TMS320C5402 DSP board can not Load the programmes

Jinqi Zhou,

It is kind of hard to guess your problem based on the description that you have given so far.  What is your target board?  Is it your own design or one from TI? 

The fact that the board boots correctly occasionally hints at a problem with your startup process.  Can you give more information on what you meant by closing and opening the constant current source.  Do you mean the power supply?  Have you verified that you have implemented RESET circuit correctly?  Also, are power and clocking done correctly?  One other point to look at is whether you have all unused pins terminated correclty, especially the JTAG signals.

From the SW side, how do you determine that the EPROM is programmed correctly?  Are you specifying Boot Mode correctly?  If wrong boot mode selected, then you would see unexpected results.  Do you have some way to verify that your SW is functional?  Testing on TI EVM or simulator would be a good choice for this type of task.

Please provide some additional information so we can get to the root cause of your problem.

Thank you for your reply.

For your questions, I’d like describe my problem in more details. The target board is my own design.

   I have written program into the eeprom of the board, and the boot mode is 8 bit parallel boot mode. But the vc5402 cannot read the program correctly.

   I measure the address pin of vc5402, a0 to a15 is all high level. And the pins of PS,IS,mstrb and iostrb show all high level , but the xf pin and ds pin showed low level.

   I use CCS to read directly the program located in 0x8000 of eeprom without and code running. The ce of the eeprom is always set to low level. The oe of the eeprom is controlled by the mstrb and r/w. Here is the phenomenon in different board:

a)   the time of mstrb enable signal is too short to let the eeprom output data. And the CCS cannot show any data in eeprom.

b) The signal of mstrb is right, and control the oe,ce,r/w and the address of eeprom is all right, is right. but still the eeprom cannot output any data. But we can read the eeprom in specific eeprom writer correctly.

Last but not least , I want to know the ESD of vc5402. The board worked well but now it can not bootload the program from eeprom, even after reset the vc5402, and the reset function was good too.

Oh, I nearly forgot asking one question to you. What do you mean that “One other point to look at is whether you have all unused pins terminated correctly, especially the JTAG signals”. For the JTAG  in my board, just one EMU1/OFF pin is set to the high level, the others are terminated nothing. Is there any effect on the vc5402? Besides, HPI and INT pins are unused, and are all terminated nothing. What do you think of it? And what should I deal with these unused pins?

I have a couple of questions to clarify what you stated above.

You say that you use CCS to read the EEPROM.  Is that working correctly?  It's not clear from your statements.  Can you connect CCS to your target and read the CPU registers to indicate that the JTAG interface is working?

The fact that you don't see any activity on the EMIF seems like the DSP might not be booting correctly.  did you follow the requirements specified in the Bootload Application Report (SPRA618B)?  The location of the boot table needs to be specified at location 0xffff in IO or Data Space. 

How do you have the CLKMD pins configured to a supported mode?  These pins are read at Reset to configure the onchip PLL.  Do you see CLKOUT begin toggling after device comes out of Reset?

According to the Data Manual the XF pin should go high if a valid Rest signal is seen by the DSP.  You say that this pin is low, so that indicates that the DSP is not initiating the Rest process which would include reading the external memory.

I have request to IC team on EDS rating for your DSP.  I do not find this information in the Data Manual.  Do you have some concern that the board has been damaged by ESD?  Do you have a second board that you can test?

The question about unused pins being terminated correctly applies to Input signals.  The EMU1 pin is of most concern.

Regards,
Tom

Hi, TommyG

Here is my answer for your questions:

1)    I connect CCS to my target and read the CPU registers and found out that the mp/mc=1, which pin is connected to low level in my board. and DROM=0 which should be 1, something is wrong with it.

2)   I follow the requirements specified in the bootload application Report and it worked well. Now it can't iniciate itself.

3)   The CLKMD pin is set all '1', and the frequance from clkout is 5MHz, half of the input freqence, as I set. But the waveform from clkout pin is no good, there are 100mV of jitter on the waveform.

4)    Yes. The XF pin is low, it indicates that the DSP is not initiating the Reset process. The pin is high when DSP get reset.

5)    I have couples of such target boards I designed. They worked correctly before, however, now they don’t work as they are supposed to. Therefore, I am considering whether EDS has damaged my boards, and want to get some useful suggestions hopefully.

6) I read the eeprom SST39VF040 through CCS, and found out the CCS can't get any data from EEPROM at all .the CE OE and W/R is all right.

     I use Oscilloscopes to probe the data output. the waveform didn't have any change. but the EEPROM could be read and written correctly under specific EEPROM programmer.