Small scale MPPT & BMS IC wanted (think ~1-2W)

[Krakonos]

Hi!

TL;DR: What is the jellybean li-ion solar charger IC?

I have this idea in the back of my head to build a nice little solar power bank. Nothing fancy, just 1W cell (possibly 2W if I feel generous). Something like the stuff chinese sell you, but actually working. I really like the idea of putting the cell onto my bag (on the lid, for example), and let it work for a day or two. This could be really useful for long trips, where I won't see electrical outlet for a month or more, and don't want to carry a lot of batteries.

I ordered some thin film solar cells from ebay, and will test if they are any good (is there any better source with reasonable pricing?). But now my question is how to get the pixies from the cell to the battery :-).

I searched online for some time, but most of the stuff is for large installations and buck topology. I have found the BQ25504, which is kind of what I want, but it's QFN only, rather limited current, and pretty expensive (which makes mistakes in soldering QFN even more expensive).

I even bought a cheap chinese charger/bank, to see how they do it, and perhaps buy a chinese IC from aliexpress and such. Well, they don't. The panel is connected via SS14 diode (if I can trust the markings) directly to the battery. I wonder if it's enough, given there are enough cells in the panel, and the maximum power point is approximately in the range of battery + diode drop voltage.

I'm already thinking how to do this on my own. Some micro, a boost converter (I really like the MCP16251/2, but the current capabilities have little reserve), and PWM the output from boost converter (set to target voltage, 4.2V) via a mosfet and proper filtering. I could implement a crude MPPT like this, and with some tuning it could start gathering electricity even under very little light. Depending on the solar cell voltage, I could even insert a bypass mosfet (some MCP boost converters already have it) and turn off the switcher if the voltage on cell is higher than the battery AND the battery is not fully charged (but I'd rather not to, as there is the need for extra ragulator, for the end of the charging curve).

But before I do that (I can't really devote much time to it right now), I wanted to ask here.

Is there any IC/technique I'm missing, that would make my life easier? How is it done in the commercial world? What are your suggestions?

Thanks!

PS: I have used the search field in here, and found one post similar to mine, but with no results at all. Everything else seems to be related to high-power stuff (or possibly scammy kickstarter).

[uncle_bob]

Hi

What exactly is your "load"? The output of the gizmo needs to be designed to work with whatever you are trying to charge.

What is the environment for the cell(s)? Unless you live in a very unusual part of the world, a "person mounted" cell is going to get less than ideal light.

How much power do you need / want / use? If the cell puts out 1 AH at 3V over a day, is that useful to you? Do you need 10 to make it worth doing?

Lots of variables.

Bob

[Krakonos]

Hi!

I'm not really sure if the questions you as are that relevant, as I specified 1W cell, which kind of implies the answers... But in short:

1) Expected load is to pump the power over to my phone or even NiMH batteries for my gps, in case of emergency.
2) I live in Central Europe. This is intended to be used in remote parts of the world on the travels, possibly tropical or subtropical. I'm expecting at least 8-10 hours a day of good sunlight.
3) 1Ah at 3V over a day is more than plenty. I would be pretty happy if I could charge my phone's battery (~2.3Ah, Li-Ion) once in 5 days. Remember, it's not primary source of power.

But again, I can do the math myself. What I really need is the amount of power I can expect from the cell. I don't think directly connecting the cell to the battery is really optimal solutions, that's why I'm interested in more sophisticated solutions.

[uncle_bob]

Hi

Ok, well, if the cell puts out full power / full voltage for 8 hours a day and that meets your needs - just put a switching regulator on it's output. Since the loads you list (phones and GPS) all have battery chargers / managers in them, there is no real need for anything else.

Bob

[Krakonos]

Hi!

That would kind of work. The downside is that in non-optimal conditions, the phone probably would not charge, or charge intermittently, and I doubt that is a good thing to do. Nevertheless, I would have to balance the solar panel's power for optimum gain, so I won't carry more than I can use, but also don't operate the cell below the optimum power point. This seems like a really unreliable and inefficient setup. That's why I asked for MPPT controllers in the first place.

Btw. I just got the thin film cell. But it's raining all day long, so I haven't tested it in sunlight yet. I'll wait for some good weather to see if it's any use.

[uncle_bob]

Hi!

That would kind of work. The downside is that in non-optimal conditions, the phone probably would not charge, or charge intermittently, and I doubt that is a good thing to do. Nevertheless, I would have to balance the solar panel's power for optimum gain, so I won't carry more than I can use, but also don't operate the cell below the optimum power point. This seems like a really unreliable and inefficient setup. That's why I asked for MPPT controllers in the first place.

Btw. I just got the thin film cell. But it's raining all day long, so I haven't tested it in sunlight yet. I'll wait for some good weather to see if it's any use.

Hi

The cell's output is primarily a function of the light input. The rest of the variables are normally pretty minor. Unless you have a tracking system to keep it pointed at the sun, just let it do it's thing.

Bob

[Krakonos]

I have done some testing with the solar cell and it's power output is way below the 1W (more like 0.2W), so it will go rest into the bin, until I decide if it's worth the while. Thanks for you answers though, I'll keep it in my mind while thinking about further work!

[uncle_bob]

I have done some testing with the solar cell and it's power output is way below the 1W (more like 0.2W), so it will go rest into the bin, until I decide if it's worth the while. Thanks for you answers though, I'll keep it in my mind while thinking about further work!

Hi

Your experience is pretty typical. The power ratings you see on many cells assume an amazing amount of light ....

Bob

[suicidaleggroll]

Hi

Your experience is pretty typical. The power ratings you see on many cells assume an amazing amount of light ....

Bob

Not really...at least not in my experience.  They DO assume you're running them at the optimal I/V combination though, which requires an MPPT.  A simple DC/DC feeding 5V out with whatever the phone decides to pull will not cut it.

OP - when you did your testing, did you bring a high power pot, programmable load, or a big set of power resistors to test the entire I/V curve?  You should get something that looks like this:

If you happen to test it with the wrong load, you'll get a very different power number than what's possible.

[Seekonk]

One cell is a joke, but with a 3W or higher 12V panel a nifty power point controller can be made with a cheap 2596 type module and a transistor with a couple of parts