PCB Design for Charging, Protecting and Boosting a battery

Hello,

this is the first time I am trying to make my own PCB and before I order it, I would be really happy if someone could look over it.

I am trying to design it with the following functions:

  1. Charges LiFePo4 Battery (not usual Li-Po) to 3.6V
    IC I thought about using: CN3058E (Datasheet: CN3058E datasheet)
  2. Protects the Battery from overcharging and overdischarging
    Using LiFePo4 protection IC HY2112. (Datasheet: https://caxapa.ru/thumbs/480772/DS-HY2112_EN.pdf)
    And Mosfet FS8205 (Datasheet: https://datasheet.lcsc.com/lcsc/1810110910_Fortune-Semicon-FS8205_C32254.pdf)
  3. Boost the voltage to 5V to use with Arudino
    Using a MT3608 (Datasheet https://www.olimex.com/Products/Breadboarding/BB-PWR-3608/resources/MT3608.pdf)
  4. Has a switch to use the board for the projects even without a battery connected
  5. Has mounting holes to screw it in the case of my projects.

This is the design I came up with so far:

It is normal to make a prototype from strip board before you lay out your PCB. That way you know the circuit will work.

Otherwise it is likely you will have to make a prototype PCB first. Then get it working by cutting tracks and adding links until it works, and then have a second PCB made that incorporates the changes you have to make. This is the way professional engineers do things.

1 Like

That way it would be very expencive and time consuming, wouldnt it?
Of course in a professional and more complex cirquit, but as a hobby?

I would have to buy all components in the regular form and in SMD
Then I would order the PCB and wait for weeks until it arrives, just to test it and order it again.

My hope was that someone already worked with the components which are very common charge, boost and protect ICs and maybe take a look if I connected them correctly before running into errors I will not know how to solve or even analyze why it is not working

Yes, eh, for the smd parts that you will need you can also get SMD to THT adapters.

I tend to first get some of the parts and get other parts from what i already have.
I have THT leds, i have the resistors to trim the MT3608, i will need the inductor (although by now i have some)
I got some SS34 diodes, and just soldered some wire onto the end to make them THT. btw that part of the schematic looks ok, although i would let the direct power and the BAT + come in thru a diode each, before the boost stage. That way there is no voltage drop at all after the boost, and if direct power is available, battery power stops automatically. The other active components i do not have any experience with, except the 8205 dual mosfet. But the package i have those in is 6 pin and tiny, i think sot-23-6.
Still i prefer to have all parts before i start top actually design the PCB from the schematic, so i can verify that the footprints will actually match the packages that i have.
I do the whole thing on a 1/10th inch experimentation board before starting the design and i actually start out on a breadboard. Once the prototype works as i want, i will copy the schematic and start to design the PCB, starting with a small amount, and usually the wait for that is a couple of weeks since i don't want to spend loads of money on DHL. Many times i have a few designs that i order to cut the shipping costs down, and i always want the sender to make sure taxes and duties are paid, to prevent extra costs.
I will the populate the PCB myself and test it to confirm it's functionality. Even then, still things may show up.

Skipping these steps will make the process to unreliable, and fault finding near impossible

Yes, but it always is. Electronics is not a cheap hobby and everything takes time. If you order a manufactured PCB with the parts, it will be expensive and there will be a very big chance it will not work the way you want it to.

Well that is the breadboard stage, start there and if you run into something, Me, Mike or someone else can have a look and see if we can find what you did wrong. Without the actual parts you can not test anything. You can probably also buy modules based 6on the parts of your choice that do more or less what you want and combine them. (not my choice, though for some one of projects i have done... again for a prototype it is an option)

There may always be parts that i have decided i can not solder, like a w5500 chip, so i bought some modules that i can attach to a designed PCB, until i know that it works the way i want, and only then will i incorporate it into the next batch of PCB's and get the w5500 soldered onto it.

Creating a manufactured PCB has all sorts of extra complications. The parts you buy from a PCB manufacturer, and how they are placed and rotated, all those things can go wrong, and you may have to deal with the PCB makes 'engineers' , which are not engineers but people that put rolls of parts on machines and rarely speak english very will, i mean it's a trip.

These days, you can get a PCB made in very small quantities and have them back in days rather than weeks.

The problem with the boost circuit, that is not just the circuit that you need to consider. The layout is critical. I have managed specialist power supply engineers and have not known any one of them get a layout correct to their satisfaction in less than three goes at the layout. The circuit itself doesn't change but the layout does. And that is with them knowing exactly sort of inductor they need.

okay, I will order the parts today and test the cirquit as soon as they arrive

Its what I would do, and its usually not difficult.

Once a PCB is designed and built it can be very very difficult to identify the problem.

Oh good. Do you have experience with SMT soldering, because i never found dip versions for quite few of the parts you need,

Not really, but I will try

A small blower really helps, and a syringe of solder paste as well. And i admit i use a magnifying glass.

This topic was automatically closed 180 days after the last reply. New replies are no longer allowed.