Maximum Power Point Tracking


Thread Starter


Hi —

I'm developing the power management system for a new concept of mine.

I'm looking for the easiest way to regulate 3.3V from a battery and charge it using an MPPT. I'll probably use gallium arsenide triple-junction cells, which will have a power consumption of under 20ma at 3.3V. It needs regulated voltage for the atmega 328p au chip and other IC's. The cells have an open-circuit voltage of 2.62V and a max power current of 14.6 mA.

I see that the SPV1040 might integrate MPPT circuits. I have to charge a 3.7V cell (approximately 500Ma).

The thing is, I have a pretty rudimentary understanding of charging circuits. Is there any efficient MPPT/Battery charge controller IC's on the market or easy to design and existing devices? I know that form factor is important since this will be going in a pretty small space.

I'd like to hear your thoughts. And if you'd like to virtually collaborate, let me know and we can get connected.

MPPT only makes sense if you're charging from a source that captures power from a fluctuating external source. Winds, Solar, Waves.

MPPT used from a battery could optimize the fastest charge, but it would not maximize the total energy taken from the battery. The battery would also likely get overly warm.

I think it might be more helpful if you share in detail what you need to do?
@Steve869 —

OK I see now. I could have explained this more precisely.

I need to charge a 3.7v battery with solar cells and then output 3.3v for the satellite's components. The space in this project is very minimal and I have NO electrical engineering experience.

Pls help.
So, MPPT requires you to have a switch-mode DC-DC converter controlled by a microcontroller.

It will continuously tune the DC-DC converter to achieve the highest charge current. A more complex circuit may alter the behavior based on battery voltage, temperature, or state of charge.

In your case, you should look for a more integrated solution. Is there a device designed to do all this in a single package? My first search would be for these items on DigiKey or something.

I have been thinking about this and believe I can provide more information for you. I haven't heard back yet, but this could help more.

Producing a 3.3V rail, the simplest approach is a linear regulator. That is going to generate more or the most heat — that will end up becoming wasted energy.

... But if you intend to keep the electronics warm, it could be useful.

You should also consider the dropout voltage of your regulator and maybe how much your 3.3V rail is permitted to sag (minimum battery voltage).

An inverting type of DC-DC converter can get you around some of these issues, but you also lose a common ground rail.

I assume you're talking about a cube sat, and in that case, for maximum performance, there will be a lot you have to balance.
Sorry for not seeing your help earlier.

It will be a pocketqube that is 5x5x5cm, so just a downscaled CubeSat.

The pocketqube will then launch other smaller satellite boards, these will not have a battery and only require a regulated voltage from 2 solar cells. The

Temperature is generally not an issue since all the chips on the satellite have been tested to low temps and can withstand the temperature fluctuations.
The 3.3v regulated voltage on the pocketqube itself is not an issue since the battery is fairly large and the idea is to keep it at more than 60% every cycle to reduce wear since it will be expected to do thousands of cycles a year, solar power is also more than enough considering the pocketqube itself will just be transmitting a simple beacon consuming around 50mw after having launched the other satellites.
Also, if this would help in anyone following this thread ...

Satellite board space is even more minimal and the solar cell produced just enough power for all the electronics considering the linear voltage regulator I previously used.

Does anyone know of a more efficient way — maybe a buck-boost converter. The SPV1040 chip would seem perfect for the job. Hypothetically I could just wire that up along with a battery protection circuit, right?

Help is really appreciated.
Switching regulators are typically more efficient. The larger the voltage difference the larger the potential efficiency gain.
Maybe something along these lines —

TI's seems to have better efficiency. But the difference is minimal.

The SPV1040 could be used? The board sats will just have direct regulation from the regulators and the mothership will charge the battery with the SPV1040 and an over-discharge/overcharge etc protection and then output through a regulator.

Thanks again.
There are always problems in using railguns to orbit anything. Atmospheric friction is one of those unsolved issues. What goes up into orbit eventually comes back down, and re-entry into the Earth's atmosphere burns up unprotected objects as their kinetic energy is converted back to heat.

Obviously the same problem occurs when accelerating an object to orbital velocities through our dense atmosphere. It is possible to reach the required velocities by accelerating in a vacuum, but after exiting the railgun aerodynamics comes immediately into play.

Anything with moderately deep pockets could build a picosatellite launch facility and locate it anywhere in the world. Give it a good 15 - 20 years.
There are definitely still some issues to work out here — like you insinuated. But, considering the fact that this kind of technology is huge in the military sphere, I would expect this to also move fast.

It is truly interesting that yes, as you said, simply getting out of the Earth's gravity is what requires the most energy!
If not already.

But low-earth orbital velocity is around 28,000 km/h or nearly. That's why the railgun "barrel" needs a vacuum inside while the sabot carrying the pico-satellite is accelerated.

Probably will need considerable length too, compared to current military railguns, to limit the acceleration to a survivable amount.

I'm not really familiar with what state-of-the-art integrated circuits can tolerate, but 1000 g would not be an unreasonable goal, maybe 10,000 g for a shorter acceleration time. I do know that the smaller the circuits get, the more tolerant they become to acceleration. The idea of it all is just so intriguing.
Hey, Julio ...

Have you gotten this sorted out yet? Me and some friends are working on a similar project and wanted to know if you want to collaborate. This thread has already gotten so long that I didn't want to wait much longer.
I think he could use a TPS6210DCNR or an AP3015AKTR-G1.

And I don't know enough about either or haven't used either enough to be ultimately an expert but ...

Switching regulators have a theoretical maximum efficiency of 100% (voltage in * current in = voltage out * current out, noting that current in will differ from current out). But there is also a variety of issues that can bring losses.

In comparison, the efficiency of a linear regulator is essentially determined by the inverse of the ratio between the input and output voltages.
Well, yeah. Switching regulators have a theoretical maximum efficiency of 100%. In comparison, the efficiency of a linear regulator is essentially determined by the inverse of the ratio between the input and output voltages.
If you aren't getting your power from a solar panel, then the regulators should work just great for you to operate a circuit from a low voltage.

You should know, they probably aren't ideal to charge a battery ... Maybe for a single charge cycle.