Running some test i've let my circuit turned on, after some time i've noticed that the arduino was heating up... it was fried... i've mesured the tension out from the regulator and i rad 10V.
Keeping the output voltage from the regulator under control i've also noticed that the voltage slowly drop down.
The circuit is designed using a 4.7 uH inductor, VCC_EN is the input tension, i've added a shotky diode at the output facing arduino so when connected to pc for programming Vin cannot blow up the regulator, capacitors and resistor was selected to follow the typical application Vin=18V, Vout=5V/1.2A from datasheet.
The PCB was designet following the datasheets indication.
Designing a buck converter is a dark art. Component selection and layout is critical.
Using the wrong type of capacitor can lead to the problems you have.
Best to buy a ready made buck converter. Pololu has a good selection.
Why did you actually use a buck converter.
The Every already has one built-in. The MPM3610.
Leo..
I want to keep arduino only for actuation and sensor reading and be sure to be able to provide sufficient power to users.
I'm also thinking of changing the Arduino to a ESP32c3 super mini, so I would still need the 5V power supply to power what I need.
I'm using the same capacitors as the datasheet indication, same value and same working voltage value.
I'm thinking to put capacitors whit higher voltage range to ensecure the minimum operating capacitance needed, but i don't know if it's actually the solution.
The first time I built it, I mistakenly assumed from the datasheet that it only needed a resistor for the enable pin (R5). Rightfully so, the track caught fire when I first started it up, so I built a voltage divider that provides 2V to the enable pin, allowing the regulator to start up without any problems.
As you can see, the design is exactly as per the datasheet. As I said before, it initially worked fine, but I don't know what happened.
Are you? There is more to a capacitor for this type of circuit than that. Most important is the ESR (effective serial resistance).
This is very true. I have managed specialist buck converter engineers, and getting all the correctly specified components, and also having heaps of equipment you would not have access to, they still take a minimum of three attempts before they get it right.
The best piece of advice on this thread by a far way.
Thanks for all responses,
Ok i've underestimated the trace width, during the design process i've put some extra trace on that but is clearly not enaught.
I started a new design and this is what i'm doing right now:
It is mirrored to make a simpler comparison with the datasheet since it is located in the bottom layer of the project. I keep the shotky diode at the output to prevent inverse tension during while programming the board.
The inductor type is very critical. It is important that it has the right quality of ferrite for the purpose.
As I said in an earlier post, even if you exactly replicate the recommended layout from the data sheet, that is only a beginning. I think you are way over your head with this project.
I don't think I'm that far off the mark; I calculated the inductor based on the datasheet, and the component type and characteristics also follow the exact model in the datasheet.
I've certainly made several mistakes, but you learn from your mistakes; you can't always buy prefabricated modules like the Pololu one.
I'm here to improve my knowledge and correct my mistakes.
Thank you all again for your replies, support, and patience!
You were not mistaken. You do only need one resistor connected between EN and Vin. It should not have caught fire. So something else was wrong with the design.
I would first determine why that happened before going any further
Since the whole project had changed a bit, I finally decided to set the regulator to 3.3V.
I took the opportunity to redesign everything step by step and obtained:
Initially, my rush to design the regulator led me to make a poor choice of capacitors. A 10uF 6.3V output capacitor wasn't nearly enough, as it was regulated to 5V, resulting in a drastically reduced effective capacitance.
Now, the capacitor is 22uF 6.3V to ensure the minimum working capacitance at 3.3V.
Regarding ESR, ceramic capacitors with X5R or X7R dielectrics will be used to keep the ESR around 4mOhm.
Obviously, the inductor will also be chosen from those recommended by the datasheet.
I also double-checked the resistors for the enable pin, which can indeed be driven with a single resistor, but apparently it's preferable to use it as a UVLO input, so a dedicated voltage divider is required.
I therefore redesigned the PCB, where, in addition to respecting the datasheet layout, I also ensured the correct routing, obtain:
(without, with ground plane and a render) where VCC_EN is the input voltage.
Of course, if you have anything else to add or suggest, I'd be grateful!
I wanted to share everything so that this topic might also be useful to anyone else who wants to improve their design of even complex components like bulk converters.
