v5 Beta FAT16 component SDHC

Post by **Benj** » Wed Jul 18, 2012 12:18 am

Hello Nicolas,

I haven't had time to test this on hardware yet but it compiles now so hopefully it should work. I need to buy in some SDHC cards to do the testing so may take a few days. In the meantime if you want to give it a go then backup the FC5_PIC_FAT16.c file in the "Flowcode xxx\v5\Components" folder and then copy in the attached file before starting Flowcode.

FC5_PIC_FAT16.c: (53.89 KiB) Downloaded 418 times

Let me know how you get on.

Nico595 · Post by **Nico595** » Wed Jul 18, 2012 12:42 am

Thank you very much Ben

,

I test this component today with my SDHC Card.

Post by **JohnCrow** » Wed Jul 18, 2012 10:14 am

Hi Benj

Does this mean the new component will work with cards bigger than 2GB ?

Also will It work formatted FAT16 or does it need to be FAT32.
I'll give it a try tonight if I get chance with an 8GB card.

Nico595 · Post by **Nico595** » Wed Jul 18, 2012 10:20 am

Hi Ben,

I ve renamed FC5_PIC_FAT16.c in FC5_ARM_FAT16.c and replaced in Components Folder.
Unfortunately, component not working with all SDCard.

- With SDCard no SDHC < 2GO, After Init FAT macro I have Retval = Dec 237 (0xED hex code error -> SECTBYTES ERROR)
(Ok with old FAT component version)

- With SDCard SDHC 4GO 15MB/s (Class 4), After Init FAT macro I have Retval = Dec 254 (0xFE hex code error -> NORESP ERROR)

I ve tried in FAT32 select and AUTO select but with the even problem.

Thank for your reply,
Regards,
Nicolas.

Post by **Benj** » Wed Jul 18, 2012 11:03 am

Hi Nicolas & John,

Ok thanks for testing and providing the results so quickly. A bit disappointed it didn't work but not giving up just yet. I will have another dig in tonight and see if I can get any further.

I was using the Microchip SD stack as a template as this works nicely on a range of cards however it is really overcomplicated and bloated and it's not until you start to try and port it that you realise just how bad it is. Code replicated in multiple places, definitions that are bypassed or not used in the code and lots of weird compiler workarounds that would likely just cause issues with BoostC and GCC. Last night I found an AVR GCC source which is really simple and fits in with the current code base. It is said to work but obviously I have done something wrong if this is not currently working with your hardware.

The sector bytes error is interesting and may help me to debug the issue.

Does this mean the new component will work with cards bigger than 2GB ?

Yes eventually I hope to be able to support larger cards and I also want to make the SPI transactions post init much faster. One use for this may be with the DSP type functionality we are doing to provide high quality streaming audio etc.

Also will It work formatted FAT16 or does it need to be FAT32.

I think FAT16 only supports cards <= 2GB so FAT32 may have to be used on larger cards, not 100% on this one. Do you get FAT as an option when you format a card larger then 2GB?

Nico595 · Post by **Nico595** » Wed Jul 18, 2012 11:21 am

Ok Ben,

For information:
My Small SDCard is Transcend 2GO formated in FAT32, OK with old FAT component version (but with speed writting problem when data on card is > 100 MO).
My SDHC Card is SanDisk 4GO 15MB/s Class 4 formated in FAT32 -> NORESP ERR

Good Job for tonight, I hope it will work

Nicolas.

Post by **Benj** » Wed Jul 18, 2012 7:12 pm

I think FAT16 only supports cards <= 2GB so FAT32 may have to be used on larger cards, not 100% on this one. Do you get FAT as an option when you format a card larger then 2GB?

Just noticed that all the SDHC cards seem to say FAT32 which I guess is understandable as a FAT16 file system would need huge cluster sizes and therefore would have large levels of redundant sectors if you had lots of small files on the card.

The quest for SDHC continues....

Post by **Benj** » Wed Jul 18, 2012 7:16 pm

The SECTBYTES ERROR is related to whether the start of the partition is found. I have not edited this code so I wonder why this is failing to work correctly?

Hm, the plot thickens.

Post by **Benj** » Wed Jul 18, 2012 7:53 pm

Right I have found a couple of bugs that meant that the sector read and write functions were not working correctly which probably explains the SECTBYTE error. I think this has now been fixed and I've again checked it compiles. I'm hoping this will at the very least allow the originally working card to fire up again. I have a v5 working card here so I will give this a go tonight if I get time.

Not sure what the issue is with the SDHC as this code looked ok, definitely worth retesting to see if it gets any further. In the example code it calls the init function 20 times in a loop checking for a return value of 0 before giving up.

Post by **Benj** » Thu Jul 19, 2012 12:59 am

No go here

Still working on it. Maybe tomorrow!

The card init is working ok but when getting to the part where it should be reading the MBR I am just getting 0's. Hopefully should be simple to fix.

Nicolas' card is firing up correctly so it looks like the issue is just with my card, still no go on the SDHC type cards yet though.

Nico595 · Post by **Nico595** » Thu Jul 19, 2012 11:03 am

Hello Ben,
(mp copy)

Better but other problem:

With SDCard No SDHC 2GO FAT32, Init FAT OK
Next step -> Scan_Current_Folder OK But name of folder is no good, "TRON" becomes "puovpvpwpxq"
Result -> Open Folder Error

With SDHC SDCard Error 254 on Init FAT

My prog is a loop of Init_Fat if Retval > 0 with a small delay after Init_Fat.
(.fcm export in mp)

- I ve tested in Software mode, exactly even problem.
- I ve tested with "mx_sdhc_flag = 1" fixed in macro, but Retval = 254 with SDHC Card.

Regards,
Nicolas

Post by **Benj** » Thu Jul 19, 2012 8:19 pm

Cheers Nicolas,

Right I'm back on with it now so I will let you know how I get on. Hopefully I can fix the file system related errors.

As for the SDHC, still not managed to get hold of one for testing yet so I am continuing with this standard class 1 SD that did not work previously, hopefully I can get somewhere with this. Annoying thing is it works great with the Microchip stack and so can work via the SPI interface!

If my SDHC card hasn't arrived tomorrow then I will nip down to the supermarket so I have one for the weekend.

This will not beat me

(famous last words)

Nico595 · Post by **Nico595** » Thu Jul 19, 2012 9:09 pm

Super Ben

I stay in waiting ...

I use FAT16 component of V5 ARM in FAT16 and FAT32 with small change in macro, ok with SPI hardware.

Post by **Benj** » Fri Jul 20, 2012 9:22 am

Hello,

Right I think I have a stable version now. At least it's working with the two standard SD cards I have, 1 which worked with the original v5 code and 1 which didn't.

Next step is getting the SDHC card working, if it's not already

I have also implemented a speed increase in the component code, if the init function fails then the code will try and dial back the speed for initialisation but then push it back up for the data transfer. Seems to be working well and confirmed running faster then the original v5 code on an ECIO PIC

FC5_PIC_FAT16.c: (56.51 KiB) Downloaded 358 times

PIC_CAL_SPI.c: (12.49 KiB) Downloaded 314 times

AVR_CAL_SPI.c: (16.87 KiB) Downloaded 334 times

ARM_CAL_SPI.c: (11.87 KiB) Downloaded 321 times

PIC16BIT_CAL_SPI.c: (21.36 KiB) Downloaded 317 times

Let me know how you get on with the new ARM CAL file, I have tested in software mode and it appears to be working.

- Edit - Confirmed AVR also working in SW mode

Nico595 · Post by **Nico595** » Fri Jul 20, 2012 12:09 pm

Hi Ben,

I test with my hardware immediately

Nico595 · Post by **Nico595** » Fri Jul 20, 2012 1:35 pm

Ok Ben,

I have Retval = 237 with this Mx_Init_Card.

If I replace Mx_Init_Card by this Mx_Init_Card it's OK with 2GO SDCard and new CAL, FAT, files.

MX_UINT8 RetVal;
MX_SINT16 i;

delay_ms(50); //Wait for card to power up

//Init SPI
%a_SPI_Master_Init(); //Call initialise function

FC_CAL_Bit_High_DDR (%a_MX_SPI_FAT_CS_PORT, %a_MX_SPI_FAT_CS_TRIS, %a_MX_SPI_FAT_CS_PIN); // Chip Select pin is disabled

for(i = 0; i < 10; i++) //Start CARD in SPI mode
{
%a_SPI_Master_Byte(0xFF); //Send 10*8=80 clock pulses
}

FC_CAL_Bit_Low ( %a_MX_SPI_FAT_CS_PORT, %a_MX_SPI_FAT_CS_PIN); // Chip Select pin is enabled
delay_us(50); //Wait 50us

//Command 0 - Goto Idle State - Resets card
RetVal = MX_Send_Cmd(0,0);

FC_CAL_Bit_High ( %a_MX_SPI_FAT_CS_PORT, %a_MX_SPI_FAT_CS_PIN); // Chip Select pin is disabled

if (RetVal != 1)
return MX_ERR_NOCARD; //Card not initialising or not present

delay_us(50); //Wait 50us
FC_CAL_Bit_Low ( %a_MX_SPI_FAT_CS_PORT, %a_MX_SPI_FAT_CS_PIN); // Chip Select pin is enabled

//Command 1 - Send Operating Condition - Reports the status of the card
i = 10000;
do
{
RetVal = MX_Send_Cmd(1,0);

FC_CAL_Bit_High ( %a_MX_SPI_FAT_CS_PORT, %a_MX_SPI_FAT_CS_PIN); //CARD-Disabled
}
while ((RetVal) && (i--)); //Loop while processing command 1

if (RetVal)
return MX_ERR_NORESP; //Card not responding

//Need a good way of applying this speed update!!!

//%a_SPI_Master_Uninit(); //Disable Low Speed
//%a_FAT.AsByteus_mode = 1; //SPI Bus Mode //Check this value
//%a_SPI_Master_Init (); //Enable Hi Speed

//Hardware Mode only
AT91F_SPI_CfgCs(AT91C_BASE_SPI, 0, (12 << FCV_SPI_SPEED) | AT91C_SPI_NCPHA | AT91C_SPI_CPOL); // 8MBaud @ 48MHz MCK

return MX_ACK_OK; //Card Ready

and MX_SPI_BYTE

MX_UINT8 RetVal;
MX_SINT16 i;

delay_ms(50); //Wait for card to power up

//Init SPI
%a_SPI_Master_Init(); //Call initialise function

FC_CAL_Bit_High_DDR (%a_MX_SPI_FAT_CS_PORT, %a_MX_SPI_FAT_CS_TRIS, %a_MX_SPI_FAT_CS_PIN); // Chip Select pin is disabled

for(i = 0; i < 10; i++) //Start CARD in SPI mode
{
%a_SPI_Master_Byte(0xFF); //Send 10*8=80 clock pulses
}

FC_CAL_Bit_Low ( %a_MX_SPI_FAT_CS_PORT, %a_MX_SPI_FAT_CS_PIN); // Chip Select pin is enabled
delay_us(50); //Wait 50us

//Command 0 - Goto Idle State - Resets card
RetVal = MX_Send_Cmd(0,0);

FC_CAL_Bit_High ( %a_MX_SPI_FAT_CS_PORT, %a_MX_SPI_FAT_CS_PIN); // Chip Select pin is disabled

if (RetVal != 1)
return MX_ERR_NOCARD; //Card not initialising or not present

delay_us(50); //Wait 50us
FC_CAL_Bit_Low ( %a_MX_SPI_FAT_CS_PORT, %a_MX_SPI_FAT_CS_PIN); // Chip Select pin is enabled

//Command 1 - Send Operating Condition - Reports the status of the card
i = 10000;
do
{
RetVal = MX_Send_Cmd(1,0);

FC_CAL_Bit_High ( %a_MX_SPI_FAT_CS_PORT, %a_MX_SPI_FAT_CS_PIN); //CARD-Disabled
}
while ((RetVal) && (i--)); //Loop while processing command 1

if (RetVal)
return MX_ERR_NORESP; //Card not responding

//Need a good way of applying this speed update!!!

//%a_SPI_Master_Uninit(); //Disable Low Speed
//%a_FAT.AsByteus_mode = 1; //SPI Bus Mode //Check this value
//%a_SPI_Master_Init (); //Enable Hi Speed

//Hardware Mode only
AT91F_SPI_CfgCs(AT91C_BASE_SPI, 0, (12 << FCV_SPI_SPEED) | AT91C_SPI_NCPHA | AT91C_SPI_CPOL); // 8MBaud @ 48MHz MCK

return MX_ACK_OK; //Card Ready

Nico595 · Post by **Nico595** » Fri Jul 20, 2012 2:14 pm

I've tried with FC5_PIC_FAT16.c version2 and new CAL file:

The better result, Without change code in component Macro Init Card is OK, but with this problem of bad folder name ...

FC5_PIC_FAT16.c version2

Benj wrote:Right I have found a couple of bugs that meant that the sector read and write functions were not working correctly which probably explains the SECTBYTE error. I think this has now been fixed and I've again checked it compiles. I'm hoping this will at the very least allow the originally working card to fire up again. I have a v5 working card here so I will give this a go tonight if I get time.Not sure what the issue is with the SDHC as this code looked ok, definitely worth retesting to see if it gets any further. In the example code it calls the init function 20 times in a loop checking for a return value of 0 before giving up.

Attachments

FC5_PIC_FAT16.c

(56.67 KiB) Downloaded 9 times

Post by **Benj** » Fri Jul 20, 2012 6:48 pm

Right I'm posting from my mobile due to lack on met at home. Just bought two sdhc cards and they are both working correctly with the current posted code

I will try and increase the reliability but looking good at this end.

Post by **Benj** » Fri Jul 20, 2012 7:01 pm

Using the current code I have 6 working cards 2 of which are sdhc 1 2GB and 1 4GB. The only card I have which isn't working is a mmc type card. May have got the prescaler values wrong for the arm so I will repeat the tests on this

Nico595 · Post by **Nico595** » Fri Jul 20, 2012 7:58 pm

Very good news

Thank Ben for your working,

My hardware is like this dev card:
https://www.olimex.com/dev/pdf/SAM7S-P64.pdf

I use SPI hardware.

For test, I can take Olimex dev card.

With Flowcode, in macro component, I have Mx_macro duplicate, may be problem with new component file .c

I search this WE why my SDCard answer 237, and SDHC 254 ...

Good luck Ben.

Nico595 · Post by **Nico595** » Fri Jul 20, 2012 11:32 pm

I've tested with new component FAT16(1) and replaced Init_Fat macro in .fcm prog but even problem, retval = 237

...

Nico595 · Post by **Nico595** » Sat Jul 21, 2012 12:07 am

Ok

,

I've imagined a problem with MBR reading in init FAT, I've placed your lower speed SPI mode in Init_FAT macro and SDCard 2GO is OK with your latest FAT component code in SPI hardware.

delay_ms(50);

retval = MX_Get_Buffer(0); //Read Master Boot Record if available

//if (retval)
//{
// %a_SPI_Set_Prescaler(4); //Try again

//retval = MX_Get_Buffer(0); //Read Master Boot Record if available

if (retval)
return MX_ERR_SECTBYTES;
//}

->

delay_ms(50);

retval = MX_Get_Buffer(0); //Read Master Boot Record if available

if (retval)
{
%a_SPI_Set_Prescaler(4); //Try again at a lower speed

retval = MX_Get_Buffer(0); //Read Master Boot Record if available

if (retval)
return MX_ERR_SECTBYTES;
}

Finally I test Prescaler value for Init and Read MBR for best SPI speed value:

retval = MX_Init_Card(); //Startup Card

if (retval)
{
%a_SPI_Set_Prescaler(1); //Try again

retval = MX_Init_Card(); //Read Master Boot Record if available

if (retval)
{
%a_SPI_Set_Prescaler(2); //Try again at a lower speed

retval = MX_Init_Card(); //Read Master Boot Record if available

if (retval)
{
%a_SPI_Set_Prescaler(4); //Try again at a lower speed

retval = MX_Init_Card(); //Read Master Boot Record if available

if (retval)
{
%a_SPI_Set_Prescaler(64); //Try again at a lower speed

retval = MX_Init_Card(); //Read Master Boot Record if available

if (retval)
return(retval); //Startup error or no card
}
}
}
}

delay_ms(50);

retval = MX_Get_Buffer(0); //Read Master Boot Record if available

if (retval)
{
%a_SPI_Set_Prescaler(1); //Try again

retval = MX_Get_Buffer(0); //Read Master Boot Record if available

if (retval)
{
%a_SPI_Set_Prescaler(2); //Try again at a lower speed

retval = MX_Get_Buffer(0); //Read Master Boot Record if available

if (retval)
{
%a_SPI_Set_Prescaler(4); //Try again at a lower speed

retval = MX_Get_Buffer(0); //Read Master Boot Record if available

if (retval)
{
%a_SPI_Set_Prescaler(64); //Try again at a lower speed

retval = MX_Get_Buffer(0); //Read Master Boot Record if available

if (retval)
return MX_ERR_SECTBYTES;
}
}
}
}

and I've replaced Prescaler number in Mx_Init_Card

//%a_SPI_Master_Uninit(); //Disable Low Speed
//%a_FAT.AsByteus_mode = 1; //SPI Bus Mode //Check this value
//%a_SPI_Master_Init (); //Enable Hi Speed

%a_SPI_Set_Prescaler(2);

return MX_ACK_OK; //Card Ready

Finally after Init, Prescaler value high for best SPI speed value:

//%a_SPI_Master_Init (); //Enable Hi Speed

%a_SPI_Set_Prescaler(1);

return MX_ACK_OK; //Card Ready

Can you tell me if it's ok ?
SPI_Set_Prescaler() is very a good idea

For SDHC I have 254 return.

EDIT:

After more test with 2GO SDCard, reading OK, problem with writting -> Macro Create_File, no file created on SDCard or file without name.

Nico595 · Post by **Nico595** » Sat Jul 21, 2012 9:23 pm

With SDHC Card (test with two SDHC Card) I have NORESP ERROR.

Problem with Mx_Init_Card ACMD command ...

In Defines of FAT Component I add code error:

//Common Defines
#define MX_ERR_NOCARD 0xFF
#define MX_ERR_NORESP 0xFE
#define MX_ERR_SDHCNORESP 0x33

In Mx_Init_Card

do
{
RetVal = MX_Send_Cmd(APP_CMD, 0x00000000); //CMD55, must be sent before sending any ACMD command
RetVal = MX_Send_Cmd(SD_SEND_OP_COND, 0x40000000); //ACMD41

i++;
if(i>0xfe)
{
return MX_ERR_SDHCNORESP; //time out, card initialization failed
}

I have error code 51 Dec -> 0x33 -> MX_ERR_SDHCNORESP -> time out, card initialization failed

After this Cmd I've created FCV for return value of RetVal

RetVal = MX_Send_Cmd(SD_SEND_OP_COND, 0x40000000); //ACMD41

RetVal = 1 Dec for %a_SPI_Set_Prescaler(); >= 1, But RetVal return must be 0 Dec (0x00)

EDIT 22.07.2012:

OUFFFFFF, in French

SDHC Init OK,

My power of SDCard is a PA7, no good with SDHC

, after powered SDCard with 3.3v supply, SDHC Card Init OK

Perfect, Thank you very much Ben

Always problem with the Create_File macro, file is created but without name, result, bug SDCard reading.

Post by **Benj** » Mon Jul 23, 2012 9:26 am

Hi Nico,

Sorry I dug out all my ARM hardware on the weekend and then realised I didn't have v5 for ARM installed at home and as my net was down (and still is!) I couldn't download a copy. I will install tonight and specifically test the create file and write functions. Thanks for helping with the testing.

Nico595 · Post by **Nico595** » Mon Jul 23, 2012 1:26 pm

Hi Ben,

No excuse, Weekend is for relaxation and no working

My card was powered with PA7 for reset, I have a problem, when I've readed card, I can't doing a new init_Fat, Card is no found. I must reset ARM for restart the reading card. I search later ...

I'm very happy FAT16 is OK with SDHC Card. I will test with others Card, faster.

For Create_File macro I think file is created without name but you must tested with your prog and hardware.

Good job Ben, Thanks

PS: If you want, you can delete my no necessary old post

Matrix user forums

v5 Beta FAT16 component SDHC

v5 Beta FAT16 component SDHC

Re: SD Card not response

Re: SD Card not response

Re: SD Card not response

Re: SD Card not response

Re: SD Card not response

Re: SD Card not response

Re: SD Card not response

Re: SD Card not response

Re: v5 Beta FAT16 component SDHC

Re: v5 Beta FAT16 component SDHC

Re: v5 Beta FAT16 component SDHC

Re: v5 Beta FAT16 component SDHC

Re: v5 Beta FAT16 component SDHC

Re: v5 Beta FAT16 component SDHC

Re: v5 Beta FAT16 component SDHC

Re: v5 Beta FAT16 component SDHC

Re: v5 Beta FAT16 component SDHC

Re: v5 Beta FAT16 component SDHC

Re: v5 Beta FAT16 component SDHC

Re: v5 Beta FAT16 component SDHC

Re: v5 Beta FAT16 component SDHC

Re: v5 Beta FAT16 component SDHC

Re: v5 Beta FAT16 component SDHC

Re: v5 Beta FAT16 component SDHC