Problem with Bootloader and EVM5515

If your LED.out works, as long as it does not use SARAM31 and program the PLL correctly, you should be able to use hex55 to generate a .bin file to boot from NOR or NAND.  Try NOR first since it is the first peripheral to check for boot image. Use the programmer in CSL to program boot image. The hex55 command to use is:

hex55 -i LED.out -o lED.bin -boot -v5505 -b -serial8.

Regards.

I will try this again exactly as you have indicated on Hex55.  This is an area that could be the problem.

One question still I do not get.  The NOR memory I believe is 16 bits and the NAND is 8 bits wide.  I don't have the schematic in front of me so could be reversed.  Should the -serial8 part of the command change for the difference in number of bits?

The other question is how do I tell the EVM board has the bootloader contained in its ROM?

Thanks

The serial8 works for both. I think it refers to something else other than the bus width.

No, you cannot tell which storage has the boot image. I would recommend using the programmer.out in CSL 3.03 to erase all 4 avaliable storage devices.

This is the display message of programmer.out. First use the gel file to change the PLL to 12.288MHz since I2C only work on 12.288MHz. After you follow the 1, 2, 3 & 4, SD should work.

Choose the device...

1x - NAND Flash [CSx: 2,3,4,5]

2x - NOR Flash [CSx: 2,3,4,5]

3 - SPI EEPROM (16 bit address)

4 - IIC EEPROM

5 - MMC

6 - SD

7x - SPI Serial Flash (24 bit address) [PinMap x: 1=MODE5,2=MODE6]

8 - MCSPI Serial Flash

141C:\Userdata\C5515\Boston_Scientific\Boot\cleaner.bin

NAND Flash...

Writing data to NAND...

Writing Sector 0...

Writing Sector 1 (MBR)...

Flushing Data...

Writing Boot Record...

Flushing Data...

Opening C:\Userdata\C5515\Boston_Scientific\Boot\cleaner.bin...

Input file opened

Writing Boot-image to NAND...

Programming... [TotalSize=256 (0x100)]

Programming... [TotalSize=512 (0x200)]

Programming... [TotalSize=550 (0x226)]

Flushing Data...

done

221C:\Userdata\C5515\Boston_Scientific\Boot\cleaner.bin

NOR Flash...

The Flash ID is 89, 8818, a2a5

Erasing chip (NOR)...

Writing data to device...

Opening C:\Userdata\C5515\Boston_Scientific\Boot\cleaner.bin...

Input file opened

Programming Complete

311C:\Userdata\C5515\Boston_Scientific\Boot\cleaner.bin

SPI EEPROM...

Writing data to device...

Opening 1C:\Userdata\C5515\Boston_Scientific\Boot\cleaner.bin...

Input file opened

Programming Complete

41C:\Userdata\C5515\Boston_Scientific\Boot\cleaner.bin

IIC EEPROM...

Writing data to device...

Opening C:\Userdata\C5515\Boston_Scientific\Boot\cleaner.bin...

Input file opened

Programming Complete

 Regards.

This was an immense help.

First the NOR write and cleaner are not working on the NOR memory.  The NAND works correct theIIC EEP:ROM works also.  But the 311C:.... to SPI EEPROM failed by never getting out of program after input file opened.

The big problem is the NOR>

Response is  after your recommended 221C:\...\cleaner.bin

input file opened

WRITE ERROR at 0x0000

Wrote 0x0000 read 0x00b2

=====

I now also tried again to use the TEST files from the EVM5515 kit.  The NAND works fine but the NOR now fails in a similar way.

This could be the problem as  there is something  in the NOR memory that cannot be erased.  It then boots to junk and never goes to the NAND or the SD card.

What do you recommend?

====

Question:   If the NOR is stuck how am I able to load programs using CC.   Does this take over the boot loader process?

Thanks

Dennis

The way I am thinking is to lift the RESET pin and ground it thus to bypass NOR. Also, have you use GEL to set  PLL to12.288MHz for SPI to work?

Hope this help.

Regards.

I will try the RESET ground later today.

I do not know how to use the GEL to set the PLL freq.  So have not done it.  This would be why the SPI has problems??

Thanks again.

Another thought.  When the reset line is low how does this only bypass the NOR?  I thought reset was to the processor.

Is there some specific way to bypass the NOR memory with one of the jumpers?

There is no jumper in this board design to bypass the NOR flash. The idea to ground the reset pin is to lift only the NOR flash chip's reset pin from the board. This is the function of that NOR flash device. You can refer to it's datasheet.

Thus, you have to be sure that the flash device indeed has junks in it that you cannot erase it. Hope you know what has been done to your board.

Regards.

Now I understand better about the reset line.  Not the master reset but the NOR chip pin.

What I understand you feel its the board is bad.  Should I send it back for repair.  We have had it about 8 months an been using it with the direct CC download.  Now we are moving to the product and must eliminate this step by using the bootloader.

The board has proven operational success with the ADS1299 part.  I need to start the board layout but first must be sure the system works standalone.

What is your recommendation.  If I lift the pin do I loose the warranty? 

Another update.

There is a BIG problem with documentation!

I just discovered that there are two versions of "programmer.out"  I mean one seems to be the general published one and the second is only found on the Specturm Digital web site under the software for the the 5525EVM board.  The latter even has a different name.  programmer_C5515_EVM.out.   Please make note of this when giving help on this specific board.  The order of even the NAND NOR selection is different in this version.

So when I run it there are no errors on writing LED.bin to either NAND or NOR memory.  It allows the Cleaner.bin to work also.

BUT>>  The LED.bin made as you suggested does NOT work in either position NOR or cleaned NOR and in NAND position.

This may have to do with the .bin file itself.  NOR memory is parallel 16 bits.  NAND is Parallel 8 bits. 

How does this get accounted for in the HEX55 build of the .bin file?

Sorry that you may have overlooked my answer stating the programmer in CSL. Yes, the programmer in CSL is intended to cover more than one HW platform thus the name has not specific any HW platform.

The serial8 option in hex55 stands for the boot table interfface format.

hex55 is the hex conversion utility shipped with CCS assembly language package to convert the executable COFF into object file format. C5515 is supported from revision 4.3.5 and newer. C5515 uses only a subset of the tool. The example command to use is:

Hex55 –i file_name.out –o file_name.bin –boot –v5505 –b –serial8

V5505 is the device family and C5515 is the next generation to VC5505 thus uses v5505; serial8 is the boot table interface format.

Regards.

The programmer.out problem is as follows:

When I use the version with CC.  It does not properly write to the NAND or NOR memory on the Spectrum Digitial 5515EVM board.  It may work on the eZ5515 stick but not on the EVM.  The only version of programmer that works (with other short commings) is the one on your web site under board software support for the EVM.  It has a different name.  Something like "programmer_C5515_EVM.out.  It appears to write to NAND and NOR correctly and run the cleaner.bin program.  Just I cannot use it to read back from the memory like the version with CC.  There is no version control on the software only a date made.  They are significantly different.

What I need is to be sure that what I write into the NOR memory is being read out.  Presently I write an ASCII file into the NAND and NOR using your version and then read using the CC version.  I get exactly what I put in on the NAND.  The NOR comes back with the same character repeated.  So it must not be working properly.  Again NOR is 16 bit parallel memory.

The problem is how do I get a .bin file written into the NOR memory (bin is only 8 bit) and the have the boot loader read it out correctly?

Hope you can help.  I have been working with TI on the boot loader.

Dennis,

Thanks for pointing out the issues. Our plan is to have one programmer in CSL which will works for all HW platform. I have been referring to the programmer in CSL 3.03. By the way, is CC refer to same version?

Othere than myself, I also had a co-work testing out the programmer in CSL 3.03 yesterday on both Spectrum Digital TMS320C5515 EVM Rev B and Rev C boards. He did successfully write to NAND, NOR, I2C and SPI with the instructions posted here. However, we have not tried the read. Is that where you problem is?

Regards.

The new version called "programmer_C5515_EVM.out.  Is found on the following link:  http://processors.wiki.ti.com/index.php/C5515_Boot-Image_Programmer 

http://processors.wiki.ti.com/index.php/File:5515evm.zip  Down load site.

Date is 5/27/11 on this one.

---====

 "programmer.out"  It is found on the Spectrum Digital web site link: http://support.spectrumdigital.com/boards/evm5515/revb/   This version is dated 2/3/2011 and is under their demo software. 

The direct link for it is:  http://processors.wiki.ti.com/index.php/File:5515evm.zip

 programmer.out. can also be found from the installation disk coming with the evm board, i.e. from CCStudio_v3.3\boards\evm5505_v1\demo

====--

There seems to be another one referenced at the following link.  I have not tried it.

http://e2e.ti.com/support/dsp/c5000/f/109/p/89749/473800.aspx#473800

====----

Just too much confusion.

XXXXXXXXXXXXXXXXX

OK  The real issue is writing a file to either NOR or NAND and have it be read by the boot loader.

If you have a Spectrum Digital 5515 EVM board what I need is a programmer.out version and a test.bin file that will load into either NAND or pref. NOR (first in the hunt) that you know works.  I don't care what you have for the test.bin file I have tried the led.out which loads from Code composer but have never been successful converting it to a .bin file that will bootload.

I would also need to know the exact commands for the HEX55 you used to take the .out to the .bin.

If you sent these three files to me via email I can use them here.  This will prove that our two boards here have a working boot loader.

Thanks in advance.

Dennis

Attached .bin will blink xf 3 times. programmer.out is the one in CSL 3.03; attached also for your reference. Also attached is a .out file which blinks xf 6 times. You can try converting it.

I will eventually create a wiki page for this and update all pages to link to this.

Hope this help.

5545.xf_blink_3.bin

3443.xf_on.out

7824.programmer.out

HEX55 exact command is copied from above to here: hex55 –i file_name.out –o file_name.bin –boot –v5505 –b –serial8

It did not work.  There is a question about the CSx.   I was told that  CS2 is  NOR  and CS4 is NAND.  What did you use?

The program.out should be using something like  221C:\file.bin   if it is in the root directory.  to program  NOR

===----====

Here was the results for one board

I used 7824.programmer.out to write the 5545.xf_blink_3.bin into both NAND and NOR.  The output stream indicated that it was successful; however, cycling power does not blink the led.

I am not clear about the "CSx" parameter.  I used a value of 4 because that is what the example used.  In any case it stated that it successfully programmed the .bin file to the board.

Am I missing a step?  Here is the output stream:

NAND

-------------------------

Choose the device...

1x - NAND Flash [CSx: 2,3,4,5]

2x - NOR Flash [CSx: 2,3,4,5]

3 - SPI EEPROM (16 bit address)

4 - IIC EEPROM

5 - MMC

6 - SD

7x - SPI Serial Flash (24 bit address) [PinMap x: 1=MODE5,2=MODE6]

8 - MCSPI Serial Flash

14

Operation to perform...

1<file> - Program device with <file>

2<file> - Read device data to output <file> 1c:\dev\c5515flash\5545.xf_blink_3.bin

NAND Flash...

Writing data to NAND...

Writing Sector 0...

Writing Sector 1 (MBR)...

Flushing Data...

Writing Boot Record...

Flushing Data...

Opening c:\dev\c5515flash\5545.xf_blink_3.bin...

Input file opened

Writing Boot-image to NAND...

Programming... [TotalSize=256 (0x100)]

Programming... [TotalSize=512 (0x200)]

Programming... [TotalSize=549 (0x225)]

Flushing Data...

done

Note: I tried cycling power at this point but it did not blink the led.  I then proceeded to write it to NOR

NOR

----------------------------------

Choose the device...

1x - NAND Flash [CSx: 2,3,4,5]

2x - NOR Flash [CSx: 2,3,4,5]

3 - SPI EEPROM (16 bit address)

4 - IIC EEPROM

5 - MMC

6 - SD

7x - SPI Serial Flash (24 bit address) [PinMap x: 1=MODE5,2=MODE6]

8 - MCSPI Serial Flash

24

Operation to perform...

1<file> - Program device with <file>

2<file> - Read device data to output <file> 1c:\dev\c5515flash\5545.xf_blink_3.bin

NOR Flash...

The Flash ID is 89, 8818, d03f

Erasing chip (NOR)...

Writing data to device...

Opening c:\dev\c5515flash\5545.xf_blink_3.bin...

Input file opened

Programming Complete

(Still no action on the led).

Yes. Both NAND and NOR were programmed correctly. Since NOR is the first peripheral to check for valid boot image, it should blink XF.

Ok. Let's confirm a few things. Is this a brand new board out-of-the-box? Is there anything displayed on the LED? Can you load the .out and see XF blinks?

Regards.

OK

We have two boards here.  They have never been modified.  They were purchased about 3 months apart.

When your .out file is run and the 1299 daughter board is removed there is one led under the board (upper right area) that blinks exactly as you indicate.  There is also a second led (for USB connection) That also blinks but very dim.  We unplug the USB connection when testing auto boot.  I am guessing the XF led is the one in the upper right area.

We have not changed jumpers

Here is my list:

JP 2 has a jumper on the upper two pins.

JP 3 has a jmpr on both pins

The WK_PU_PD  is on the upper side of the two pins

JP12 is on the upper two pins

JP6 is jumpered on both pinsVIN SEL is jumpered to the two pins on the left

That is it for jumpers.

Great! We are onto something.

First, revC & revD C5515 EVM should not have any daughter card. What board do you have? I suspect you have an old board.

Second, XF should have been written next to the LED. The upper right should be the one. Could you confirm?

Regards.

We have Rev. C

The daughter card is a 1299 D.A applications board.   I unplugged it to make sure  no interfearance.  I does not help.

XF is the LED in the upper right area near to the line of jumpers starting with JP2 at the top edge of the board.

Hope this makes it clearer.

Both LEDs blink on our boards.  Not the same type of blink.  The XF is clear on off with a long time period.  The other LED on the left side near the USB control chip.  It blinks real dim and about 1 per second.  When the program first runs it is on for a couple seconds..

Dennis

With the xf_blink_3.bin file, XF should blinks 3 times quickly and pauses a bit longer and then 3 blinks again. The .out should blinks 6 times quickly and then pauses and repeat6 blinks again. It you use the cleaner.bin to erase the boot image, pattern gone.

Regards.

I made generalities about the blinks but only made the .OUT version to work.  So saw 6 bilinks as you indicated.  Never saw the 3 blinks as tha is the bin file which never ran in boot loader mode.  The cleaner does erase the previous file.  I believe the cleaner is not needed before each new programming as the programmer erases the NOR first as indicated.

The second LED to the left on the board is not very bright.  VERY dim but it does blink at least in the .OUT file running.

I got Paul's requests for  testing the CLKOUT pin.  Will do it tonight.

Undersdand how to use programmer with no white spaces.  Don't load SARAM31 with anything.  Certialy the LED blink program does not do this and initializes the interrupts correctly as it works on your board.

Please check our jumper config to see if there is a problem with the LDO settings.  The board came this way however they are set.

I am curently snow out in Texas and won't be able to provide jumper setting on my board till Monday.

However, Spectrum Digital ships board with a boot image in SPI which displays on LCD. So default setting should be able to boot xf_blinks_3.bin.

Regards.

OK

Will wait out your snow.

I will look for the default settings on the board and see if there are some differences.  Both boards are the same from the factory as far as I know.  But there is something that is different or they would work.

Talk more on Monday unless you think of something.

Thanks for the help.  I am sure we will make this work.  Just something we are missing.

Dennis

Measured CLKOUT   Here is the approximate numbers.

The clock is 12 us high and has a period of about 28 us.  It starts upon power up or reset.  It lasts 6 ms and then goes back low till another reset.  No other activity.

Hope this along with the jumper settings is a good indication of something good foe a solution.

Thanks

Dennis

Dennis,

What you are seeing is what bootloader is doing to CLKOUT. CLKOUT defaults to ON and bootloader disables it to save power.

As to board's jumper settings: bot my revB & revC boards are the same. From top right down:

JP2: 2-3 RESET; JP5: 2-3 WK_PU_PD; JP6: 1-2JP39: 2-3 5V; and on the top left, JP3 1-2 UART_EN. All of these should not affect booting boot image.

I would recommed:

1) use the cleaner.bin to erase NOR, NAND, I2C and SPI again; confirm XF stays on;

2) try xf_blink_3.bin on all 4 peripherals;use cleaner.bin in between to confirm XF blinks or not blinks;

3) convert the blinks 6 times .out file to .bin and try #2 again.

Sorry for such a long experiments, but I am out of ideas on your case.

Regards.

No problem with "sorry for experiments"  This is how to solve problems this complex'

OK 

I have cleaned and it appears XF is on.   But the CLKOUT still quits at the same 6 ms place.  I do not have the SD card plugged in either.

I will now try to clean the I2C and SPI.

Question  You mention "try all 4 peripherals".  What do you mean.  This must be the CS2, CS3 type connections??

I am not sure which ones are real.  The schematic indicates only one chip select is connected to each memory type.   If you mean the 4 types of memory there is actually 5 including the SD card.

I will wait for this answer and then proceed with programming further.

Dennis

Thank you for understanding.

I meant NOR, NAND, SPI and I2C. The CS instructions are in early entries as: 141 for NAND, 221 for NOR, 311 for SPI and 41 for I2C. I2C requires setting PLL to 12.288MHz. You can do that via GEL file. The GEL that shipped with CSL3.03 has various PLL settings. Once you have included in Target Configurations, it is avaliable via scripts->C5505EVM_Configuration pulldown menu.

Regrads.

How can I check (except for running cleaner.bin) on memory locations to see if they are really erased?

CC has a feature under one of the tabs called memory.  I cannot seem to get it to look at all of the memory locations.  Is there some limitation.  How do you get around it.  It seems the boot loader may have found  a section of memory to run that prevents other programs.  The sequence is NOR and NAND.   IF NOR was really clean then NAND should work.  Perhaps NOR is indicating clean erased but not.

What does your first NOR address on a C rev board have in it when clearn?

Dennis

Dennis,

For this blinks XF example, I always use the cleaner and observes XF to confirm. Since you are using same, should be enough to duplicate/confirm.

The memory browser in code composer has Program/Data/IO. They are closely related to the device. For example, each C5515 peripherals have its only memory space and clock gating feature, clocjk has to be enabled to view.

Regards.

I was able to perform your suggested cleaning and testing.

I did not succeed in making your blink binary work.  I had some very strange results.  Below is not a complete log of what  haapend but it is an indication that the first time through the cleaning process the last one (41c:\cleaner.bin) did not work.  I tried again and had different results.  I believe I enabled the GEL file as that memory erased.  When I finally had run the clean over all four three times there seemed to be a clear on all memory. 

Then I programmed the blink file and no blink.  XF led always on at about the same time as the power LED.

What do you recommend.  Is there any way I could send you one of our boards?

=====---======

Trouble Writing Memory Block at 0x1021 on Page 2 of Length 0x1:

Error 0x80000002/-1143

Fatal Error during: Memory,

The memory at 0x00002044 continually indicated it was 'not ready'

All memory operations currently in progress were aborted in order

to regain control of the processor.

This is considered a catastrophic event, but the debugger should

still be able to access memory and CPU registers.

System state has been altered.  It is strongly advised

that the processor should be reset before resuming execution,

=-=-=-=======================

GEL: Error loading file 'C:\SpectrumDigitalBoard\evm5515_EVM_BSL33_revb\evm5515_v1\gel\evm5515.gel': function 'StartUp()' already defined

Trouble Writing Memory Block at 0x1021 on Page 2 of Length 0x1: Error 0x80000002/-1143 Fatal Error during: Memory,  The memory at 0x00002044 continually indicated it was 'not ready' All memory operations currently in progress were aborted in order to regain control of the processor. This is considered a catastrophic event, but the debugger should  still be able to access memory and CPU registers. System state has been altered.  It is strongly advised that the processor should be reset before resuming execution, 

Power Failure on Target CPU: Error 0xA0001020/-1062 Error during: Execution, Target,  Device driver: Power Loss Detected on Target CPU. This could also result from disconnecting the emulator from the target. It is recommended to RESET EMULATOR.  This will disconnect each  target from the emulator.  The targets should then be power cycled or hard reset followed by an emureset and reconnect to each target. 

Failed to remove the debug state from the target before disconnecting.  There may still be breakpoint opcodes embedded in program memory.  It is recommended that you reset the emulator before you connect and reload your program before you continue debugging.

Trouble Writing Memory Block at 0x1021 on Page 2 of Length 0x1: Error 0x80000002/-1143 Fatal Error during: Memory,  The memory at 0x00002044 continually indicated it was 'not ready' All memory operations currently in progress were aborted in order to regain control of the processor. This is considered a catastrophic event, but the debugger should  still be able to access memory and CPU registers. System state has been altered.  It is strongly advised that the processor should be reset before resuming execution, 

Trouble Writing Memory Block at 0x1009 on Page 2 of Length 0x1: Error 0x80000002/-1143 Fatal Error during: Memory,  The memory at 0x00002014 continually indicated it was 'not ready' All memory operations currently in progress were aborted in order to regain control of the processor. This is considered a catastrophic event, but the debugger should  still be able to access memory and CPU registers. System state has been altered.  It is strongly advised that the processor should be reset before resuming execution, 

GEL: Error loading file 'C:\SpectrumDigitalBoard\5515evm_ProgramBootorClean\5515evm.gel': function 'StartUp()' already defined

=======----------=======\

Tried      41c:\ etc    Worked qluickly

The tried 311c\  etc  failed as follows

Choose the device...

1x - NAND Flash [CSx: 2,3,4,5]

2x - NOR Flash [CSx: 2,3,4,5]

3 - SPI EEPROM (16 bit address)

4 - IIC EEPROM

5 - MMC

6 - SD

7x - SPI Serial Flash (24 bit address) [PinMap x: 1=MODE5,2=MODE6]

8 - MCSPI Serial Flash

311c:\boot\cleaner.bin

SPI EEPROM...

Writing data to device...

Opening 1c:\boot\cleaner.bin...

Input file opened

WRITE ERROR! at 0x1

Wrote 0x0001  Read 0x0000

Total Number of Word Read = 2

=============---==========

Ran all of them from 41 (error again)

Then 311 ran fast and worked

Then 221 ran fast and worked

Then 141 ran fast and worked

Then tried 41 again this time

41c:\boot\cleaner.bin

IIC EEPROM...

Writing data to device...

Opening c:\boot\cleaner.bin...

Input file opened

Programming Complete

So very confusing.   LED is still on.   Disconnected and power down.

Then ran programmer and loaded the blink_3.bin   (renamed from your file 5545.xf_blink_3.bin)

Choose the device...

1x - NAND Flash [CSx: 2,3,4,5]

2x - NOR Flash [CSx: 2,3,4,5]

3 - SPI EEPROM (16 bit address)

4 - IIC EEPROM

5 - MMC

6 - SD

7x - SPI Serial Flash (24 bit address) [PinMap x: 1=MODE5,2=MODE6]

8 - MCSPI Serial Flash

221c:\boot\blink_3.bin

NOR Flash...

The Flash ID is 89, 8818, 3dbc

Erasing chip (NOR)...

Writing data to device...

Opening c:\boot\blink_3.bin...

Input file opened

Programming Complete

====----=====]]]

Then reset

LED does not blink.  Stays on.

How many C5515 EVM you have? If more than one, do they all behavior the same? Is this a brand new board that you can exchange? Do you have a local FAE to help you?

Regards.

We have two boards.  They are about 6+ months old.  They were purchased directly from TI web site.  One is in San Jose CA (this one I could ship so you could test it).  The other is in Washington State.  They both act the same even though purchased about 3 months apart.  They both are rev C.

Paul M was our TI rep of sorts.  He has kept track of the project so we never had a local FAE.

Dennis

Dennis,

Please contact Paul M and loop him into your issue. Let him know my email address: stsang@ti.com so he can contact me via TI internal email to assist you further.

Regards.

I will have Paul Mawson contact you.  

Hi Steve, I have been fighting the same problem for some time now, and I have found some helpful information here.

However, I inherited this project and am not familiar with TI tools, so only 2 questions remain:

How do I make sure "it does not use SARAM31?"

How do I "program the PLL correctly?"

Thanks, Jim

Jim,

Yhanks for the feedback. I have updated the C5515 Boot Image Programmer page with answers to your questions. Please refer to this page:

http://processors.wiki.ti.com/index.php/C5515_Boot-Image_Programmer

Regards.

Steve

Hi Steve, I finally had time to try the recommendations on the updated wiki, but still no luck.

I did a lot more digging on my own and notice some differences between the clock instructions in the wiki and in the DSP System User's Guide, SPRUFPOC, rev 1/2012 section 1.4.3.2.6... This is my C version that I have tested with an XF flashing program:

*((ioport volatile unsigned*)0x1c1f) = 0x0001;

The order matters, plus the register 3 init value should be 0x0806, not 0x8006 as shown in the wiki. Also, I calculate 0x017c for the multiplier to get 50MHz, but I haven't yet verified that...

One strange thing is that I can successfully bootload a little test program, but after trying to bootload my program, I can't even run it from CCS4 anymore. Right now I am running with CCS4 and the program is hanging in a tight loop at 0xfff061, which looks like a bootloader address to me, but it shouldn't even be running...?

Any ideas will be appreciated...

- Jim

Hi Steve,

I just want to let you, and the world, know that I tried booting an old version of this program, and it works.

So I don't know precisely what problem I introduced, but it definitely isn't hardware or "system" stuff.

Thanks, Jim

Jim,

Official documents like User's Guide or Technical Reference Manual above wiki. Wiki can be modified without any records.

I cannot find SPRUFPOC. I looked up C5515 DSP System User's Guide, SPRUFX5D, on page 28 there are some examples on setting PLL when CLK_SEL=L.

The example in wiki provides a 100MHz when CLK_SEL=L. This isetting should provide a 50MHz.

 //  program PLL to 50z with CLK_SEL = 0
   asm(" *port(#0x1c20) = #0x8BE8 "); //PLL Control 1 reg
   asm(" *port(#0x1c21) = #0x8000 "); //PLL Control 2 reg
   asm(" *port(#0x1c22) = #0x0806 "); //PLL Control 3 reg
   asm(" *port(#0x1C23) = #0x0201; //PLL Control 4 reg

Hope this help.