The power issue is pretty tough I think. Do you want to allow charging from both sides (12v dc jack and eoma-5v0) or does the power from the card only act as additional power for the system (without charging and just taking the load off the battery)? How do you want to control the power switch (SY6280) and detect whether the card is passive or active (taking power/providing power)?

These are just a few thoughts I wanted to share …

Julius


On 21.02.2017 03:21, Luke Kenneth Casson Leighton wrote:


---
crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68

On Mon, Feb 20, 2017 at 1:07 PM, Internet <internet@devpi.de> wrote:
I took another look and i am pretty satisfied with the results ;).

yeah, thank you for encouraging me to do it.  
 
However, a few things still need attention (i wrote it down on rhombus-tech.net)


ok let's go through them:

PCB3 - Power Board

  • connect thermistor for BQ24193 USB OTG Charger between BAT_NTC and GND (not TS which is wired to the fuel guage)
yep, cut/paste error there

  • change DVCC to SYS-V (otherwise LEDs won't work)
yep done
  • pin 2 of connector J2 (to PCB1) should go to VCC-5V0 (we need power for the EOMA68 card)
ah no.   rrright, the way it works is:

* EOMA68 Card Power is actually "OTG-like".  as in, you should expect it to be *reversible* on demand, because there could actually be an OTG charger plugged into the card, providing power *in* to the Housing.  this is why the SY6280 is required.

* the laptop's main components need a *separate* 5.0V power rail, for the USB Hub, Audio, LCD, USB2 connectors etc. etc.

* power could actually come *in* through the EOMA68-5V0 line and has to go *in* to VBUS... which... argh, is missing so that needs sorting.

* power can also go *out* through the EOMA68-5V0 line (from VBUS) when the bq24193 is put in "boost" mode (this is for OTG power compliance) ... which is missing and entirely defeats the purpose of *using* the bq24193 in the first place :)

*sigh* so still have to sort that out, creating a two-way power path between VBUS and EOMA68-5V0 at *specifically* 5.0v, coping with 1.5A sustained current, and under user-control... preferably without using too many components.  SY6280's are my favourite "power cutting IC".  there's one spare pin on PCB3 J4 which could be used to control its enabling.

my first guess at a circuit would be a schottky diode pointing from EOMA68-5V0 to VBUS to act as "power in to VBUS" and the SY6280 to act as "power out".  however, a *five amp* power regulator is going to be needed on VIN to ensure that the input voltage never exceeds 5V (which it could at the moment).

thoughts appreciated.

.






_______________________________________________
arm-netbook mailing list arm-netbook@lists.phcomp.co.uk
http://lists.phcomp.co.uk/mailman/listinfo/arm-netbook
Send large attachments to arm-netbook@files.phcomp.co.uk