General / first impressions:

Build quality is poor. My LCD had a few scratches in the reflector, and I read that someone else had the same problem. The keyboard connector's VCC pin wasn't soldered down, so it only made contact if you flexed the connector a certain way. The board is laid out nicely inside, but the soldering of the buzzer was messy and the diodes on the serial port were in places where they weren't originally intended to go. The keyboard buttons seemed to use tin contacts instead of the sturdier gold used on the non-button side of the board..

The software, although it looks good, is clumsy. Maybe it's that I'm used to the touch screen of my Palm, maybe it's poor design. The escape key is right next to the enter key, and accidentally pressing it clears your entry. The "del" key is where the backspace key usually is, but it only deletes characters to the right of the cursor; if you make a mistake, you have to hit the left arrow (above the "y" key) and then the "del" key. When you edit an entry, there is no 'insert' mode, meaning that to insert a character you have to retype the rest of the line.

Somewhat mysterious is the fact that some keyboard characters can be entered multiple ways. For example, "*" is 2nd-E or ctrl-7. A "-" is either a 2nd-A or a 2nd-E. Surprisingly, the enter key works as a clear in the calculator mode, not as an equals.

Also mysterious is the fact that the keyboard connects to the left of the Vista, but exits out the right side of the keyboard-- leading to a cable that is always in the way. This is probably a leftover from use of this keyboard with a daVinci organizer, which connected in the back of its cradle.

The installed batteries were almost dead, even though the pull-tab that was supposed to disconnect them seemed to be working properly. When I replaced them with the thoughtfully provided spare set, it worked better (the screen didn't dim as much), but the low-battery indicator was still on.

On the plus side, the processor is fast and the screen has a good resolution. This unit has a lot of promise, but is hindered by sloppy user interface design and lack of (software) expandability.
 

Keyboard connector:
 

Pin	Color	Signal
1	Red	VCC (4.5v when vista off, 4.1v when on - with my original set of low batteries)
2	White	Data
3	Green	Clock (direction unknown)
4	Black	GND

Note: pin 1 is indicated on the keyboard cable by the arrow on the connector. It is closest to the serial port connector.

The keyboard is very similar to the daVinci keyboard, but the keys may be relabeled and the connector is different.
 

Serial port connector:
 

1/8" jack Pin	Signal
tip	RX
middle	TX
shell	GND

I'm not sure of the levels; there is some sort of conversion in the DB-9 connector (and surprisingly, an LED that will never be seen).
 

Parts

IC1: chip-on-board, small. Looks like it handles reset and keyboard. May be the processor.

IC2: chip-on-board, smaller than IC1. Is probably a 256Kx8 ROM or FLASH, and may also be the processor. I'm guessing it's 256 KB because the Vista package says "768kb memory" and I can only account for 512KB in IC3. The exception would be if it uses some sort of compression... even 7 bit encoding or a fixed Huffman code would help, and the processor is certainly fast enough to do this.

IC3 is the flash memory, 512Kx8, FLASH, 90nS, 32 pin standard TSOP package, 5 volt-only, 256-byte pages, two 16K boot blocks, 1K cycle life. Under the chip is what looks like pads for either chip-on-board memory, or a smaller package surface-mount part. The cycle life seems surprisingly short -- hopefully all memory is used (to maximize life) and deletions aren't really erased (which would limit the number of erase cycles), but this would make the device insecure.
    W29C040T-90
    2019086Y2
    027AJRA
    Manufacturer is Winbond

IC4: chip-on-board, 101 pins, looks like an LCD driver.

IC5: chip-on-board, 101 pins, looks like an LCD driver (it connects to a lot of vias, and I didn't check to see if these went to the connector at the bottom of the LCD (not visible in the picture above)) or the processor (unknown type).

IC6: chip-on-board, 101 pins, looks like an LCD driver.
 
 

Self test mode

Press Reset while holding the on/off button
Press 1-8 for a screen test, "." for a flash memory test, or "right" to continue with normal reset.

The flash memory test consists of an Erase test and a Write test (which makes the LCD go dim). It won't run if there is data in memory (it'll display "Data exist"). When it's done, it'll display "Flash Test OK"
 

PC serial port
Seems to be 57600 baud, 8N1.
This protocol seems to be similar to one used in other Royal organizers.

A '@' is sent upon power-on.
pressing the link button seems to send a CR (maybe something else, too)
often the link will fail on the first character, but other times will time out (as if it checks the first byte for validity, and if valid expects a whole message)
sometimes it sends back an 'L'.

"A whole lot of good that does!" you say? Well, I haven't investigated it too much yet because of (1) the hope that someone else has already deciphered a similar protocol, (2) I don't want to install their PIM software on my PC if it is of the same caliber as the Vista software, and (3) I can't determine if the "no reverse engineering" clause for the PC software license is enforceable and/or applicable to the PIM/Vista link (to be sure, I'm not using their software, as per their agreement).
 

Memory Layout
The on-board memory contains the boot code and probably a lot of system calls to form a minimal operating system.

Without a FLASH chip, the unit will boot to a "Please Wait..." screen, where gets hung either trying to execute code out of FLASH, or read some data structure out of it. The screen test functions of the self test work, but the FLASH test doesn't (no error is displayed; it just hangs).
 

back to my homepage