Handhelds.org - Open source for handheld devices

UserPreferences

DellAximX5Hardware

Dell Axim X5 hardware information collection

[WWW]http://handhelds.org/moin/moin.cgi/DellAximX5

Authors:

Disclaimer

This document has NOTHING to do with Dell Computers. The information collected here is the result of the efforts of a few people involved in the "Linux on Dell Axim X5" project. It has been collected using all available documentation from network, as well as miscellaneous tools written specifically for this (such as tester and HaRET).

DMA channel state

MECR=0x2
DINT=0x0

1st column is Channel control. The rest could be found in documentation for PXA250 processors 

DMA #0=0x8  DCMD=0x0 DSADR=0x0  DTADR=0x0
DMA #1=0x80000000 DCMD=0x9023ca60 DSADR=0xa1fbd560 DTADR=0x40500040
DMA #2=0x8  DCMD=0x0 DSADR=0x0  DTADR=0x0
DMA #3=0x8  DCMD=0x0 DSADR=0x0  DTADR=0x0
DMA #4=0x8  DCMD=0x0 DSADR=0x0  DTADR=0x0
DMA #5=0x8  DCMD=0x0 DSADR=0x0  DTADR=0x0
DMA #6=0x8  DCMD=0x0 DSADR=0x0  DTADR=0x0
DMA #7=0x8  DCMD=0x0 DSADR=0x0  DTADR=0x0
DMA #8=0xa0000000 DCMD=0x6022c030 DSADR=0x40500140 DTADR=0xa1fcf000
DMA #9=0x8  DCMD=0x0 DSADR=0x0  DTADR=0x0
DMA #10=0x8  DCMD=0x0 DSADR=0x0  DTADR=0x0
DMA #11=0x8  DCMD=0x0 DSADR=0x0  DTADR=0x0
DMA #12=0x8  DCMD=0x0 DSADR=0x0  DTADR=0x0
DMA #13=0x8  DCMD=0x0 DSADR=0x0  DTADR=0x0
DMA #14=0x8  DCMD=0x0 DSADR=0x0  DTADR=0x0
DMA #15=0x8  DCMD=0x0 DSADR=0x0  DTADR=0x0

DMA: Pocket PC uses at least:

GPIO configuration (HaRET output)

GPIO# D S A INTER | GPIO# D S A INTER | GPIO# D S A INTER | GPIO# D S A INTER
------------------+-------------------+-------------------+------------------
  0   I 1 0    FE |  21   I 1 0       |  42   I 1 1       |  63   O 0 0
  1   I 1 1       |  22   O 0 0       |  43   O 1 2       |  64   I 1 0
  2   I 1 0    FE |  23   O 1 0       |  44   I 1 1       |  65   I 1 0
  3   I 1 0    FE |  24   O 0 0       |  45   O 1 2       |  66   I 1 0
  4   I 0 0 RE    |  25   I 1 0    FE |  46   I 1 2       |  67   I 1 0
  5   I 1 0       |  26   O 0 0       |  47   O 0 1       |  68   I 0 0 RE FE
  6   I 0 0 RE FE |  27   I 1 0 RE FE |  48   O 1 2       |  69   I 0 0 RE FE
  7   I 0 0 RE    |  28   I 0 1       |  49   O 1 2       |  70   I 0 0 RE FE
  8   I 1 0    FE |  29   I 0 1       |  50   O 1 2       |  71   O 0 0
  9   I 1 0    FE |  30   O 0 2       |  51   O 1 2       |  72   O 1 0
 10   I 0 0 RE FE |  31   O 0 2       |  52   O 1 2       |  73   O 1 0
 11   I 0 0 RE FE |  32   I 0 0       |  53   O 1 2       |  74   O 1 0
 12   I 0 0 RE FE |  33   O 1 2       |  54   O 1 2       |  75   I 1 0
 13   I 0 0 RE FE |  34   I 1 1       |  55   O 1 2       |  76   I 1 0
 14   I 0 0 RE FE |  35   I 1 1       |  56   I 1 1       |  77   I 0 0
 15   O 1 2       |  36   I 1 1       |  57   I 0 1       |  78   O 1 2
 16   O 1 2       |  37   I 1 1       |  58   O 1 0       |  79   O 1 2
 17   I 0 0 RE    |  38   I 1 1       |  59   I 1 0       |  80   O 1 0
 18   I 1 1       |  39   O 1 2       |  60   O 0 0       |  81   I 0 0
 19   O 0 0       |  40   O 1 2       |  61   I 1 0    FE |  82   I 0 0
 20   O 0 0       |  41   O 1 2       |  62   O 0 0       |  83   I 0 0

Legend:

This table shows all GPIO's that have been found so far:

MediaQ 1132

LCD

Axim X5 uses the MediaQ 1132 chip to control LCD. A fully functional driver to control backlight, brightness and contrast has been written specifically for Dell Axim X5.

FRAME BUFFER

Frame buffer is statically mapped to physical address 0x14000000, it uses 16 bit format and its resolution is 240x320 pixels. PocketPC sets framebuffer start address + 0x2000 (e.g. 0x14002000 or 0x14042000 depending whether you want synced or unsynced access - see mq1132 docs).

BACKLIGHT

Controlled by Pulse Width Modulator #0 on PXA. The brightness (range 0..1) == PWM_PWDUTY0/PWM_PERVAL0. PWM_PERVAL0 is 254 in Windows, so you have 254 gradations of brightness. In Linux you have 1023.

UART

Normal UART available is fully documented in 278522-001.pdf. FIR has not been tested yet, but should work with the pxa_ir driver. The IR tranceiver is enabled via GPIO63 on MediaQ 1132 chip.

USB

There is a USB Device Controller built-in into the PXA but it is not used. Instead, Axim uses the UDC that is built into the MediaQ chip. This means we'll have to write an UDC driver for MediaQ (doh). MediaQ 1132 contains a USB host (OHCI compatible) and it is rumoured that the outputs are available on the bottom connector, thus it is possible to modify the OHCI driver so that it works with MQ1132. This would mean that you will be able to connect any USB device (say, a 20GB portable HDD or a scanner) to your Dell (which PocketPC isn't able to do).
USB cradle insertion process:
USB cradle removal process:

Update: A driver has been written to drive the UDC. At least the g_ether gadget driver works fine with it (1.5 to 2 times faster than in WinCE! 500k/s PC->PDA and 375k/s PDA->PC).

PCMCIA/CF

All stuff that can be found in 278522-001.pdf (PXA250) and/or 27869301.pdf (PXA255) is valid also for Dell Axim.

PCMCIA/CF insertion process:

PCMCIA/CF removal process:

Interrup requests from CF cards are routed through MediaQ chip GPIO2. This is a level-triggered interrupt, thus after you process one interrupt you must disable and then re-enable that specific interrupt in order to get next one.

SD/MMC

It looks like these cards are handled in SPI mode (a four-wire interface) via the MediaQ chip (hell knows why, PXA2xx has a built-in MMC controller). When a SD card is inserted, MediaQ register SP00R bits 0-1 are set to <code>01</code> (this enables GPIO 50-53 to act in SPI mode). Also the pull-up registers for GPIO50-53 are activated.

Touchscreen

Touchscreen is controlled by wm9705 AC97 codec. For some reason the Y channel readings are very noisy.

Battery

The main battery voltage is measured via the BMON input of the wm9705 codec. The battery capacity is either 1400mA<em>h and 3400mA</em>h. Battery capacity is detected via MediaQ GPIO66 pin: if it's 1, it s a 1400mA*h battery; if it is 0, its the <code>extended</code> battery.

Sound

Dell Axim uses the wm9705 audio codec. Amplifier power is controlled by GPIO23 & GPIO24 on MediaQ 1132:

I'm not sure what every GPIO apart does. Also a TPA721 amplifier is used (datasheet can be downloaded from the Internet) but it is not important from programmer point of view since it is completely invisible to software.

Flashing the ROM

If you press "Contacts" and press the "Scroll" lever right down until it clicks, then press Reset, then release Reset but keep Contacts and Scroll Push for at least 10 seconds!) you'll get this (provided you have a CompactFlash storage card inserted into the slot): 
        LCD initialized Ok.!!
        P25 EBoot Ver. xxxxx
        SDRAM 64MB Cleared!
        Dell Axim X5
        Hardware Device ID
        xxxxxxxxxxxxxxxxxx
        Manufacturer Code
        xxxxxxxxxxxxxxxxxx
        UUID Method1 Variant = 8
        UUID Method1 Version = 1
        CF card detected!
        Init CF card ok!
        Determining Card Type....
        Storage Card Found!
        Load Image From CF Disk....
        !! PLEASE DO NOT TURN OFF !!
        No matching *.IMG file found!

If you don't have a CompactFlash card in slot, you'll enter the Diagnostic tool which is not too interesting from our point of view :)

So replacing WinCE with Linux in ROM is just a matter of time now. :-) It also means you dont have to fear to brick your Axim with an incorrect ROM; you always can re-flash it with a correct image (well, I haven't tried it but should be true :-).

JTAG

Intro We need to find out the exact JTAG connection, as it's not available from Dell. A Dell guy named Gary (Lerhaupt? http://www.lerhaupt.com/foo/archives/2003_06.html) has been able to connect JTAG to Axim, according to #handhelds.org logs, so it IS possible. However of course I'm sure he won't (be able to) tell us...

Philipp Blundell has one Axim with desoldered CPU - he just needs to trace the lines to JTAG pins but didn't find the time yet...

Marcel Telka said that the TDO pin might actually be the yellow (gold) pad between PXA and crystal (near the corner of the XTAL) of the picture at [WWW]this place

Sounds likely that some pads there are JTAG (after all the JTAG balls are in the upper area of the middle line of the PXA255).

The actual test points

These test points have NOT been functionally verified yet, so they might still be (partially) incorrect!

AndreasMohr, 040425: In fact I tried to do a JTAG connection recently, but I was UNSUCCESSFUL (JTAG Tools "detect" didn't find anything, whereas it works with my JTAG-capable Access Point), and one cause of that might be voltage potential differences between PC power supply and Axim PS. I should try to implement an opto-coupled version of my JTAG adapter sometime...

But OTOH it might very well be that we have errors in our JTAG pin list below.

So whatever you intend to do with the information below, do it CAREFULLY!

Suspected/measured JTAG pins (balls) on the PXA255 are:

Annotated pictures of the JTAG test points on Axim PCB back and front side can be found [WWW]here (back) and [WWW]also here (front)

[autumn 2004] Phil Blundell now told me that TCK is most likely wrong:

<pb_> oh ho! TCK seems to be wrong.

> pb_: really??? hmm. would be useful to know the correct pin :)

<pb_> AndiM: on your picture, you have TCK on the TP directly south of RN509.

But it appears that it's actually the TP a little way northwest of that, below

the second of the four chip resistors (counting from the right)

I haven't had the time to test it yet (my desktop PC died, and I didn't want to use my notebook parallel port, and I got my first job and moved homes, so this had to wait way too long)

How we did it

Thanks to the help of Philip Blundell (pin measurements on desoldered device) we were able to find out the JTAG pins on the Axim (using my special measurement method to close in on potential candidates first). Greg Gilbert's generous donation of a sacrificial Axim made this possible.

Philipp just checked the pins I suspected below, and it seems he had great success! (the only JTAG pin not identified yet is TDI)

Unfortunately most of his pins don't correlate with what I measured.

AndreasMohr, 0402xx: In order to find out the test points leading to JTAG balls, I used a very small conductive wire, moved it under the PXA255 to one ball line which contains some JTAG related balls, made sure to keep the wire as much as possible in contact with all balls of that ball line (by pulling the wire tightly), then tried to do some resistance measurement against various test points nearby.

When I got a winner (multimeter beeps), then I wrote down some note and then tried to verify later whether it actually is one of the JTAG pins by cross correlating with the 90 degree oriented ball line which also has that JTAG ball we suspect to be the one...

Unfortunately I wasn't able to identify too many JTAG test points that way...

Sure sounded much better than killing a whole Axim by de-microwaving a whole PXA255 Axim PCB, but it didn't help much.

Further JTAG info

Further JTAG information can be found here: http://www.xda-developers.com/jtag/ (they actually bake an XDA PCB to find out about the JTAG interface!) Visit http://axl.freeyellow.com/iPAQ/jtag.htm for a description on how to prepare the iPAQ JTAG interface... Some helpful hardware experimentation hints: http://handhelds.org/hypermail/ipaq/173/17371.html