The Libre Solar project started in 2017. It was an already existing set of electronics developed by Martin Jäger, who began in 2015. It all started with a Arduino based charge controller for photovoltaic systems. It was back on Arduino base with low entry and later developed to a next generation of electronic charge controller with a STM 32. The higher demand regarding resolution and calculation time required the upgrade of the hardware component.

Further we built the Libre Solar around or with that component and started prototyping. There were different prototyping phases of different components. It was always developing one component and testing it. Based on the test we implemented new requirements. An example would be power classes and how to get a higher current out of the system. There were different development stages from the component. Now it’s a set of hardware components used as charge controllers. Besides, it’s a different set. It’s for battery management systems and a modular setup, in case you want to use different types of batteries. You need a different type of components, for example, a battery management system for lithium ion batteries, if you want to use them.

It was developed in Hamburg by Martin Jäger, who’s the main developer and founder of the Libre Solar project. He mainly developed this by his own in the beginning. Later, people started to kick in, mainly in the firmware development. First the hardware was development of the PCB electronic layout and afterwards the firmware. In the firmware development was another contributor who kicked in mainly concerning the modularization of the software. We have one main firmware repository where we can modularly choose the different type of hardware and the firmware adapts to that.

The main benefit from the Libre Solar project was learning about about power electronics. Martin Jäger doesn’t come from the field of electronic development, he’s a mechanical engineer. For him, it was tinkering around and learning about power electronics and software development. It’s mainly the benefit for him as a developer. When I jumped in concerning the educational resource, the benefit was to have open source resources for redesigning new components, like our Libre Solar Box where we can develop an application out of the electrical component from the Libre Solar project. This was the benefit from my point of view as a student. The benefit for the people who are using the component is to build it by themselves. We have test fields of friends who are living in rural areas. They build up the off grid energy system completely by themselves. For example, there were users in South Germany who rebuilt the charging sensor node in a rural area. Besides, we have research projects. As example, in Rwanda in Africa, we built up a test system with these components. The benefit there for the people is that they can run an energy grid system with open source hardware components. They are now able to redesign it by themselves. But the knowledge level is still needed to know how to adapt and how to redesign it. There’re different layers of benefits.

The first development stage was voluntary funded. All the development was made free. At the education project we derived open educational resources out of the existing design documents. This was funded by the city of Hamburg by the Hamburg Open Online University. This was one funding pot mainly for creating educational resources. And then the research projects. Another funding was a research project from Scotland together with partner startups who are interested in this field. This was financed by a research fund from Scotland.

I think this was from Scotland.

I don’t know the size of the overall funding on the project. I have numbers for the educational part. It was around 70,000. It were two funding rounds where we got money. For the Rwanda project I have no numbers.

With our student group, we replicated the Libre Solar component, it was a community based workshop run in a FabLab. In the planning process, we got a bill of materials and saw what components we needed. We divided the replication into PCB replication and the actual solar box replication. Within PCB replication, we decided between to order a full assembled PCB or to order the circuit board without parts on it. It was like the PCB without components. With the ordered components we went to the Fab Lab and did the usual process steps to do for electronics. Afterwards it was about testing to see if all the components were assembled in the right direction and currency. Then we flashed the PCBU. Finally, we had a ready to use charge controller.

So the development process for the PCB mainly took place in GitHub, as as a collaboration platform. The repository is placed in GitHub. For the electric design, I’m using KiCAD as an offline development tool, to design the component. And then, you placed ready design files in place in Github, and make some comments. We used the Git workflow, you have there and yeah. So this about development, using the tools on GitHub and then iterating.

The electrical design is mainly developed by Martin. I have to look up if there’re some other comments. Especially in electric designing, hardware is designed without many contributors because you need technical knowledge. It’s like the usual software kit workflow where you have other people contribute, make a pull, push or merge requests or give comments. That’s a typical good workflow.

About development process issues, I can’t tell much because Martin is the main developer. I know there are issues regarding contribution guidelines. Especially for the firmware, you stick to a code style but different people are working with different styles and the code gets a little bit unstructured and messy. For people to learn the code, it’s important to have a common structure. This was one case of a contribution guideline for the hardware design, but I can’t say too much because I’m not too much contributing in it. If you don’t have proper documentation, people don’t have a good starting point where they can jump in the development. A proper documentation is crucial.

The same counts for modular development that people can contribute to specific modules or sub-components of your hardware. From the development and production side, the main issues was safety, quality and secureness. When you work with batteries some people have been insecure on how to handle the safety issues. You can’t shortcut them otherwise they will explode. There you need expertise and a guide how to trust the people you’re working with. Some of us had an electrical background. The people, who didn’t have the background in electrical engineering were more insecure about handling batteries.

This was one issue besides the quality testing. For quality parameters, there was not so much documentation done. About this, we depended on expertise.

We had Martin as the main developer with the main expertise. If he wasn’t there, it would have been harder to do the quality test.

Martin made most of the decisions because he had the main expertise. What he said should fit. Other people had expertise which been discussed. For the educational part, I was the project leader on the administration side but I didn’t make any decisions. It was made in a group of seven people by consensus.

It’s been mainly electronics (PCBs).

I would call it more a set of products like a portfolio. You had different components for different requirements. You could mix them to build up a system out of it.

No, it’s similar to charge controllers. One product is like DC-DC converters and charge controllers with different types depending on power class. The second type is a battery management system which have different types depending on how many cells you want to connect. The third are communication gateways to connect to server-based applications.

First was the electrical classification.

No, the panels, batteries, cable and plugs were sourced. Everything was sourced. There was no mechanical frame. We had a solar box. This is one application but it’s not the main component of the Libre Solar project, it’s in our student project. The solar box has a mechanical frame we can consider. The Box is one application you can make. Firmware would be another class.

We declared on our website the different development stages. I don’t have it all in my head. If you look here, you can see the development stage declared for each product.

Yes, definitely. We had a maker space in South Germany who replicated one charge controller completely by themselves, we didn’t even know about. Afterwards they sent us a link to the forum when he discussed it. They replicated it and used it for a sensor node as well. This is one independent replication of the component.

We have a forum in the project. Within the forum, you can see a lot of people who are interacting and asking questions, because they replicate the components by themselves. I don’t have an exact number.

The hardware development of the electrical layout is mainly done by Martin Jäger. I have to look in the GitHub[*] if there are more contributors. I’m doing the educational part. I’m designing the resource contents for education. Katherine, who was in the Open Hardware Summit, works in the educational part because she’s in the library and does a little bit of software for data visualization. On the data software side are contributors, but we aren’t in a local group here in Hamburg. It’s distributed in the world.

[* Notes from the open make team: not many more contributors spotted on GitHub, although the projects are often forked.]

Yes, everything around the hardware components within the Libre Solar project.

My wish is to build up Libre Solar as the Arduino for renewable energy systems. I always wanted to have an open hardware base for learning and building up its own renewable energy system for everyone. It’s for industrial, private or educational use. We aim for a ready to sell product and we are nearly there. It’s just because of the CE certification. We already did a few tests in the laboratory for measuring the CE standards but we were out of range and didn’t get the CE certification. That’s why we can’t legally sell the products, but it’s a goal.

Right now we have the supply shortage of micro-controllers which causes the prototyping process to hold, because we can’t source STM 32 micro-controllers. In the beginning, the goal is to have a set of hardware components ready to sell with a portfolio for educational reasons and research institutes. There was a connection with the University of Freiburg and different other universities are interested. This is a low entry point because you don’t need the proper certification if you just use it in research environment.

This would be one channel. For example, we think of sensor nodes for agricultural applications, which we test here in Hamburg for loading stations for mobile farm. It’s energy supply, we can use it for nearly everything which needs electrical current.

The hardware is just a load for us.

It’s not the main goal. I’m a researcher now. I talked with Martin about the publication of one paper, for example, from Joshua Pierce on Hardware X. There you can publish open source hardware projects, but it’s not the main focus.

Not specifically, our feedback or the issues is that we make an iteration of the documents and republish the documentation. If we see that there is something missing within the documentation, we redevelop the documentation and publish it. But it’s not explicitly a publication about the issues. We’re thinking about a blog. If you have a blog post, you can say: ‘here is something we made’. It’s a little bit like issue tracking using GitHub. While we were in a replication workshop, we didn’t write down the issues. We just make an iteration of the documentation.

Yes, no publications.

I could have a look in the repository. It’s mainly for the hardware with common information like design files, CAD files and interactive bill of materials. We didn’t provide an assembly guideline for electronics but we made a guideline how to source the components. It’s like our educational resource website where we put in everything concerning assembly processes.

We published it only on GitHub. There is linking to the wiki from Open Source Ecology in Germany. The key entry in this platform is describing a so called Open Nano Grid. This would be another application you can make with hardware components from the Libre Solar project.

GitHub was like the poster in this time, Martin Jäger choose it. I think it was the most common collaboration platform for software development and electronics. It’s really comfortable because you have a version controlling and all this DevOps functionalities behind it.

The starting point is a little bit tricky. For people, who come from outside and aren’t aware of software, it’s hard to interact with GitHub or Git-based repositories.

No, I think it wasn’t a high quality notes document that we had. We didn’t have the time to make it in a qualitative documentation to publish.

Yeah, that’s one of the main main components and maybe the amount of contributors. If you have more contributors who feel more safe with the thing you develop, you will get more feedback and you can optimize the product better. When Martin Jäger is the only developer here, he always doesn’t know if the design is the best design. The feedback on hardware design is crucial besides a lot of contributors.

It would be easier because now you’re developing and you maybe know that the quality is quite good but you don’t have the direct feedback. Because we don’t sell the product, we don’t have a customer environment feedback process. For now, there’s not so much qualitative feedback for the design.

From the independently replicated parts we don’t have any qualitative feedback. I can just assume and you can analyze the forum.

It’s easy for people who have a technical or software development background. We did the entry workshop about how to use Git. (Yeah, okay.) A colleague of ours did a little workshop. The graphical interface of GitHub makes it much more easy using it and it was like trial and error or working with it together. It helped to focus on the main features. With time the people got used to it. It’s like a learning curve for platforms.

Yeah, it wasn’t directly to GitLab from Open Source Ecology but to the wiki from Open Source Ecology. (Open Source Ecology Germany is running a wiki). There they can maybe just search for the entry. It was like one article where they mentioned the Libre Solar project. But we are documenting it mainly on GitHub. Other people, who are reusing the component, are documenting it and the key focus was ecology. They helped to disseminate or promote the project in a way. Some people informed others that there’s this component you can use to build up an open nano grid. This was the system they introduce. Therefore, they link to the Libre Solar project and say that this is the open hardware component you can use for building up such a system.

We provide information on different platforms to draw more people for us and the Open Source Ecology. That project pays to show the interlinking or the collaboration between the organizations. Indirectly, it’s to draw more people and to reach out from the different organizations.

I wasn’t involved there, this was before my time. Martin Jäger attended at the POC 2021 meeting in Paris. I don’t know how he got there and under which circumstances. He was asked for something that I don’t know. He got to know different people from this from the Open Source Hardware community and they’re still connected there.

The main clustering was the hardware, education and maybe the website. Then it’s three or four if you want to consider the forum as an information base. It’s like the hardware information. This was mainly the repository you have on GitHub. The hardware information is consisting of the design files, BOM file, readme file and an overview. This is the technical documentation of the hardware. It’s one main information cluster. For all the components, a separate or a dedicated repository exists with these informations.

The second information cluster is the learning cluster. It’s the open education resources. The main idea behind it is to give a basic information about the used technologies within a system. It’s divided in two chapters. One is a system view to have an overall view of what we are here talking about regarding the nano grid system, renewable energy and PV battery. The second part is the component part. It’s to dive in into the components. For example, how is a charge controller built up? What does a charge controller contain? It has a power electronics, measuring and communication stage. I’m clustering it in the sub-components of the hardware. The idea behind it was that people who want to contribute in the development have a starting point of what do they want to develop. If you have an electrical engineering background, you don’t need this because this what you learn in university. But if you don’t have and you’re interested in renewable energies, you can use the educational resources to get the basic overview.

The second part in the educational part is the production. It’s about how to replicate and produce such components. I’m oriented on the life cycle process. We have development and production stage and now the usage stage. Each component has a user manual. For learning environment, user usage and documentation it would contain the data acquisition, data analyzes of energy systems and improvements. But this isn’t yet developed. The fourth stage would be recycling or upcycling of such components.

The forum was a typical forum where people can add questions and threads. It’s very good because you see people contributing in this topic. There is one about the development of a new charge controller and this was a very active thread where people give their feedback to it. It went on for some time. It results in a new development of a hardware. This shows that different people come with different requirements and make up their mind for the development of a new 20 ampere maximum power point checker.

It’s made with static site generator called viewpress and a git repository. There’s a framework and broad usage. Within viewpress are different content blocks. This forum is based on this course an open source forum platform.

GitHub is the main backbone of all information and data. It’s linked to two other front end websites.

Yeah, everything is published. For example, the main development of the battery management system, which got funded by energy access, is within one repository. Energy access funds open hardware project in the field of renewable energy.

The creating of more content is limited by the amount of people, time and financial resources. Everything is on a voluntary basis.

Successful is the ready to use prototype we developed. I would declare it a prototype because of this legal issue of the CE certification and not call it a product. It’s ready to use and people are using it with off grid applications all around the world. We have a feedback loop from the research project how the system is running: It’s running. This is the main success. The documentation is a success, too. In my opinion, it’s a very high quality documentation of all the components, it’s very structured. We were able to develop educational resources within such a hardware project which isn’t typical. The contribution in the hardware design is improvable. This is still a gap. It would be nice to have more contributors dedicated on the electrical hardware design. We had some reviewers which was good. We need to know how to improve the EMC within the electrical component. There was a review but not an active contribution to this.

A review was there because it was from friends. We had the direct contact. It wasn’t on a GitHub sphere or distributed community sphere. It would be nice to have a review and improvement suggestions for the design. A complete redesign isn’t the goal behind it. If someone would redesign it and make a suggestion, that would be very cool. But it’s not the main goal.

They don’t contribute in the electrical electrical layout. I think, the users are mainly people who need the current and want to have renewable energy. The main reason is always the software part. You have open interfaces where you can extract data out of it and you can control data out of it. You have the operating power for the system. You can use it for your requirements. If the current flow is not 99%, these users doesn’t care about if they see a failure. Till now, they wouldn’t contribute to erase this failure to improve function.

They’re building it by themselves. I think because the documentation is there. They’re replicating it. We had one hardware where we made a little batch of 100 components and this was sold. Now the batch is empty. We had no CE certification for one component, therefore it was only for educational and tinkering purpose. Another example, at the CCC Congress we had around 10 and we sold three out of it. It were Universities from France and UK. The third I don’t know anymore, but there were people who were interested and said to buy a new one, because they have a research topic and it costs like 50 or 60€. There were distributed server applications and they wanted to test how servers can run on renewable energy systems. This was the perfect use case for them because they can extract all data out of the charge controller and they can correlate with their process data. They are still running on those parts and sharing with us information.

If we would have people who manage this feedback and community work, we could get more qualitative feedback out of it.

At this time, I was a member of the Open Source Ecology association in the field of open source hardware community. I was attending some lab meetings. I was definitely convinced by the idea of open source hardware and I was intrinsically motive interested in renewable energy systems. I studied electrical engineering with the focus on automation systems and therefore my interest are energy systems and the automation aspects behind it. When I moved to Hamburg, there was one member of Open Source Ecology who said to me that there is the Libre Solar project and Martin Jäger, the developer, lives in Hamburg. This was a coincidence. Besides I found a call from the open Hamburg online university to contribute to a project of open education resources. Afterwards I made a concept to build a box which I’ve seen in the Open Source Ecology wiki. It was a solar box. But it used different types of PCBs of solar charges. At this time, it was a solar charger developed by a Canadian developer. So my idea was to take this solar box and take the Libre Solar components out of it. I already had the electrical expertise to design this a little bit in my head.This was when I called Martin Jäger and introduced him to this idea and he was interested. That’s how I jumped in. It was mainly within the development of open educational resources.

It was a funding project of the Hamburg Open Online University. They had a funding round and they were very interested of students who contribute. At this time it was mainly professors, researcher and staff. I was like a student which they were very interested in. This time I got funded. It was like 25,000 Euro. It was like the first funding project for me. I was a completely inexperienced with such things. I was just a student.

My professor was the mentor because that was needed from the Hamburg open online university. I was a student and the project leader of this funding project. That’s how we started. In this project we didn’t pay any people. This was my idea to have people who are intrinsically motivated who would join without money. We took all the money for buying hardware, to rent fab labs and the machines and producing a promo video. It’s a little YouTube video of introducing open source hardware. All the money wasn’t spend on people but on external resources.

The second grant we get was used to pay student assistants. They created the content. Our idea was bigger and it was to much to write this content all in our free time.

It wasn’t directly to the Libre Solar project. With the beginning of this open educational resource, I started a student group with my professor as a mentor and we called the student group “collective open source hardware”. „Cosine eight" was the name behind it. We were the seven people in the student group. We mainly replicated the Libre Solar hardware and redesigned the solar box.

My old supervisor, or professor, wasn’t in the content. He wasn’t really part of it. He was very interested and he gave me all the possibilities. I started to introduce this project in all courses within the university. I jumped in 20-25 courses and introduced it to the students with his help because he introduced myself. It made it a little bit easier to introduce my project because he said it’s a good project and you should listen to it. In the group we were students and Martin Jäger has a PhD degree. He worked part time in the automotive industry and part time developed the Libre Solar project. The others were all mainly master students from electrical engineering, renewable energy systems and mechanical engineering. In the Hamburg open online university we had some communication designers who did the logo and other icons. They designed it for us or helped us to design it. They were employed at the Hamburg Open Online University. It was a service they gave us for free.

I was introducing it to all the courses. It acquired them by myself and some of them I knew already. With two of them I had direct contact and the others were interested after introducing it. They wrote me an email and told to be interested.

We had two students who made their student project work in this area. They took a solar box and made the testing. They tested all the parameters of the solar box, for example current versus saturation. But it didn’t went so good to be honest because they didn’t document anything. It was for them to do it but not on a qualitative level. They just did it for getting the credit points without the open source community documenting. It was a project work they had in one course.

No, it were only two of them. They said they had to do a project work. They asked if they can do this project work in this field. For myself, I didn’t get any credit points because we didn’t have a project work where we could use this for. But I did my master thesis in the Libre Solar project. I asked my professor if I can do my master thesis there. I introduced the topic, it was about control systems

I think, Martin Jäger’s was happy that people were interested in his project. In terms of community reach out, this supported community building for him. The main idea of the student group was to get more students in the field of open source hardware. In 2017, no students knew what open source hardware is. This was in my field, I was very pity that no one knew what open source hardware is, especially engineering students, because there’s so much potential behind. It’s still the case. I think, this was for Martin a motivation because we talked about it. He said that’s a good idea to disseminate the idea of open source hardware within engineering students. Some students perceived a benefit, they could get credit points because they had this project work where they got credit points. For the others, it was fun. You’re playing football and others are tinkering in a Fab Lab. For some, it was the first time of manufacturing. For example, I studied electrical engineering and I’ve never picked and placed a PCB. I’ve never saw a PCB production. It was in theory taught but we never experienced it in the laboratory. It provided a learning effect. Application oriented learning I would call it now.

Martin Jäger got a little bit of this money because he has this high education level and he put a lot of time in it. We said to Martin to write a receipt because he’s working so much in this field. Even though he didn’t want it to but we said that’s money for your work. Besides we did the video I mentioned. It was a video creator who got paid for doing information videos.

In the second phase we got students paid as student assistants and paid students for creating the content.

We had two. I think they worked for a year or a year and a half. There was Corona and maybe it was extended by half a year.

All of this was at the Hamburg Open Online University.

The coordination of the development was quite easy because Martin was the only developer at this time. The software development coordination went through Git. The research projects had a typical research project coordination with Martin as a partner. But I don’t know the exact constellation of the research project he worked with. For our student work the coordination was made by me. We had a telegram channel where we had ad hoc communication. We built up a GitHub repository where we had issue-based coordination. This was a little bit of our project management behind it. Besides, we had weekly meetings at a Fab Lab. This was every Thursday, it was meeting in the evening with discussing and thinking about the design and requirements. It was in the Fab Lab St. Pauli.

Martin was there before us. His first prototype and first testing system was build in the Fab Lab St Pauli. At the Fab Lab St. Paul was a solar panel charging a battery. This is why we went to the Fab Lab St Pauli, too.

The university didn’t gave us a room. We asked for it. They brought us then in an old laboratory, where we could hang around. But this isn’t a good place to work. Then they build up a Fab Lab but it wasn’t open for students. We asked for it but it’s just for start-ups and some ecosystem share stakeholders.Further, there was a Fab Lab with 3D printers. But it wasn’t a Fab Lab where you can like tinker around, it was just for 3D printing. The university didn’t have the infrastructure for us. That’s why we went to a Fab Lab.

We were seven people.