Fez Hydra on Linux

I think this is an interesting discussion - I wonder what the difference is between the Rev 1.1 board and the Rev 1.2 board (@ James that might be something you can answer on GHI’s time :wink: only if you can get time, that is )

There was some improvement on power supply and decoupling capacitors. Nothing changed that would effect the code, which is very confusing on why there is a problem running linux. We are interested in knowing and will look into it when there are no high priority NETMF tasks.

Raj, what if we ship you a 1.2 board?

@ Gus - it is difficult to configure linux on hardware that is not working perfectly

Changes to power or decoupling should not cause this issue, but since 2 people have had the same issue it could be something with the revision. If someone with the rev 1.1 could test, that may help to verify.

If not, you could ship a 1.2 board for further tests.

@ Gus - Forgive me if I am wrong, but I remember reading somewhere that the rev 1.1 Hydra had the RTC crystal pre-installed, I can’t imagine that it would make a difference, but just to cover the bases.

That may be a lead. I am away from my development system right now and can not verify but the Linux boot may start with the slow clock and then switch clocks. If there is no crystal installed the slow clock may fail.

I could be wrong on this but would not rule out the possibility.

Actually, I believe that you are correct. Having the clock fail might send the kernel into kernel panic since the clock is one of the most basic functionality. If this happens during the initial boot phase of the base kernel, then it may not have yet set up STDIO to report the error.

@ malconxx

Will it be possible for you to solder a crystal on the Hydra ?

The hydra did not with me, a friend is doing tests with it.

Update

1.GHI shipped me a 1.2 rev board for testing. This is pretty awesome of them to support a community project.
2.Out of the bag the Hydra did not boot Linux
3.I stuck in a RTC crystal that I had lying around and wiggled it to make contact
4.Linux boots without any problem

taylorza was correct in pointing out this difference between the 1.1 and 1.2 boards.

Please solder in a 32.768Khz crystal and you should be good to boot with a 1.2 rev board.

Gus, I could ship the board back to GHI or to someone who needs one delivered in the Toronto area, please let me know.

Hello,

will it be possible to provide step by step instructions on how to solder 32.768 crystal on Hydra 1.2 board
-Thanks
-Alok

Its as easy as getting an appropriate crystal, connecting it to the location on the board, and soldering it.

1.Get the correct crystal eg. : VT200F-6PF20PPM Seiko Instruments | Crystals, Oscillators, Resonators | DigiKey
You should be able to get this from any parts source,

2.Once you get it insert the leads into the board on the Q1 footprint / silk screen, see attached image

3.Solder from the other side of the PCB and you are good to go

Thank you very much Rajesh, as always, you were a great help

Hello,

I soldered 32 Khz crystal , now I get following messages and it hangs , I have pressed reset button severla times which is evident in following messages

RomBOOT

Start AT91Bootstrap…
1SRAM REG 0x64
Downloading image…
*** f_open, File name: [u-boot.bin]: error!
RomBOOT
Start AT91Bootstrap…
1SRAM REG 0x64
Downloading image…
*** f_open, File name: [u-boot.bin]: error!
RomBOOT
RomBOOT
RomBOOT
Start AT91Bootstrap…
1SRAM REG 0x64
Downloading image…
*** f_open, File name: [u-boot.bin]: error!
RomBOOT
Start AT91Bootstrap…
1SRAM REG 0x64
Downloading image…
*** f_open, File name: [u-boot.bin]: error!
RomBOOT
Start AT91Bootstrap…
1SRAM REG 0x64
Downloading image…
*** f_open, File name: [u-boot.bin]: error!

what could be wrong.

-Thanks
-Alok

Alok

Looks like you do not have u-boot.bin file on the sdcard.

Rajesh,

these are files which you provided earlier in this forum , I also built u-boot.bin according to instruction provided on web page Home - GHI Electronics in both the cases same error shows up which I am copying and pasting below

RomBOOT

Start AT91Bootstrap…
Downloading image…
*** f_open, File name: [u-boot.bin]: error!

looks like something wrong with u-boot.bin, could you please provide these files again.

-Thanks
-Alok

It may not be the file.

Try this

1.Delete partition(s) on SD Card
2.Create new BOOT partition using a different size, eg. if you had 100MB before change to 150MB now
3.Copy the 3 files into the BOOT partition
4.Test

If it fails, please try with a different SD Card to verify.

Rajesh,

I deleted 50Meg partition and re-created 100 Meg partition (FAT16 type) and copied three files (boot.bin,u-boot.bin,image.bin) , and inserted in microSD Card attached to hydra and I get following error, I have tried with two different SD card.

Start AT91Bootstrap…
1SRAM REG 0x64
Downloading image…
*** f_open, File name: [u-boot.bin]: error!

-Thanks
-Alok

in file ./driver/MEDSdcard.c

in function

unsigned int load_SDCard(void *dst)
{
DWORD dwAddress;

unsigned int ByteRead = 0;

unsigned char ret;

ret = MEDSdcard_Initialize(&medias[0], BOARD_SD_MCI_ID_USE);

memset(&fs, 0, sizeof (FATFS));     // Clear file system object

res = f_mount(0, &fs);
if (res != FR_OK) {
    dbg_log(1, "*** f_mount: error!\n\r");
    while (1) ;
}

res = f_open(&fileObject, OS_IMAGE_NAME, FA_OPEN_EXISTING | FA_READ);

if (res != FR_OK) {
    dbg_log(1, "*** f_open, File name: [%s]: error!\n\r", OS_IMAGE_NAME);
    while (1) ;
}

in last 4 lines , my hydra is failing and it keeps on looping in while loop , I have verified that u-boot.bin file is on MicroSD card exists

-Thanks
-Alok