New: XDA launches forum for app developers. Discuss coding, tools, marketing, and more.
XDA Developers Android and Mobile Development Forum
Forgot your password?
THREAD CLOSED
Tip us?
 
AdamOutler
Old
#411  
AdamOutler's Avatar
Elite Recognized Developer
Thanks Meter 8377
Posts: 4,838
Join Date: Feb 2011
Location: Louisiana

 
DONATE TO ME
Im going to be looking at tapping the lines from sensors. Camera, tilt, light, compass...
Quote:
Originally Posted by jrloper View Post
Windows is a finicky, needy, hormonal pain in the .... You get the point.
Check out my developer pages. Add me to your circles on Google Plus.
 
AdamOutler
Old
(Last edited by AdamOutler; 23rd May 2011 at 03:34 PM.)
#412  
AdamOutler's Avatar
Elite Recognized Developer
Thanks Meter 8377
Posts: 4,838
Join Date: Feb 2011
Location: Louisiana

 
DONATE TO ME
gosh, I wish we could unlock all the secrets of the FSA-9480 chip.
Quote:
Originally Posted by jrloper View Post
Windows is a finicky, needy, hormonal pain in the .... You get the point.
Check out my developer pages. Add me to your circles on Google Plus.
 
TheBeano
Old
(Last edited by TheBeano; 23rd May 2011 at 11:19 PM.)
#413  
Senior Member
Thanks Meter 134
Posts: 348
Join Date: Jul 2010
So it seems quite likely that:
  1. BOOT_ON in the resistor constant name does indeed mean the BOOT pin is on, so the BOOT signal from the FSA9480 will be high for 619K and 301K resistors on the ID pin, no other external inputs required
  2. The MANUALSW2 I2C register that we were playing with earlier does indeed set the BOOT pin high and low.
So, the BOOT pin from the FSA9480 is not directly connected to the OM[5] pin. If it does affect it at all, as was hinted at by the BOOT_MODE signal on the block diagram, then there is some additional logic involved somewhere.

The phone boots up automatically with the 523k (BOOT_OFF) and 619k (BOOT_ON) resistors. So the BOOT signal, if it is used at all, doesn't tell the application processor to boot up.
 
AdamOutler
Old
#414  
AdamOutler's Avatar
Elite Recognized Developer
Thanks Meter 8377
Posts: 4,838
Join Date: Feb 2011
Location: Louisiana

 
DONATE TO ME
Quote:
Originally Posted by TheBeano View Post
So it seems quite likely that:
  1. BOOT_ON in the resistor constant name does indeed mean the BOOT pin is on, so the BOOT signal from the FSA9480 will be high for 619K and 301K resistors on the ID pin, no other external inputs required
  2. The MANUALSW2 I2C register that we were playing with earlier does indeed set the BOOT pin high and low.
So, the BOOT pin from the FSA9480 is not directly connected to the OM[5] pin. If it does affect it at all, as was hinted at by the BOOT_MODE signal on the block diagram, then there is some additional logic involved somewhere.

The phone boots up automatically with the 523k (BOOT_OFF) and 619k (BOOT_ON) resistors. So the BOOT signal, if it is used at all, doesn't tell the application processor to boot up.
True. I think the boot signal is a signal for the kernel to start battery charging display, or actually boot Android.
Quote:
Originally Posted by jrloper View Post
Windows is a finicky, needy, hormonal pain in the .... You get the point.
Check out my developer pages. Add me to your circles on Google Plus.
 
TheBeano
Old
#415  
Senior Member
Thanks Meter 134
Posts: 348
Join Date: Jul 2010
It loads the battery charging kernel if the JIG signal is off and the power switch is off, and it has power on the USB socket (VBUS_IN). It's the JIG signal that switches on the phone automatically in the factory modes, we know this from the circuit diagrams in the service manual. So the BOOT signal goes somewhere else.
 
AdamOutler
Old
#416  
AdamOutler's Avatar
Elite Recognized Developer
Thanks Meter 8377
Posts: 4,838
Join Date: Feb 2011
Location: Louisiana

 
DONATE TO ME
Quote:
Originally Posted by TheBeano View Post
It loads the battery charging kernel if the JIG signal is off and the power switch is off, and it has power on the USB socket (VBUS_IN). It's the JIG signal that switches on the phone automatically in the factory modes, we know this from the circuit diagrams in the service manual. So the BOOT signal goes somewhere else.
Then why do the "boot off" resistors only activate charging mode and not boot?

The kernel loads up whenever power is applied. The boot on resistors actually turn the phone on.
Quote:
Originally Posted by jrloper View Post
Windows is a finicky, needy, hormonal pain in the .... You get the point.
Check out my developer pages. Add me to your circles on Google Plus.
 
TheBeano
Old
#417  
Senior Member
Thanks Meter 134
Posts: 348
Join Date: Jul 2010
Quote:
Originally Posted by AdamOutler View Post
Then why do the "boot off" resistors only activate charging mode and not boot?

The kernel loads up whenever power is applied. The boot on resistors actually turn the phone on.
Yes, I should say the JIG signal switches it on when there's no USB power. If there's no USB power then the BOOT_OFF resistors still switch the phone on. But after the SBL is loaded all the decisions are made by reading the resistor value through I2C, so there's no need for the BOOT signal at that point. So it must be doing something else.
 
AdamOutler
Old
#418  
AdamOutler's Avatar
Elite Recognized Developer
Thanks Meter 8377
Posts: 4,838
Join Date: Feb 2011
Location: Louisiana

 
DONATE TO ME
Do you agree that i2c over USB is not looking possible?
Quote:
Originally Posted by jrloper View Post
Windows is a finicky, needy, hormonal pain in the .... You get the point.
Check out my developer pages. Add me to your circles on Google Plus.
 
TheBeano
Old
#419  
Senior Member
Thanks Meter 134
Posts: 348
Join Date: Jul 2010
Quote:
Originally Posted by AdamOutler View Post
Do you agree that i2c over USB is not looking possible?
Yes it seems very unlikely that there is anything like that supported by the FSA9480.
 
AdamOutler
Old
(Last edited by AdamOutler; 24th May 2011 at 02:30 AM.)
#420  
AdamOutler's Avatar
Elite Recognized Developer
Thanks Meter 8377
Posts: 4,838
Join Date: Feb 2011
Location: Louisiana

 
DONATE TO ME
I'm still exploring for i2c, it's cheaper then JTAG, and, according to Google IO, its supported by the Android Open Accessory Standard ( which is an Arduino mega + usb shield + extra crap.. the extra crap brings the price to $400). So, it still makes sense to explore that.

Since the chip has a maximum i2c speed of 400khz, and the processor has a maximum i2c speed of 3mhz, 400khz would be the logical i2c bus speed on the platform. The following modification is required for Arduino to communicate at this speed:

under linux:
1. edit the twi.h in the wire library
Code:
sudo gedit /usr/share/arduino/libraries/Wire/utility/twi.h
2. change line 32 from
Code:
  #define TWI_FREQ 100000L
to
Code:
  #define TWI_FREQ 400000L
Speed is set to 400khz for any arduino sketch compiled there after.
Quote:
Originally Posted by jrloper View Post
Windows is a finicky, needy, hormonal pain in the .... You get the point.
Check out my developer pages. Add me to your circles on Google Plus.

THREAD CLOSED
Subscribe