A few years ago, [Lukas Fässler] needed a solar charge controller and made his own, which he has been improving ever since. The design is now mature, and the High Efficiency MPPT Solar Charger is full of features like data logging, boasts a 97% efficiency over a range of 1 to 75 Watts, and can be used as a standalone unit or incorporated as a module into other systems. One thing that became clear to [Lukas] during the process was that a highly efficient, feature-rich, open-sourced hardware solution for charge controllers just didn’t exist, at least not with the features he had in mind.
Data logging and high efficiency are important for a charge controller, because batteries vary in their characteristics as they recharge and the power generated from things like solar panels varies under different conditions and loads. An MPPT (Maximum Point Power Tracking) charger is a smart unit optimized to handle all these changing conditions for maximum efficiency. We went into some detail on MPPT in the past, and after three years in development creating a modular and configurable design, [Lukas] hopes no one will have to re-invent the wheel when it comes to charge controllers.
28 thoughts on “High Efficiency, Open-Sourced MPPT Solar Charger”
Hi David Langridge at Solar Focus, Perth, Western Australia ;-)
The first IIRC some 20 years ago to pursue a tight analog design
in a narrow market place relying on only one outlet not market savy :-(
Good experience where a keen designer left all marketing and sales
to someone who knew nothing of marketing and sales, perhaps the
designer was far more interested in alternator design finding
another marketer and sales person not skilled either…
Oh so many lost opportunities, so many permutations not addressed…
Suffice to say, by all means design, be aware of the markets and how
to progress don’t automatically assume others you don’t know are good at
anything just by mere claim.
What is it about alternative energy people and “cannot compose a coherent sentence much less a paragraph”?
Jesus christ man, I’ve read this three times and I still have no idea what your comment says.
I’m not native speaker :) and I thought that my lack of English is a reason I can’t understand the comment above. Thanks man!
im dumber for reading that. now ill never build again.
Excellent work, looks very professionally engineered and made.
This would go perfect with the DIY solar panel HaD article.
“The various interfaces from SPI over I2C to USB make it easy for your application to communicate with this charger. ”
Considering were MPPT’s are usually located, I’d go with a different interface.
What interfaces do you think suit better and why?
Certainly USB is good for this project, considering its’ small size. And he has it plugged directly into a smartphone. I think it is a good choice for what he is attempting to achieve.
If you were really keen I’d go for an RF link, or RS485 for a wired connection. But it would be a bit harder to just plug your smartphone into!
I don’t want to be negative… but an efficiency of 97% ?!?!
I would like to read more details about how that was achieved, first the calculations and then the technical details. Why do I ask you may wonder? Well 97% is pretty darn good and considering the wide span of 1..75Watt well that almost sounds to good to be true. The project page even states “Efficiency > 97%”, that almost approaches the efficiency of a bare wire and then doing some conversion as well. I would like to say “hats of, project well done”. But for now I remain skeptical.
As I could not find any serious information in how this was really achieved (and/ or measured). It read like a tell-sell commercial, trying to scare the viewer of with technical terms that hold no real info. Was the high efficiency DC-DC controller made from scratch or was the design (and 97% claim) copied directly from the datasheet of a regular DC-DC converter that only does 97% when switched off? The project page just isn’t clear about that.
Please tell us more, so we can learn, seriously, this is potentially interesting stuff.
PS: regarding efficiency and losses, sometimes even the cables towards the electronics can do a few percent. I assume these were left out of the equation as the project focuses on the converter and not the complete installation.
Well typically efficiency numbers are for the ideal conditions, not worst case. For power supplies it’s often in the 50-75% utilized range. Higher or lower gets you worse efficiency.
I looked at the schematic and the inductor is 100uH (or 68uH in another revision).
I am starting to wonder if I am simply misunderstanding the meaning of MPPT.
I have a solar setup for camping. The highest load device is a 240V fridge that runs from an inverter.
I first bought a solar regulator like this that even had MPPT in the model number (MPPT30). It was 30 Amp.
I have a 240Watt 24V Panel (36V) and four 12V batteries (13.8V), total of about 300Ah. With the MPPT30 I would get 6.5 Amps from the panel at 36Volts and 6.5A to the battery at 13.8V
I then bought a proper MPPT regulator and it now I get 6.5A from the panel and about 17A to the battery at 13.8V
Clearly the first regulator was just PWM output instead of being a DC-DC converter.
Inside the units, the first one had a very small inductor and the DC-DC converter of the second regulator had a large inductor.
The one in this article has a very small inductor.
So am I misunderstanding the meaning of MPPT?
To the layman, looking for an inductor is one way to differentiate between a PWM controller and MPPT. It’s not the complete story though and you’d need a proper look at the topology.
The size of the inductor is related to the switching frequency. If you can operate at very high frequencies, you can reduce the inductor. Switching frequencies usually follow power levels so below 100W it’s possible to operate in the 100’s of KHz. 100uH is not unrealistic for that frequency.
By the look of the project page, it is doing genuine MPPT.
There is an example there that shows the volts/current input and the volts/current output of the gadget indicating that it is doing proper conversion (ie: the high input volts get converted down, while low input current gets converted up).
It would seem the efficency is a result of the measurements done internally (by the micro) of the input power (Vi x Ii) and the output power (Vo x Io). Accuracy is probably not huge, but still pretty good.
But this was for a specific “screenshot”, and would not apply across the whole range. It is also for a fairly modest in/out conversion ratio, and fairly modest power throughput too.
Quite reasonable figures, and not all that surprising (at least to me anyway).
I’m really fond about technology. But I’m also pragmatic.
Sometimes, the best component is not always the best. It’s like using high quality components for solar panels with the best results vs cheap components to obtain acceptable results.
I read somewhere in the net an (australian) article about cleaning your solar panels. The OP said that cleaning is not free (you pay somebody to do it or you increase your personal risks – falling- if you do it yourself) and you need to do it every year or sooner. It’s better (money spent) to purchase bigger solar panels to minimize the power loss of dusty panels (and because rain will eventually clean this for free too). I found his comment clever.
It’s also like imported food vs local. There are always hidden costs.
I remember a guy at school who calculated the cost of “green” recyclable glass bottle vs plastic (PET or PVC). And he made a survey of what people thought. Most of them forgot all the hidden costs (cleaning the bottle means hot water, detergents, weight and truck loads from the store to the cleaning company and return) that increased the cost of green washed glass bottles. And at the end, PET (because no chloride molecules) was far better than glass bottles about the total energy spent.
Anyway, nice work from him. Really
A *very* nice project, congrats!
on sad side:
Who want to bet on time it takes a way to ali/bab/xpss? Dealer asks 6-8 weeks for v0.1 with SOIC ARM on board, place your bets players.
I built an mppt – maybe one of the first – in the 80’s. I almost hit high 90’s efficiency using synchronous rectification (I had both N and P power fets…), 24v output and a very special hand-wound inductor, mostly analog with a PIC controller to do the “hunting” for the MPP and control for 3 stage charging. I’ve been off-grid since around ’79.
Jan is right – with the work that took it wasn’t going to be a commercially worthwhile project as it was. Cost too much.
It wasn’t high 90s over the whole range of possible inputs – really dim sun wasn’t that hot due to the usual tare, and at the high end the usual I^2R issues. As part of the project I also “chose wisely” re the panels (then > $6/watt). It was amazing how difficult those last few percent of goodness were – it took a few revs to get there, and even then I was no beginner at pwm power switching. Cruising controls wound up with a version very close to mine (then the name changed to blue sky, and then…well, I bought Trace/Xantrex once I built a much bigger system (now Schneider?).
Names all changed a lot in that business as players came and went or got bought.
Buying second-hand panels that are (were) once part of a domestic roof top installation has become really cheap, but these panels are usually in the much higher wattage range… 150W and up. 250W being pretty common.
I would have liked this unit to be able to handle these bigger panels, but yes, cost would go up.
Also, after reading the project page (without digging down into details), the main output can be 6 to 13 volts. I would have liked a top limit of 14 or 15 volts (so I can charge a “12 volt” LiPO4 pack properly).
A question I will be putting to the devloper/s, what will happen if I connect a higher wattage panel? Will it simply current limit to 75 watts, or will the smoke be released…
I have half a design somewhere for a charger based on a TPS3405(I think that’s the number) LED driver. It’s almost perfect, so long as your source has an isolated ground because it’s low side sense. And it just so happens wall warts and solar panels do, unless you’re getting fancy with wall wart splitters going to other things.
Only 1.5A, but I haven’t seen any other way to build something with that small of a parts count.
I should actually build and test that thing one day…
Next project idea for Lukas: Same thing but 2kW-ish and 48V nominal batteries for a full-on off-grid house..
Is it not the case that by the time you get to 2kW panels, the variability of the panels becomes an issue, or some get dirty, or shaded etc, and you’re effectively MPTT’ing the lowest-output panel?
I think you ideally want an MPTT per panel, but big enough for the larger 150-200w domestic panels.
See solaredge power optimizer. $75 problem / challange – comms are closed source
is possible to make device:
Can I carge my phone AND inthis same time put power from solar, panel, AAA battery, fruit power, dynamo and 2 different accumulators 7.2Ah i 20Ah (different type)?
the trouble is differens source of power not next device
You mean like this one https://hackaday.com/2018/07/29/vampire-charger-is-a-rugged-anything-to-5vdc-converter/ that was featured on these pages only a couple of days ago…?
Please be kind and respectful to help make the comments section excellent. (Comment Policy)