Hello, Sotiris here! š
š Today I want to discuss about some ways to reduce the timing of Power Electronic design process.
āļøā Letās first have a look at the different steps, to see how reducing time is possible!
The Power Stage R&D process:
āš¤ Which process does a hardware design engineer follow, to get a working PCB prototype?
- Specs for a prototype
- Ideas about implementation, topologies, attacking certain problems etc.
- Reference designs. You donāt have to invent the wheelā¦
- Custom circuits CAD software, based on experience, reference designs, previous iterations
- Layout, EMI considerations, thermal management, signal integrity and another 100 problems
- Ordering parts, finding replacements and waiting for the parts to arrive. ā°
- Assembly of the prototype PCB
- Testing, debugging, corrections and a roller-coaster of emotions š
You know what? Over the years I got bored of the process above. Iāve repeated the steps so many times, both for small and big projects, from the concept idea up to the working prototype, that Iāve developed a good sense of the time required for each design stage.
What is this, you might ask, a self-affirmation post of an engineer? š Iāll pass!
WAIT, itās not what you think ā
Jokes aside, many times Iāve found myself in weird places thinking about how to speed up the R&D process.
š„ āImagine having an idea on Monday and a PCB ready to be tested on Wednesday!
I believe that apart from the speedup, there is a real efficiency benefit in this scenario: on Monday you have the problems, questions, the hunger and the excitement to find out if your design works. But then, all these emotions that feed creativity, get you focused and at the end of the day get you the golden idea, fade away as time passes, and we go from that Wednesday to 3-4 weeks aheadā¦
š āYeah, nice, but how can we get closer to this ideal scenario? I donāt live in Houston, Texas next to Mouser warehouse, nor do I live next to a big factory, like JLC, to get my PCBs a couple of hours after they are manufacturedā. How can this be solved?
š The answer for me is building a stock of hard-to-get components in your lab, taking charge of the ordering process, knowing exactly how many days it takes for things to get to you, and making arrangements to be sure you donāt have to stop the process because of 1-2 parts. You can really minimize the waiting time once you get hold of the logistics! Questions like āAre customs involved in the shippingā or āWhat are the incoterms of shippingā might make or break you.
The magnetics R&D process:
š Magnetics designs share a lot of common problems with PCB design. Itās custom work that entails a design software, an engineer to design a component, the BOM to create a magnetic component, a fab house to wind your part and finally shipping.
š Manufacturing and logistics are taken care of from Frenetic online in this case though! Our services integrate design with actual assembly and shipping, providing you with a sample at your facilities usually within 2 weeks.
All the engineer has to do is design the component and ask for a sample. Thatās itā¦
āHmmmā¦ Good... but 2 weeks feels like a long timeā¦š
ā
āI want that transformer/inductor within a week, to minimize my R&D timeā you might complain.
š After all, I did preach previously about making sure that 1or 2 parts donāt hold back your project.
How can Frenetic help you solving this problem?
ā
We have created a library of built and tested samples for you to choose and order right away! This will give you all the flexibility you need to get things going as fast as possible.
Letās figure out a PV application together.
Say we are interested in PV solar applications. A typical solar panel 1.6m length, 1m width and about 35mm thickness will output 250-300W if the sunās irradiance is high enough, like 1000W/m2, at 25Ā°C cell temperature. Of course, these are optimal conditions and truth is that panels wonāt āexperience that much sunā, plus the temperature of their cells wonāt be a fixed 25Ā°C either. As the maximum power drops during the day the voltage at which that power can be extracted drops linearly.
Without getting into details, a 18V-35V voltage range is about right for most solar panels. Looking for an application note that describes a 250W DC/DC converter to extract power from a single solar panel I came across Infineonās app note [1].
