Project Logs

Collapse

• Charger issues

  Chris Hamilton • 08/13/2015 at 11:13 • 0 comments

  So far I haven't had any replies back from TI. I have power supplied from VBUS entering the BQ24296, but I believe the fuel gauge logic is preventing charge mode. The design is just too complicated compared to using a TI fuel gauge, so that will be our next step.

• Still working on new design

  Chris Hamilton • 07/19/2015 at 09:41 • 0 comments

  I have been testing this design and there are some weird situations. I had thought to sell this version, but I don't think it necessarily works. I'll be working on some changes to the logic driving it asap. Possibly, I will end up just setting the Qi charger priorities to a fixed mode and improving the fuel gauge layout.
  

• ST NRND'ed the STC3115AIJT?

  Chris Hamilton • 04/11/2015 at 08:12 • 0 comments

  OK, so it looks like ST is pulling out of fuel gauges. Great... Well, honestly even though I loved the STC3115AIJT, I had done a bunch of work trying to fully integrate the chip in. Now I am looking at the BQ27531-G1 which works with a similar series of chargers and actually 'runs' the charger and filters uc requests. So maybe it will be the awesome default discrete does most things, i2c enables all features USB/Qi Charger I meant to make.

• Testing

  Chris Hamilton • 04/04/2015 at 18:11 • 0 comments

  Yield is very low and required re-working. The spacing per component is cramped, but it is just too much solder paste for the amount of area. It takes approximately 30-40 minutes to complete one which effected the slump and workability with my new 4300LF paste wasn't holding up well either. I think I'll stick to SMD291SNL for more complicated boards (and in general). I am going to be testing the board, but it will be expensive to sell at this point. I will try to minimize the Qi wireless capacitor count as much as possible.

• Design Complete

  Chris Hamilton • 02/10/2015 at 18:42 • 0 comments

  Initial design is off to OSHPark. This design is 4 layers, but kinda breaks the rules in having what is essentially 3 signal layers. The ground plane is solid, but there are many high current traces, so they are actually placed on the bottom layer instead of a split internal VCC layer. This allows the design to be functional and fit within the 4 layer stack for OSHPark. The analog ground is somewhat isolated, though there may be issues with some near traces. The inductor is well isolated from traces.

View all 5 project logs