Author Topic: Schematic review  (Read 820 times)

0 Members and 1 Guest are viewing this topic.

Offline sephtronTopic starter

  • Newbie
  • Posts: 5
  • Country: au
Schematic review
« on: January 04, 2021, 04:40:26 am »
Hey all, newbie here with my first schematic.

I'm building an OLED watch with an ATMega328p and PCF8563 RTC. It has three buttons to be connected to the same external interrupt. Also included an MCP73831 for the battery charging circuit.

I'd love some feedback as I'm sure there must be something I've missed or misunderstood. 

Thanks! :)

 

Offline Peabody

  • Super Contributor
  • ***
  • Posts: 2049
  • Country: us
Re: Schematic review
« Reply #1 on: January 04, 2021, 06:16:26 am »
I haven't studied the whole thing, but your P-channel mosfet after the battery is oriented backwards.  The drain has to be connected to the battery, and the source to Vcc.  This allows the mosfet's body diode to block USB voltage from getting back to the battery.  See Figure 5 of this app note:

http://ww1.microchip.com/downloads/en/AppNotes/01149c.pdf

When a button is pushed, how are you going to identify which one was pushed?
 
The following users thanked this post: sephtron

Offline sephtronTopic starter

  • Newbie
  • Posts: 5
  • Country: au
Re: Schematic review
« Reply #2 on: January 04, 2021, 07:30:52 am »
Oh, yeah thanks for pointing that out.

> When a button is pushed, how are you going to identify which one was pushed?

By code, got the idea here: https://www.hackster.io/Svizel_pritula/10-buttons-using-1-interrupt-2bd1f8

My prototype seems to work, so far.
 

Offline Peabody

  • Super Contributor
  • ***
  • Posts: 2049
  • Country: us
Re: Schematic review
« Reply #3 on: January 04, 2021, 03:20:40 pm »
Sound like you have everything covered.  One thing you might look at is the MCP73871 charger chip.  It looks like it has the load sharing function built in, so you wouldn't need the mosfet or the diode.  But I've never used one, so don't know how well it works.

Edit:  It turns out the MCP73871 only comes in a QFN package.  That would be a deal breaker for me.
« Last Edit: January 04, 2021, 03:26:52 pm by Peabody »
 

Offline sephtronTopic starter

  • Newbie
  • Posts: 5
  • Country: au
Re: Schematic review
« Reply #4 on: January 04, 2021, 04:44:04 pm »
Thanks, I've prototyped it with modules and an Arduino board so far. So here's hoping the PCB version works haha.

>  It looks like it has the load sharing function built in, so you wouldn't need the mosfet or the diode. 

Ohh, I'll give it another look. So with the charging circuit, I was referring to this existing project's design : https://blog.zakkemble.net/diy-digital-wristwatch/

He explains it more here: https://blog.zakkemble.net/a-lithium-battery-charger-with-load-sharing/

When i looked it up, the MCP73831T seems to come in a SOT-23-5 package. It can be found in JLCPCB's parts list:
https://jlcpcb.com/parts/componentSearch?searchTxt=C14879
https://datasheet.lcsc.com/szlcsc/Microchip-Tech-MCP73831T-2ATI-OT_C14879.pdf
« Last Edit: January 04, 2021, 04:45:51 pm by sephtron »
 

Offline pigrew

  • Frequent Contributor
  • **
  • Posts: 680
  • Country: us
Re: Schematic review
« Reply #5 on: January 04, 2021, 09:13:32 pm »
Except for the previously-noted backwards MOSFET, everything seems OK.

However, there are a few things I'd do in my design:
  • Add load capacitor (likely ~20 pF) on the RTC's OSCI pin.
  • Add decoupling caps near the MCU's VDD pins. (100 nF each according to Atmel).
  • Add resistor to short USB's shield to GND.
  • Swap the direction of the LED, so that it's connected to VDD, and the MCU would have an open-drain output.
  • Add pull-up on the display's nCS, so that it doesn't get confused during power-up.
  • If the XTAL pins of the MCU are unused, they should be grounded (perhaps through a 0-ohm resistor).
  • Verify battery has over-discharge protection.
  • I'm not familiar with AVR, but do you need a programming header? Are the button pins, and INTC pin, all interrupt capable (to wakeup the MCU)?
  • (maybe) add a 100k to 470k pull-up on the SPI_SI net because inputs are bad for power consumption.
« Last Edit: January 04, 2021, 10:09:27 pm by pigrew »
 
The following users thanked this post: sephtron


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf