Hi,
Processor: 6211, Big Endian
Emulator: XDS510PP Plus
Customized Target Board, having SDRAM as the memory.
I am facing the problem with SDRAM communication. I am not getting the
data back from the SDRAM exactly what I am writing in a particular
location. if I am writing 0x00000000 then in response i am getting
0xaf0000. I am using the exactly same SDRAM chip as on the
TMS320C6211DSK. and Modified my GEL file exactly same as giving in gel
folder of the CCS. Please Help to overcome this problem.
Experiment: -
SDRAM
Started by ●October 11, 2006
Reply by ●October 12, 20062006-10-12
Hi Arun,
Try writing standard patterns of 0xAAAAAAAA and 0x55555555 and see what you
are getting. This might give you some clue.
Good luck.
Regards,
Kalpesh
-----Original Message-----
From: c... [mailto:c...] On Behalf Of Arun
Kumar
Sent: Wednesday, October 11, 2006 10:21 AM
To: c...
Subject: [c6x] SDRAM
Hi,
Processor: 6211, Big Endian
Emulator: XDS510PP Plus
Customized Target Board, having SDRAM as the memory.
I am facing the problem with SDRAM communication. I am not getting the
data back from the SDRAM exactly what I am writing in a particular
location. if I am writing 0x00000000 then in response i am getting
0xaf0000. I am using the exactly same SDRAM chip as on the
TMS320C6211DSK. and Modified my GEL file exactly same as giving in gel
folder of the CCS. Please Help to overcome this problem.
Experiment: -
Try writing standard patterns of 0xAAAAAAAA and 0x55555555 and see what you
are getting. This might give you some clue.
Good luck.
Regards,
Kalpesh
-----Original Message-----
From: c... [mailto:c...] On Behalf Of Arun
Kumar
Sent: Wednesday, October 11, 2006 10:21 AM
To: c...
Subject: [c6x] SDRAM
Hi,
Processor: 6211, Big Endian
Emulator: XDS510PP Plus
Customized Target Board, having SDRAM as the memory.
I am facing the problem with SDRAM communication. I am not getting the
data back from the SDRAM exactly what I am writing in a particular
location. if I am writing 0x00000000 then in response i am getting
0xaf0000. I am using the exactly same SDRAM chip as on the
TMS320C6211DSK. and Modified my GEL file exactly same as giving in gel
folder of the CCS. Please Help to overcome this problem.
Experiment: -
Reply by ●October 12, 20062006-10-12
Arun-
> Processor: 6211, Big Endian
> Emulator: XDS510PP Plus
> Customized Target Board, having SDRAM as the memory.
>
> I am facing the problem with SDRAM communication. I am not getting the
> data back from the SDRAM exactly what I am writing in a particular
> location. if I am writing 0x00000000 then in response i am getting
> 0xaf0000. I am using the exactly same SDRAM chip as on the
> TMS320C6211DSK. and Modified my GEL file exactly same as giving in gel
> folder of the CCS. Please Help to overcome this problem.
What if you write 0x55005555? Do you get 0x55af5555? If so, then you might
carefully check byte enable signals on your SDRAM for D23..D16 lines? These lines
could be shorted, not connected, etc. on your custom board.
-Jeff
> Processor: 6211, Big Endian
> Emulator: XDS510PP Plus
> Customized Target Board, having SDRAM as the memory.
>
> I am facing the problem with SDRAM communication. I am not getting the
> data back from the SDRAM exactly what I am writing in a particular
> location. if I am writing 0x00000000 then in response i am getting
> 0xaf0000. I am using the exactly same SDRAM chip as on the
> TMS320C6211DSK. and Modified my GEL file exactly same as giving in gel
> folder of the CCS. Please Help to overcome this problem.
What if you write 0x55005555? Do you get 0x55af5555? If so, then you might
carefully check byte enable signals on your SDRAM for D23..D16 lines? These lines
could be shorted, not connected, etc. on your custom board.
-Jeff
Reply by ●October 16, 20062006-10-16
Hi Sir,
when I write 0x55005555 to the location in SDRAM and read back, I got
the same data, No error.. My problem starts from here: -
Step:-1 I switched off all the power supplies.. and restarted the target..
Step:-2 I booted up the target through Emulator XDS510PP and write to
a location in SDRAM
Program: -
void main()
{
/* Clearing all the global interrupts*/
/* Initializing the MAR registers for the Chip Select CE#0 by writing
0x1 to the MAR-8,9,10,11 registers */
*(unsigned volatile int*)0x80000000 = 0x0;
printf(" The data read back=%x\n",*&(*(unsigned volatile int*)(0x80000000)));
/* Reading back twicew thrice etc..*/
printf(" The data read back=%x\n",*&(*(unsigned volatile int*)(0x80000000)));
printf(" The data read back=%x\n",*&(*(unsigned volatile int*)(0x80000000)));
printf(" The data read back=%x\n",*&(*(unsigned volatile int*)(0x80000000)));
}
Step:3 Out put is: -
af0000
0
0
0
Step:4 I am wondered at it.
Please Help sir. I am trying my best
On 12/10/06, Jeff Brower wrote:
> Arun-
>
> > Processor: 6211, Big Endian
> > Emulator: XDS510PP Plus
> > Customized Target Board, having SDRAM as the memory.
> >
> > I am facing the problem with SDRAM communication. I am not getting the
> > data back from the SDRAM exactly what I am writing in a particular
> > location. if I am writing 0x00000000 then in response i am getting
> > 0xaf0000. I am using the exactly same SDRAM chip as on the
> > TMS320C6211DSK. and Modified my GEL file exactly same as giving in gel
> > folder of the CCS. Please Help to overcome this problem.
>
> What if you write 0x55005555? Do you get 0x55af5555? If so, then you might
> carefully check byte enable signals on your SDRAM for D23..D16 lines? These
> lines
> could be shorted, not connected, etc. on your custom board.
>
> -Jeff
>
--
World Belongs to few enthusiasts who keeps kool
when I write 0x55005555 to the location in SDRAM and read back, I got
the same data, No error.. My problem starts from here: -
Step:-1 I switched off all the power supplies.. and restarted the target..
Step:-2 I booted up the target through Emulator XDS510PP and write to
a location in SDRAM
Program: -
void main()
{
/* Clearing all the global interrupts*/
/* Initializing the MAR registers for the Chip Select CE#0 by writing
0x1 to the MAR-8,9,10,11 registers */
*(unsigned volatile int*)0x80000000 = 0x0;
printf(" The data read back=%x\n",*&(*(unsigned volatile int*)(0x80000000)));
/* Reading back twicew thrice etc..*/
printf(" The data read back=%x\n",*&(*(unsigned volatile int*)(0x80000000)));
printf(" The data read back=%x\n",*&(*(unsigned volatile int*)(0x80000000)));
printf(" The data read back=%x\n",*&(*(unsigned volatile int*)(0x80000000)));
}
Step:3 Out put is: -
af0000
0
0
0
Step:4 I am wondered at it.
Please Help sir. I am trying my best
On 12/10/06, Jeff Brower wrote:
> Arun-
>
> > Processor: 6211, Big Endian
> > Emulator: XDS510PP Plus
> > Customized Target Board, having SDRAM as the memory.
> >
> > I am facing the problem with SDRAM communication. I am not getting the
> > data back from the SDRAM exactly what I am writing in a particular
> > location. if I am writing 0x00000000 then in response i am getting
> > 0xaf0000. I am using the exactly same SDRAM chip as on the
> > TMS320C6211DSK. and Modified my GEL file exactly same as giving in gel
> > folder of the CCS. Please Help to overcome this problem.
>
> What if you write 0x55005555? Do you get 0x55af5555? If so, then you might
> carefully check byte enable signals on your SDRAM for D23..D16 lines? These
> lines
> could be shorted, not connected, etc. on your custom board.
>
> -Jeff
>
--
World Belongs to few enthusiasts who keeps kool
Reply by ●October 17, 20062006-10-17
Arun,
I didn't see the step where you init the EMIF ?
Do you have a GEL file that does that ?
- Andrew E
Arun Kumar wrote:
>Hi Sir,
>when I write 0x55005555 to the location in SDRAM and read back, I got
>the same data, No error.. My problem starts from here: -
>
>Step:-1 I switched off all the power supplies.. and restarted the target..
>
>Step:-2 I booted up the target through Emulator XDS510PP and write to
>a location in SDRAM
>
>Program: -
>void main()
>{
>/* Clearing all the global interrupts*/
>/* Initializing the MAR registers for the Chip Select CE#0 by writing
>0x1 to the MAR-8,9,10,11 registers */
>
>*(unsigned volatile int*)0x80000000 = 0x0;
>printf(" The data read back=%x\n",*&(*(unsigned volatile int*)(0x80000000)));
>/* Reading back twicew thrice etc..*/
>printf(" The data read back=%x\n",*&(*(unsigned volatile int*)(0x80000000)));
>printf(" The data read back=%x\n",*&(*(unsigned volatile int*)(0x80000000)));
>printf(" The data read back=%x\n",*&(*(unsigned volatile int*)(0x80000000)));
>}
>
>Step:3 Out put is: -
>af0000
>0
>0
>0
>
>Step:4 I am wondered at it.
>
>Please Help sir. I am trying my best
>On 12/10/06, Jeff Brower wrote:
>
>
>>Arun-
>>
>>
>>
>>>Processor: 6211, Big Endian
>>>Emulator: XDS510PP Plus
>>>Customized Target Board, having SDRAM as the memory.
>>>
>>>I am facing the problem with SDRAM communication. I am not getting the
>>>data back from the SDRAM exactly what I am writing in a particular
>>>location. if I am writing 0x00000000 then in response i am getting
>>>0xaf0000. I am using the exactly same SDRAM chip as on the
>>>TMS320C6211DSK. and Modified my GEL file exactly same as giving in gel
>>>folder of the CCS. Please Help to overcome this problem.
>>>
>>>
>>What if you write 0x55005555? Do you get 0x55af5555? If so, then you might
>>carefully check byte enable signals on your SDRAM for D23..D16 lines? These
>>lines
>>could be shorted, not connected, etc. on your custom board.
>>
>>-Jeff
>>
>>
>>
>
>
I didn't see the step where you init the EMIF ?
Do you have a GEL file that does that ?
- Andrew E
Arun Kumar wrote:
>Hi Sir,
>when I write 0x55005555 to the location in SDRAM and read back, I got
>the same data, No error.. My problem starts from here: -
>
>Step:-1 I switched off all the power supplies.. and restarted the target..
>
>Step:-2 I booted up the target through Emulator XDS510PP and write to
>a location in SDRAM
>
>Program: -
>void main()
>{
>/* Clearing all the global interrupts*/
>/* Initializing the MAR registers for the Chip Select CE#0 by writing
>0x1 to the MAR-8,9,10,11 registers */
>
>*(unsigned volatile int*)0x80000000 = 0x0;
>printf(" The data read back=%x\n",*&(*(unsigned volatile int*)(0x80000000)));
>/* Reading back twicew thrice etc..*/
>printf(" The data read back=%x\n",*&(*(unsigned volatile int*)(0x80000000)));
>printf(" The data read back=%x\n",*&(*(unsigned volatile int*)(0x80000000)));
>printf(" The data read back=%x\n",*&(*(unsigned volatile int*)(0x80000000)));
>}
>
>Step:3 Out put is: -
>af0000
>0
>0
>0
>
>Step:4 I am wondered at it.
>
>Please Help sir. I am trying my best
>On 12/10/06, Jeff Brower wrote:
>
>
>>Arun-
>>
>>
>>
>>>Processor: 6211, Big Endian
>>>Emulator: XDS510PP Plus
>>>Customized Target Board, having SDRAM as the memory.
>>>
>>>I am facing the problem with SDRAM communication. I am not getting the
>>>data back from the SDRAM exactly what I am writing in a particular
>>>location. if I am writing 0x00000000 then in response i am getting
>>>0xaf0000. I am using the exactly same SDRAM chip as on the
>>>TMS320C6211DSK. and Modified my GEL file exactly same as giving in gel
>>>folder of the CCS. Please Help to overcome this problem.
>>>
>>>
>>What if you write 0x55005555? Do you get 0x55af5555? If so, then you might
>>carefully check byte enable signals on your SDRAM for D23..D16 lines? These
>>lines
>>could be shorted, not connected, etc. on your custom board.
>>
>>-Jeff
>>
>>
>>
>
>
Reply by ●October 17, 20062006-10-17
Arun-
> when I write 0x55005555 to the location in SDRAM and read back, I got
> the same data, No error.. My problem starts from here: -
>
> Step:-1 I switched off all the power supplies.. and restarted the target..
>
> Step:-2 I booted up the target through Emulator XDS510PP and write to
> a location in SDRAM
>
> Program: -
1) Initializing MAR registers is not enough. What about other EMIF
control registers?
2) If you are initializing all EMIF regs using .gel file, then what
happens if you repeat this test using CCS memory display / change only;
i.e. upon entering CCS, without running your program?
-Jeff
> void main()
> {
> /* Clearing all the global interrupts*/
> /* Initializing the MAR registers for the Chip Select CE#0 by writing
> 0x1 to the MAR-8,9,10,11 registers */
>
> *(unsigned volatile int*)0x80000000 = 0x0;
> printf(" The data read back=%x\n",*&(*(unsigned volatile
> int*)(0x80000000)));
> /* Reading back twicew thrice etc..*/
> printf(" The data read back=%x\n",*&(*(unsigned volatile
> int*)(0x80000000)));
> printf(" The data read back=%x\n",*&(*(unsigned volatile
> int*)(0x80000000)));
> printf(" The data read back=%x\n",*&(*(unsigned volatile
> int*)(0x80000000)));
> }
>
> Step:3 Out put is: -
> af0000
> 0
> 0
> 0
>
> Step:4 I am wondered at it.
>
> Please Help sir. I am trying my best
> On 12/10/06, Jeff Brower wrote:
>> Arun-
>>
>> > Processor: 6211, Big Endian
>> > Emulator: XDS510PP Plus
>> > Customized Target Board, having SDRAM as the memory.
>> >
>> > I am facing the problem with SDRAM communication. I am not getting the
>> > data back from the SDRAM exactly what I am writing in a particular
>> > location. if I am writing 0x00000000 then in response i am getting
>> > 0xaf0000. I am using the exactly same SDRAM chip as on the
>> > TMS320C6211DSK. and Modified my GEL file exactly same as giving in gel
>> > folder of the CCS. Please Help to overcome this problem.
>>
>> What if you write 0x55005555? Do you get 0x55af5555? If so, then you
>> might
>> carefully check byte enable signals on your SDRAM for D23..D16 lines?
>> These
>> lines
>> could be shorted, not connected, etc. on your custom board.
>>
>> -Jeff
>>
> --
> World Belongs to few enthusiasts who keeps kool
> when I write 0x55005555 to the location in SDRAM and read back, I got
> the same data, No error.. My problem starts from here: -
>
> Step:-1 I switched off all the power supplies.. and restarted the target..
>
> Step:-2 I booted up the target through Emulator XDS510PP and write to
> a location in SDRAM
>
> Program: -
1) Initializing MAR registers is not enough. What about other EMIF
control registers?
2) If you are initializing all EMIF regs using .gel file, then what
happens if you repeat this test using CCS memory display / change only;
i.e. upon entering CCS, without running your program?
-Jeff
> void main()
> {
> /* Clearing all the global interrupts*/
> /* Initializing the MAR registers for the Chip Select CE#0 by writing
> 0x1 to the MAR-8,9,10,11 registers */
>
> *(unsigned volatile int*)0x80000000 = 0x0;
> printf(" The data read back=%x\n",*&(*(unsigned volatile
> int*)(0x80000000)));
> /* Reading back twicew thrice etc..*/
> printf(" The data read back=%x\n",*&(*(unsigned volatile
> int*)(0x80000000)));
> printf(" The data read back=%x\n",*&(*(unsigned volatile
> int*)(0x80000000)));
> printf(" The data read back=%x\n",*&(*(unsigned volatile
> int*)(0x80000000)));
> }
>
> Step:3 Out put is: -
> af0000
> 0
> 0
> 0
>
> Step:4 I am wondered at it.
>
> Please Help sir. I am trying my best
> On 12/10/06, Jeff Brower wrote:
>> Arun-
>>
>> > Processor: 6211, Big Endian
>> > Emulator: XDS510PP Plus
>> > Customized Target Board, having SDRAM as the memory.
>> >
>> > I am facing the problem with SDRAM communication. I am not getting the
>> > data back from the SDRAM exactly what I am writing in a particular
>> > location. if I am writing 0x00000000 then in response i am getting
>> > 0xaf0000. I am using the exactly same SDRAM chip as on the
>> > TMS320C6211DSK. and Modified my GEL file exactly same as giving in gel
>> > folder of the CCS. Please Help to overcome this problem.
>>
>> What if you write 0x55005555? Do you get 0x55af5555? If so, then you
>> might
>> carefully check byte enable signals on your SDRAM for D23..D16 lines?
>> These
>> lines
>> could be shorted, not connected, etc. on your custom board.
>>
>> -Jeff
>>
> --
> World Belongs to few enthusiasts who keeps kool
Reply by ●November 2, 20062006-11-02
I notice you mentioned that you can't write EMIF configuration register
correctly. maybe you would lower the EMIF input clock somewhat.
Hi to all... I got n serious accident 15 days back. so could not
>explain my problem last few days.. sorry...
>
>I checked my setup today again.. the problem is still as it is... I
>did many experiments that I am going to explain, please suggest me
>some thing from these experiments
>
>Step:-1
>I powered on my Target Board.. and run the program for SDRAM test...
>
>Step:2
>I got the data mismatch at every 32 successive location as: -
>Data Mismatch at location-1
>Data Mismatch at location:-32
>Data Mismatch at location:-64
>Data Mismatch at location:-96
>Data Mismatch at location:-128
>
>Step: 3
>I halted the running program and loaded the program again, I did not
>Switched off the lsupply any where and run the program again.
>
>This time: - My data is matched all the location up to 128th locations
>but again I got the same statement of mismatch from the location
>Data Mismatch at location:-160
>Data Mismatch at location:-192
>Data Mismatch at location:-224 onwards
>
>Step:4
>I halted the program without switching off the power and run the program again
>This time I got the data matched again up to the location 224 and got
>the problem again from the location
>256 onwards successively on 32th locations.
>
>Step: 5
>I powered off the system and again run the program
>I got the data mismatch at every 32 successive location as: -
>Data Mismatch at location-1
>Data Mismatch at location:-32
>Data Mismatch at location:-64
>Data Mismatch at location:-96
>Data Mismatch at location:-128
> and the Step:2,3,4,5 repeated
>
>The same program is running fine on the Starter Kit of TI but not on
>my target board.
>I am not using the GEL file as when I configured the EMIF Locations it
>is giving errors like, Trouble in writing CPU target memory at
>0x01800004, 0x01800008,0x0180000C all SDRAM registers..
>
>So I am running the program without GEL file.
>
>Please Suggest some thing
>
>Regards
>Arun Kumar
>
>--
>World Belongs to few enthusiasts who keeps kool
>
>--
>World Belongs to few enthusiasts who keeps kool
Hi to all... I got n serious accident 15 days back. so could not
>explain my problem last few days.. sorry...
>
>I checked my setup today again.. the problem is still as it is... I
>did many experiments that I am going to explain, please suggest me
>some thing from these experiments
>
>Step:-1
>I powered on my Target Board.. and run the program for SDRAM test...
>
>Step:2
>I got the data mismatch at every 32 successive location as: -
>Data Mismatch at location-1
>Data Mismatch at location:-32
>Data Mismatch at location:-64
>Data Mismatch at location:-96
>Data Mismatch at location:-128
>
>Step: 3
>I halted the running program and loaded the program again, I did not
>Switched off the lsupply any where and run the program again.
>
>This time: - My data is matched all the location up to 128th locations
>but again I got the same statement of mismatch from the location
>Data Mismatch at location:-160
>Data Mismatch at location:-192
>Data Mismatch at location:-224 onwards
>
>Step:4
>I halted the program without switching off the power and run the program again
>This time I got the data matched again up to the location 224 and got
>the problem again from the location
>256 onwards successively on 32th locations.
>
>Step: 5
>I powered off the system and again run the program
>I got the data mismatch at every 32 successive location as: -
>Data Mismatch at location-1
>Data Mismatch at location:-32
>Data Mismatch at location:-64
>Data Mismatch at location:-96
>Data Mismatch at location:-128
> and the Step:2,3,4,5 repeated
>
>The same program is running fine on the Starter Kit of TI but not on
>my target board.
>I am not using the GEL file as when I configured the EMIF Locations it
>is giving errors like, Trouble in writing CPU target memory at
>0x01800004, 0x01800008,0x0180000C all SDRAM registers..
>
>So I am running the program without GEL file.
>
>Please Suggest some thing
>
>Regards
>Arun Kumar
>
>--
>World Belongs to few enthusiasts who keeps kool
>
>--
>World Belongs to few enthusiasts who keeps kool
