The 12V is so you can use an SLA battery. As for other battery types, I don't know the charge controllers to use, and some of them are critical to use one otherwise fires or dead batteries are the result.
I would look first at the total load per day you require, running on the battery. Then hunt for a suitable battery and charge controller, then select the solar panel to suit that. But most solar charge controllers are designed for Lead Acid or Sealed Lead Acid batteries.
You may get away with the MCP73871 but you do need to work out the power requirements first. The MCP73871 has a limited charge capability so may not be big enough.
 

kittykite

Joined May 26, 2017
3
Hi so I can tell you O.P. that most Lithium Cells are 3.7V, so the 7.4v cell you're talking about is just two cells in series.

What you need is a basic solar regulator... that has adjustable parameters. A 3.7v battery when fully charged is 4+v (lets say) and when drained is 3.# V (i'm not sure)... they usually have a chip that reads the voltage and when it drops below 3.#V... the device stops drawing from it not to ruin the battery. SO your solar regulator will be set to stop charging at 'max 4+v'... recover charge at 3.#V or w/e

float, discharge stop and reconnect should be adjustable.

I'm not sure what you mean by solar booster. But the regulators job is to make constant the incoming voltage of the solar panels... which obv will vary based on the sun etc And you choose the regulator model (10A, 20A, 30A) based on what devices you're running through it.... a phone maybe draws 1A max... same with digital cameras.. I guess.

gl
 
Top