Power Electronics Profiles: The Hardware Engineer

Why is hardware interesting?

Working in power electronics in the age of AI is like making paintbrushes in the Renaissance. With our converters and power supplies, we will give AI the power it needs, as we already give so many things for others to make their works of art. Within power electronics, control and firmware are becoming more and more prominent and sophisticated, so this hardware stuff can seem uncreative. Seems like the ultimate, doesn’t it?

However, power hardware is also very interesting. But not because technology is advancing in this field, which is also the case, but mainly because the hardware engineer is in direct contact with reality, which is more creative than the best artist. Electronics is almost never very difficult, but it is always complex, because it involves many components and variables, which communicate and monitor each other. Moreover, in power electronics, it is more difficult to modularise, i.e. to create totally independent stages that are decoupled from each other. They could affect each other: heat up, cause interference, overloads, failure modes that are difficult to foresee, or leave no room for each other, even touch and shunt or damage each other’s mechanical integrity during a short-circuit or by a bad installation. So, by mere combinatorics alone, there are many ways for your power electronics to fail. And the reality, be sure, is always aware of all of them.

In each project, he will pick out a few, the most instructive ones, to show you that you don’t know how to design. The trick in validating a converter is to recognise them in time and correct them. Therefore, the hardware designer learns much more from the unexpected things that happen in the lab and in the field, than from the best research paper or the latest semiconductor model. Whether a converter turns out well depends on the work you put into it, but it also depends on that attitude and the goodness of chance, just as it does with a great work of art. Because, frankly, we hardware guys often have no idea why our hardware works. We tend to see lots of ways it wouldn’t, only luckily they don’t materialise. The rest of you would be scared if you knew how many things we miss, the solutions we improvise and the ones we come up with by chance or due to a mistake.

I suppose the same thing happens to painters, and what in the end appears to be a stroke of genius starts out as a blur, which the landscape then helps to correct. A design engineer’s job is never a jack, horse and king, but I suppose some can rely a little more on regulations, design guides and theoretical calculation and simulation models. They are like a painter of English gardens: realistic, but peaceful. The power electronics paints jungles: the least important thing is the painting and the most important thing is the leopard two metres away. The design error, in other words.

There are always many of them lurking, but you can’t complain because reality is almost never so cruel: it usually moves the leaves a bit, so that you notice it, if you are attentive. Years ago, an experienced designer told me something that I would always remember: When you see an anomaly in your prototype that you don’t know how to identify, stop and take the opportunity to study it. If you let that opportunity pass, it will appear later on…, when you have many units installed in the field. That’s why, in power electronics, the engineer has to pay attention and spend many hours testing the converter he is developing. He should be camped next door, on the test bench.

A firmware engineer who has been working with the same micros for 30 years would not have an easy task in adapting to the latest control platforms. A digital electronics designer from the 1990s would also have a hard job developing high-speed buses like today’s ones. However, an outdated power hardware engineer would have comparatively less trouble going from designing with thyristors to designing with the latest semiconductors. There are more guarantees that he or she will be able to adapt and do it just as well. Similarly, the designer of your mobile phone charger could move on to developing megawatt converters, or the other way around. He will not be up to date with the components and their physics, the topologies, solutions and typical design criteria, the instrumentation, the final application and its pitfalls. He will have a lot to learn, and he would not be able to lead a new design for years, but the most important thing he would already possess: the habit of dealing with physical reality and its occurrences. Physics changes with scale, but its quirks do not..

The hammer does not work for the blacksmith

On the other hand, one of the dangers of technology is that it makes things too easy for us. That tools do the work for us, think for us and we settle for it, believing that everything works when it doesn’t. For some reason, in power electronics that doesn’t happen. You usually get the measurements wrong or have reason to question them. Simulation models are always missing something, even more if you expect them to solve everything in a hardware problem. And the most important calculation method is the rule of three. In the end, sophisticated calculations and simulations, whether they are Simulink, Spice / PLECS, or FEM, are primarily used to feed intuition, which is what we then use to accept their results as correct. These simulations and calculations are made with simplified models of the real system, but in order to simplify reality, you have to understand it intuitively. No tool is going to avoid this effort, which is iterative:

When we move to prototyping, it is to add the measurements to the loop and iterate again: measurements > intuition > calculation > simulation > redesign > measurements >…

In fact, if the prototype works right the first time, it is a bad sign. That’s because you haven’t looked hard enough, and it will fail later, when there is a lot of equipment deployed in the field.

Another feedback from reality: the trampled ego

Everyone in the field of engineering makes mistakes and suffers setbacks. However, compared to other areas, there is something very special about power electronics: it contradicts you faster. And it usually does it out loud and in public. As with the rest of electronics, it is comparatively easy to set up your own set-up to test many of your ideas. In fact, you can learn at home with inexpensive and varied projects. But potentially, if you’ve made a mistake something will do very strange things, burn out or even explode. It’s certainly not as bad as a bridge falling down, but everyone will know about it. Every day, the power electronics make it clear that you’re not as smart as you thought you were. It’s relentless feedback that keeps coming back to you. However that’s exactly what keeps you awake, trying to improve on what you did wrong… as soon as the upset wears off. It’s a great filter for the overweening ego. Maybe that’s why I meet very few difficult people in the profession, let’s call them….

At times when serious design errors appear in a project, having good colleagues is crucial to support and resolve the situation. You won’t care about silly things like looking bad for being wrong, or not being right, or your colleagues suggesting better solutions than yours. Moreover, sometimes they will help you, sometimes you will help them. Instead of seeing them as a threat, you end up appreciating working with people who the more competent are, the better.

The trap of the details

In a company we all have ideas and suppliers, starting with the chair or the notebook. We have to pay attention to detail, and we make mistakes, sometimes with uncalculated consequences. The designer of a converter has to be very involved with his people, because electronics requires attention to detail. A very small component can bring down a whole piece of equipment and the installation it powers. The power designer is like the manager of a company where every detail is taken care of. In the end, everything works well thanks to that inexplicable ingredient that appears when, together, things are done with dedication.

However, as a great industrial master taught me, the best is the enemy of the good. To make the best equipment you have to take small, realistic steps. In fact, it will only come after years of projects and releases. That’s why the hardware engineer has to be able to give up on perfection all the time. He can always keep thinking about every detail, and there will always be some other, even better option, but he has to make design decisions and move forward. A good idea implemented well is better than a good idea implemented poorly. That’s why you have to spend so many hours of your available time testing and polishing the mistakes that matter, from the obvious ones to the hidden ones. But you can’t try to correct all the mistakes you can imagine – you have to intuit which ones are going to appear in reality, pay attention because others will appear, and not chase the rest. If the leaves move, there is usually a leopard, but the leaves have not always moved, even if it seems that way. Nor can you insist on imposing your ideas, however good they may be, or on understanding every last subtlety, such as the curious physical effect in turn, but nothing relevant to the application or topology. Electronics is designed with black boxes, but in power electronics the boxes are grey, and you can’t waste time you don’t have trying to clarify them all.

Often the idea, the design and the components are great, but they are simply not good enough because they are not polished. Other times none of the three is anything outstanding, and yet they get the job done. They work acceptably for their purpose, because they weren’t great to begin with, but the major mistakes have been rectified and the rest are no problem. We don’t always fully know why there are design flaws that never appear. You can’t calculate; at most you can take notes. For some reason, it is the same with hardware as with people and organisations: a design does not work well when it has no mistakes, but only when it has received enough care.

Trainee of everything, master of nothing

In power electronics, the hardware designer is first and foremost an awake ignorant person. Alert in case the leaves move, as we have already said. But also ignorant, because he has to know a little about a lot of things, enough to be up to date and to listen to the experts on each one. His job is to use limited knowledge to work with each component supplier, to explain to his specialists what he needs, and to realise if they have misunderstood or even got it wrong. Knowledge is not everything and is not an end in itself, but as if it were. The designer should want to know everything.

Here is a list of the skills:  

– Power semiconductors and drivers

– Passive power components: inductances, transformers, capacitors, resistors…

– Analogue, industrial and digital cards and electronics

– Voltage, current, temperature, pressure sensors…

– Electrical design: protections, switchgear, cables and plates, electrical panels and cabinets…

– Electromagnetic compatibility

– Thermal and climatic conditioning

– Mechanical layout design

– Error analysis

– Laboratory work, measuring instrumentation and tools

– Technical and scientific ‘general culture’: machines, electrical installations and networks, control systems, batteries, RF, industrialisation, notions of firmware, physics, chemistry…

 – Power electronics applications: photovoltaic equipment and parks, wind farms, automotive, railways…

From all the skills listed above, none of them is superfluous for the power electronics hardware engineer. Every day you will be jumping from one to another, and you will have to study many of them again to remember them, catch up or not make too much of a fool of yourself. The good thing is that the learning curve is exponential: at the beginning, with little effort you learn or remember a lot, unless you are bad at it or can’t find the information. But after a certain level in a subject is when you start to relate the ideas within it, and then another curve starts, that of a deeper understanding. Up to that point you had really just parroted the ideas you were told, thinking you understood them. Unfortunately, for the poor hardware engineer, it’s usually all in the initial phase: he or she may enjoy the first step, but hardly ever goes up to the second one.

However, to be in hardware, you need to know a little bit about each of these areas, and at least one of them, even if you will never know as much as the specialists. The reason of course is that they are not from your company and cannot do all the work for you. In fact, as we have already said, their job is to link different areas so that everything fits together in the final design.

Diverse profiles

The problem is that it is obviously impossible to be very good at everything. Therefore, within a converter development company there must be other hardware designers who are just as good at the areas in which you are weak. In addition, we have already talked about the importance of having a lot of eyes dedicated to spotting mistakes during development. Those eyes will be hardware, but also validation, control or any other speciality. In power electronics, ‘synergy’ and ‘diversity’ are not just fine words: just as knowledge adds up, so does working with people of different backgrounds, skills and experience. It adds up when you least expect it, even without realising it. Many ideas come out of unconscious imitation of something someone told you or that you have seen somewhere else, in some gadget that might not even have anything to do with a converter. The most important place in an engineering company is the coffee machine, and if it is a power machine, even more so. Around it debates and technical and personal anecdotes with the most varied casuistry are heard. This happens because the professional and personal interests of the designers are usually very varied. Among the hardware designers, I have met people with all kinds of hobbies and ways of being:

 – One was a model aircraft builder and always had a helicopter in the boot.

– Another tinkered with the car and tuned it.

– Another made a hot-air balloon that was never heard of again after it was launched.

– Some are very academic, others never read a research paper.

– When they studied, some were very nerdy and others just barely passed.

– I have worked with two hardware engineers who first did vocational training and then went to university. That gave them a lot of experience.

– Some of us are very “sciencey” and need to understand everything. Others are very engineering minded: they know what solution we need to go towards even without understanding it in detail.

– Some of us think in the abstract or see “the invisible” better: electric, magnetic and electromagnetic fields, eddy currents, maybe the pressures in a cooling circuit, or where air and heat go…. Others are good with “the visible”: they have a good feel for where to put each component inside the converter, or how to integrate it into the final system.

-Whether they know more or know less, whether they are aware of it or not, they all love physics. Power electronics is about tuning physics and seeing it in action in front of your eyes.

para programar sus ocurrencias.  Some look like freaks, others look like normal people.

– We’ve all heard the saying “you have to know or have the phone of the one who knows”. In power electronics, we all have our phones. 

– Some know very well what the converter they are making is for, others just look at its insides. Everyone has the same amount of fun.

– Some are very handy, others are not, but all of them love to experiment at work. Some of them at home too. It doesn’t matter if he can’t fix it: for the hardware engineer, a broken washing machine is an opportunity. On Monday he’ll go to work dirty and soaked, but more technically literate.

– Nevertheless, the hardware engineer is a bit of a divo. The mechanical designers will have to find a way to locate their components and the firmware designers will have to find a way to program their ideas.

It is very important for designers to get along well with each other, to share what each one knows, to give priority to whoever is best for the project, and not to insist on imposing that brilliant idea that you have come up with but that, I’m sorry, is not going to work. That’s why building a good converter development team is difficult, and not only because of the current shortage of professional profiles in each speciality of power electronics. It is a challenge for Human Resources, which has to deal with different types of people. They can’t all be the same, neither in skills nor in personality. That would be a disaster. It’s like recruiting for a pirate ship: some have to get by with the spade and others with the sails. And none of them should run off with the treasure.

Everyone is the customer

Hardware is only the beginning: power without control is worthless, said a tyre advertisement years ago. To achieve the required performance, the hardware is designed around what the control wants to achieve and the control around what the hardware can support. In addition, it is necessary to work with all the other development teams: Firmware, Mechanical Design, Validation… The project manager of a converter, who is often a former hardware designer, needs a lot of judgement to move the project forward, while at the same time it is necessary for each group of specialists to be able to polish their work.

You will often have to rely on them and allow them to spend scarce time clarifying anomalies or ruling out possible errors that they sense, but which sometimes turn out not to be real. For hardware people, as well as mechanical designers, helping industrialisation people is crucial. They are almost another speciality in their own right. And the factory has to be a second home for hardware after the lab, because a design that cannot be easily and efficiently mass produced and tested should not even exist.

He will also have to work hard with Procurement to get the costs right and avoid supply problems. For the hardware designer, everyone is a customer. That’s how he has to see it. But, above all he has to be clear about who the end customer is. It is not the project manager. It is not even the one who pays. The end customer is the field technician who is going to have to come in at 3 o’clock on a Saturday morning to see why your converter has stopped. The hardware engineer has to listen to the field technician more than anyone else in the world, because he is even more in touch with reality. He experience this first-hand.

It’s worth it

If you are in power electronics, you will be working with very enthusiastic people and just the right amount of ego. You will have the impression that in other areas of technology others are doing more sophisticated things than you, but you will be closer than they are to the best teacher – reality. And you will not stop learning. But I remind you again: if you’re hardware, it’s not as pretty as it sounds. You’ll never be an expert at anything, and you’ll know it. To make matters worse, the next day you’ll have forgotten what you learned the day before, because you’ll be on to something else. Don’t worry, tomorrow you will enjoy learning the same thing again. And then you will come up with a new idea. That will be the one that makes the difference.