A Little History of OpenMagnetics and Magnetic Synthesis

This week is somehow arbitrarily special, so I would like to tell you a little history of OpenMagnetics.

A hot July day I found myself coming back by train from Malaga to Madrid. I had visited Ezequiel Navarro , Juan Fernandez Reina and Claudio Ca?ete Cabeza in PREMO Group , where we spoke about power electronics, industry business, specialized talented people and above all, magnetics, over a few drinks in a chiringuito by the Mediterranean sand.

It’s funny how magnetics make people that work with them speak together in a way like few other technologies do.?

We spoke about the past, on how easy companies and vendors copy and paste designs and manufacturing concepts for all kinds of magnetics. Lack of innovative solutions and only repetitive customized solutions.?

We spoke about the present, current market opportunities, requirements from latest power electronics applications related to magnetics and of course, current design tools for magnetics. The famous “killer Excel sheet” came out, and the conversation pivoted immediately to how software and AI can help and improve the present bottlenecks.?

We spoke then about the future, discussing unexisting tools and systems that will help us to focus on how to innovate in better magnetics and avoid getting stuck on leakage inductance calculations, thermal performance comparisons with accuracy or parasitic effects of the windings.?

That conversation and the tipsy humming of the train got me thinking on how most of what I had done up to that point, which included hardware architecture and proprietary simulation products for my employers, had amounted to practically nothing.

I had taken knowledge from the electronics community just to increase the wealth of my boss and his investors, who, when the shit hit the fan, didn’t flinch on sacrificing me to save face.

I have wanted to create something for myself for a time then, a project where I could do things my way, make my own mistakes, and why not, get the glory from my work that has been going to my bosses before.

On that train, after that conversation and that whisky, both ideas, giving back to the community and doing my own project, clicked together. I arrived home, bounced the idea with my wife, and that same night I registered the domain OpenMagnetics.?

At that moment it was just a vision for creating an open-source repository of magnetics models, a place where the knowledge produced by the community in books and papers were reflected back in usable optimized algorithms for everybody to use for free. I didn’t really know what I was getting into.

In the next months I started building my idea little by little. First a data structure to describe a magnetic, then software architecture to process it. Core reluctance and losses models came after it. 3D models of the cores, and a webpage where people can access them. With time I finished the winding losses models (which was the main topic of the conversation on Malaga and ironically the first module I wanted to build) and the leakage inductance.

At that moment it felt right to start putting everything together and build upon it, so the Core Adviser was born. After that I attacked what I thought would be a low hanging fruit, the insulation.?

Several months of reading standards and talking with tape manufacturers placed me in the position of starting to build something that has been bugging around my head since I heard Bob White talk about it.

Why am I telling you this, dear reader? As that was the beginning and the path, today’s version is an important stone in the road that I wanted to share with you.?

Today the first open-source and free Magnetic Synthesis platform is released.?

Last night I uploaded the (hopefully) stable version of the Magnetic Adviser to OpenMagnetics, which is a tool designed to accelerate the development of magnetic components by proposing several full designs from the electrical requirements and the waveforms from our converter.

What does it mean? You just specify the insulation requirements you need to comply, the turns ratios and inductance, the voltage and current from your circuit, and in a few seconds you get several BAM (Best Available Design) designs with the automatically chosen cores, wires, coil distribution, margin tape and solid insulation; everything optimized for your inputs.

And of course, you get the simulation of every operating point you inserted: Core losses, ohmic, skin and proximity winding losses, magnetizing and leakage inductance; all with the best available models in the state of the art.?

And the best? every line of code that calculates all that is open-sourced, available for you and with verification tests for every model.

We say in Spanish “Es de bien nacidos ser agradecidos” so I cannot finish this article without thanking those who have helped me achieve this stone in the OpenMagnetics path, either by discussing ideas, providing kind words and support, or by offering me a job compatible with doing this on my free time (and may the muses forgive me I forget someone):

Ad Soetens , Ana Aparicio , Claudio Ca?ete Cabeza , George Slama, Ian Rensing , Javi Santana , Jochen Baier , Juris Vencels , Marcelo S. , Max Soriano , Nicola Rosano , Pablo Sánchez , Richard Blakey , Riccardo Tinivella .

And the most especial thanks to Noelia Jiménez Fernández , the One who held me together, kept my ideas grounded, listened to all my rants and transformed them into musings.

This was the beginning, but the end is far yet, much more to come!