thesolarpowerexpert.com

The Solar Power Expert Blog

Introducing the v4 Charge Controller

by Chris - March 21st, 2012.
Filed under: Charge Controller. Tagged as: , , , .

A member of the free charge controller mailing list pointed me toward Tim Nolan’s charge controller design. I took several days to really study his design. It incorporates the same design ideas as previous versions of the Free Charge Controller, but his design is much more simple and elegant. Not only is the design of the buck converter robust, it also solves many of the practical problems mentioned with earlier designs of the Free Charge Controller. To be more specific, it solves the following issues:

  • The design is built on top of an Arduino, which means it can leverage the large user base and code base that the Arduino platform has built up over the years.
  • The voltage and current monitoring circuitry and buck converter circuitry use a minimum number of components and are built on top of an Arduino protoboard ‘shield’. This means that an expensive, custom PCB is not nessessary. This further mitigates the problem in open source hardware projects by making it easier to get hardware into the hands of people who are interested in contributing
  • Because a protoboard is used instead of a custom PCB, it’s easier for experimenters to try different hardware – such as different coils for different voltages, or bigger coils and MOSFETs in order to push more power through the system.


Developing a Kit

After finalizing the v2 design, I spent a few hundred dollars to have custom PCBs manufactured and purchased parts to build 10 kits to sell. I quickly realized that building kits is not a trivial process. It’s expensive, time consuming, and easy to mess up. Not only that, the process made me realize that I need to devote my efforts to the areas of the Free Charge Controller project where they are best used: namely design, documentation, and promotion. The functions of kitting, pricing, logistics, and customer interaction, while very important, are not areas that I am particularly proficient in or represent the best use of my time.

For that reason, I reached out to Jameco, a well known distributor of electronic components. They were very receptive to the idea of working with me to build a kit. There was a bit of work to convert Tim Nolan’s original design, which was based on Digikey parts, to find equivalent components out of their catalog. They were willing to stock parts they otherwise wouldn’t and were very flexible in working with me to develop a kit. Two weeks ago they sent me the first prototype kit and I have spent the time building and documenting the new design. We are collaboratively working through the final bugs introduced by the hardware changes.


v4.01 Free Charge Controller Schematic

As you can see from the schematic below, the design is very similar to Tim Nolan’s original design. The theory of operation from his hardware notes is completely the same. (Note: the component designations, like R1, in his notes refer to his schematic, not mine.)

Charge controller schematic

v4.01 Free Charge Controller Schematic

You can also download a pdf copy of the charge controller schematic.


Plans From Here

The development kit that Jameco will soon have available for order is a huge step up from my original efforts to build a kit. For one thing, the elimination of the custom PCB and many components cuts the cost to about half. The cost of my original kit was around $120, and the kit from Jameco will be in the ballpark of $60 to $70. That’s almost a 50% decrease in price, which I think will prove a big difference in getting developmental hardware into the hands of contributors.

While my initial efforts at documenting the build process and creation of basic software will be a big help to those attempting to build it, I’m really looking forward to when ten people participating on the mailing list have built one. At that point, we’ll be able to discuss an optimal layout, building strategy, and write documentation around that. (While my personal build notes will be published here on my blog, the ‘official’ build instructions will reside at freechargecontroller.org so that everyone can participate in the documentation process.)

Another huge advantage the community can enjoy at that point will be fine-tuning the design. I envision the present design to be able to handle 12 volt solar panels on the order of 30 to 60 watts. Even if we can collaboratively improve the efficiency of the design, there will be physical limits in terms of voltage and current. I look forward to when we can begin to agree on the realistic specifications that we want to the design to meet. Once that is accomplished, we can then begin to finalize a printed circuit board (PCB) layout for the creation of an Arduino Shield – in order to make the assembly process even easier for people to build.

One reason I’m so excited about this development kit is that it will allow people to easily ‘hack’ the kit for customization. Do you need a charge controller that can pass 20 amps? Add bigger MOSFETs and a bigger coil. The development kit will be a perfect springboard for experimentation and expansion. I envision many different versions (forks) of the free charge controller will spring up from this design.

11 Responses to Introducing the v4 Charge Controller

  1. Hello, very nice. I just want to say that the board should really be hackable to support 20 amp panels or 24V 10 amp panels, because such panels are common nowadays and for smaller loads, the commercial controllers are cheap enough.

    just my 2 cents

  2. Thanks Andy! You are right that they are common, but they aren’t standard. Unfortunately, there are no standards when it comes to solar power. Configurations at 12, 24, 36, and 48 volts are common as well as at many different amp loads. However, 12 volt panels are still very popular with both off-grid and grid-tie systems. My article, Choosing Your (Solar) Path In Life, helps explain some of these concepts a bit better, and it has links to articles for more information.

    The original specification for the Free Charge Controller was for a 12 volt, 60 watt panel. It will be hard to accommodate panels of much more power than that due to current capacity of the wires used on the development kit. However, adjusting the design to handle higher voltages is very practical. It will take testing though.

    While the price for the development kit makes it significantly more expensive than some commercial controllers, the idea is not to compete with commercial controllers. This development kit will appeal to people who value hack-ability, customization, open source IP, and repair-ability over cost.

  3. This is fantastic! I’m sitting down right now with your V3, Tim Nolan’s, and a few others schematics. I’m making a few modifications to Tim’s design myself like adding reverse polarity protection, LVD, and battery draw current monitoring. What are you using for IC1 rather than the MAX4173?

  4. We’re using the LTC6101 IC from Linear Technology.

    Jameco was willing to stock this IC in the SOT-23 package (which is easier for soldering), just for our kit. Isn’t that awesome!

  5. Hello, I’m glad to know that there is a kit to develop prototypes. I have to make a car battery charger powered by a solar panel for a training school for auto mechanics. I must try
    Is it possible to know when the kits will be available from Jameco?

  6. We don’t really have a date set. The good news is that we’re really close, however we’re currently waiting for a part manufacture to stock Jameco with a part needed for the kit. Once they get the stock, they can send a piece to me and I can verify that the circuit is working correctly. It’ll be a few weeks. The last thing I want to do is have kits sent out and have it turn out that you don’t have the right parts.

  7. Wow, this is great.. There are lots of cheap power converters, and there are lots of high-quality more expensive converters, but nothing I can fix (or maybe modify to directly produce a regulated 48V).

    Long term, I think we have to get to the point where we can come up with a single chip silicon solution that has switching, sense, and high-side drivers all integrated into the same package… which is why I decided I needed to develop and promote the Open Silicon power converter concept.

    The same experience in power converters is useful for load applications.. like DC lighting and motor drivers, which is what I started out looking at.

  8. Trying to do MPPT with 12v nominal panels is self defeating. It’s much more practical to run higher voltage arrays (100v or so) by putting panels in serial, and have 24 to 48v battery banks.

    Steve, a offgrid “real” solar power expert, designer, and installer

  9. Hey Steve,

    Actually, the controller is designed to run with 18v nominal panels, then drop the voltage down to charge a 12v battery. You’re right that it is more efficient to run higher voltage arrays, however high voltage is not safe for a DIY / hobbyist type circuit.

    The real idea is to create a simple, working, foundational circuit upon which other people can then alter and experiment with to obtain greater currents, higher voltages, better software, etc.

    Cheers!

    Chris Troutner

  10. I think to start with it should be able to handle 70v for it to be a true and worth while mppt project. So you can at least input 10 amps on a off grid 24/48vdc battery bank. Not to compete with commercial manufacturers but to actually get a worth while charge on the batteries. I would be interested in helping

  11. I have looked I think just about everywhere, and can’t seem to find the code for the MPPT. Do you know where I can find the lastest version?

Leave a Reply

Your email address will not be published. Required fields are marked *