This page summarizes the revisions of the GTA02 hardware, both pre-release and planned future revisions.
The internal codename of the Neo FreeRunner was GTA02. Compared to the previous GTA01, it has the same physical dimensions, screen and RAM, but more flash storage, a faster processor, plus wifi and accelerometers. The next major revision is expected to be GTA03.
All Neo FreeRunners that have been released to the public at this time are revision 6. There are a few persons with r5, but they were manually adapted to the state of r6.
Estimated time line
|TODO: It should be nice if there are some dates next to the milestones (See: To-Do List)|
- GTA02v3 design finalized - DONE
- GTA02v3 power supply flaw found, GTA02v4 needed - DONE
- GTA02v4 design finalized - DONE
- GTA02v4 small number produced for Openmoko internal developer evaluation - DONE
- GTA02v4 thorough evaluation by Openmoko internal developers - In progress
- GTA02v4 flaw found, GTA02v5 needed - On hold while further development on drivers is done to ensure no further hardware bugs. - In progress
- GTA02v5 design finalized - DONE
- GTA02v5 small number produced for Openmoko internal developer evaluation - Done (around 20)
- GTA02v5 thorough evaluation by Openmoko internal developers - Done
- GTA02v5 produced in moderate volume - Done
- GTA02v5 goes on sale to community via web store - Done
First generation of prototypes that was given to internal Openmoko software developers. Total 30 pcs fabricated.
- It is working just fine, but still based on 2440, with external NAND/SDRAM and no NOR flash
- Using the PCF50633 05 N3 due to 04 N3 not available, re-work power for basic schematics verification
- Using GTA01 SIM socket
- Add external debug port
- Still using Global locate A-GPS
- ATAG_REVISION: 0310
Second generation of prototypes, Total 50 pcs run at Taipei SMT factory MOUNT
- Ideal is have 256 MB NAND on Samsung package, Due to chip availability Start using S3C2442 B43
- Using correct PMU PCF50633 04 N3
- Change new SIM socket
- Change to u-blox A-GPS
- Change LCM power from 3.3v to 1.8v
- USB power switch layout/pin assignment mistake, could not verify USB host supply 5v function
- GPS function verified ok with good sensitivity
- ATAG_REVISION: 0320
Production verification version, 2007/10/11 28 pcs fabricate at FIC SuZhou
- Still using S3C2442 B43 for hardware verification
- Using control pilot run to verify S3C2442 B54 chips
- ATAG_REVISION: 0330
Mass production release candidate version 1
2 weeks after v3 gerber out, release the v4 gerber, and 2007/10/20 20 pcs fabricate at FIC SuZhou
- Change LCM power from 1.8v to 3.3v for display stability issue
- fabricate another 200 pcs for yield rate/production verification
- fabricate 50 pcs with S3C2442 B43 (128 MB NAND) for quality comparsion
- USB host power chip have some output voltage stability issues with Vb/Vcc comes from different power source, need layout change to fix the issue
- Battery Coulomb design not working on A4
- ATAG_REVISION: 0340
Mass production candidate version 2/Mass production version
- First batch fabricate 2008/1/14 at FIC SuZhou
- Mass production A5 trial run start from 2008 March, including some resistor/capacitor change compare with inital 100 pcs prototypes A5, and prototypes for GTA02 developers was tracked in the Prototypes Page
- Coulomb counter issue fixed
- USB host power switch fixed
- Need add capacitor for PMU Vbat input for stability issue, this could be done by direct SMT or hand rework
- Need rework (still using SMT in production) add capacitor for PMU Vbat input for PMU stability issue.
- Need manual rework GSM IR UART path a 100k pull down for better GSM deep sleep
- ATAG_REVISION: 0350
GTA02 mass Production version change list
- PMU's LED power error: PMU potential damage issue
- NOR FLASH enable WP: User can write data into NOR FLASH.
- CE CS/RS fine tune: Audio's background noise too high
- I2C pull high resistor: The resistor is too high and signal is distorted
- GSM leakage current: TX_MODEM has a pull high resistor on IO_3V3
- Power consumption: Disable keep active function
- SDIO clock and esd protect resistor
- Refer to Datasheet: R1526 to 33K
- GSM modem on pin: The R1018 is too small and has some leakage current
- LED driving transistor: When GPIO is on, the transistor will be draw more current on LED. This is component change fix, do not need change PCB or re-work.
Mass production candidate version 3/Mass production version
A6 will be fine tune version of A5, only minor schematic change for better product quality and version control. Capacitor and resistor change A6 also on mass production A5
- First 100 pcs start from 2008 mid April, and factory make component placement mistake on GSM, second 100 pcs PCB arrive time TBD.
- Add capacitor space for Vbat, reduce the SMT effort
- Add GSM IR resistor for better GSM deep sleep
- Reserve 3 GPIO for hardware version control
- Fixed LEDs power usage (from about 150mW of v5 to about 25mW)
- ATAG_REVISION: 0360
GTA02 A5 to A6 changes
- Power Glitch on VB_SYS: Add capacitor on layout, Mass production A5 also apply this change.
- G-sensor separate these interrupt pins: At A5, each accelerometer INT1/INT2 connected to same line, at A6 only INT1 was connected.
- GSM_modem power source Reduce power's ripple when the phone is talking
- Keep active Disable keep active function, just fine tune
- GPIO for version control
- GSM RX_IR has some noise Add resistor and reduce GSM RX_IR noise and gsm can't enter suspend mode easily, apply on mass production A5.
- LED driving transistor apply on mass production A5.
- LCM's VDDIO We can totally power off LCM's power, save about extra 1mA.