Segger J-Link and Ford KW2000

[g1dtf]

Hi

I am having problems using a segger j-link and J-mem to read memory of the TMS470 in a Ford KW2000.I can connect ok with j-flash and read out the TM470 flash data without any problem. However when I try to connect with j-mem the data is not read and when I hit disconnect the set cycles the power on/off and rests. If i try to read the mem using either mem or savebin CLI commands I get a can't read memory error and the power then cycles/resets as with the disconnect.Very occasionally the unit connects ok with both j-mem and CLI commands reading out the correct memory data. Whilst it is working I also do not get the power cycle/rest problem on disconnect. Can anyone offer any sugestions of what I may be doing wrong. I have connected the j-link to the KW2000 using a 30cm length of ribbon cable using connections posted by Dunker in an earlier thread.

Thanks

Andy.

[Dunker]

I've been having a look tonight.

Make sure your wires are as short as is practical - mine are about 12cm.

Go to Segger website and get the latest software version (4.40b latest but I'm using beta 4.41b).

I use 600khz for initial connect and then 8000khz. I use 600khz with J-Mem.

Sometimes increasing the reset delay can help (CPU tab, init sequence).

J-Mem always has been flakey with Ford TMS. Connect with J-Flash, close J-Flash, immediately connect with J-Mem.

To get the stub data using J-Flash (obviously load the TMS470-288 project file first - I use this for both 288 ans 689), connect then Target / Read back / Range... / then enter 3FB0 and 3FFF. Then save data.

Obviously keep eye on radio display - if its not lit up to start with CPU might be asleep.

[g1dtf]

Thanks I will try shortening the cable and using your settings. I forget to mention in my previous post when the set resets on disconnect (when not working/reading) the display does go off before the set restarts up. Will update the thread on results both positive or negative.

Andy.

[g1dtf]

I have redone the cabling shorter and tried your settings. Things seems to be working a lot better now and getting j-mem connecting ok nearly every time. Hopefully its not just a flash ib the pan and my problems have been sorted.

Many Thanks Mate

Andy.

[Dunker]

Good to hear. It will never be perfect, I dont think it will ever be unless its on a development board - but then I doubt it. Did you ever get that radio unlocked?

[g1dtf]

It was one with a 689 MCU and bodged it with a no code hack by altering some code in the EEprom by trial and error.

Just need to handle a few more sets now rather than just this one to confirm everything is working ok but looking good so far.

Andy.

[Dunker]

If you still have the set and can be arsed, post the stub data and eeprom data so I can generate the reset eeprom data (with a code of your choice if you want). The more data we have the better.

Either way, no probs.

[g1dtf]

I dont have the set any more but I can send u the stub and eeprom data from this 288 if u want it.

[Dunker]

Thanks but I've got all the data I need. I've got the encryption / decryption sorted.

[alexics]

Can you run the JTAG without the TRST? I mean using the long winded reset or will it fail?

[g1dtf]

Hi

Has anyone got any info or reading the TMS470R1VF48B MCU in a Sony Ford CD3XX-CDI-ISLAND-KW2000 with the Segger J-Link.

Thanks

Andy.

[Dunker]

I've not looked but if its not in the sample project files have you tried just using the 288 project. I would have thought that there's not a lot of difference between them - memory size etc. The datasheet will tell you - then you can edit an existing project and save to a new project file to suit.

[mobilart]

Hi

Has anyone got any info or reading the TMS470R1VF48B MCU in a Sony Ford CD3XX-CDI-ISLAND-KW2000 with the Segger J-Link.

Thanks

Andy.
Use Project TMS470R1B1M.jflash

[g1dtf]

Thanks guys. Will give it a try next time I get hold of one and actually have time to play with it.

[g1dtf]

Its been a while since my last post but I am still having problems getting J-flash to connect with a 689 CPU. I have a 288 on the bench which connects fine with j-flash and J-mem also works ok to read the code out. However I have a 689 which when I try to connect j-flash comes back with the error " could not read target memory failed to connect". J-mem also does not connect properly to read out the code data. I am using j-flash 4.40b with the 288 project file.

Thanks

Andy.

[Dunker]

It's a strange one. The odd radio does give problems but overall they are readable. I'm only in Castleford so if you are ever coming up this way you could call in and compare what happens with your J-Link and mine. Let me know.

[g1dtf]

I have now updated to latest verion 4.56c which in turn offered a firmware update which I applied.. I can now connect to the 689 with J-flash and read the stub data etc. However I am still unable to get j-mem to read the ram to display the code. I have checked with a 288 and that is still working ok with the new software. Ufortunately I have only this one 688 to play with so cant comment to to if its just a problem set. J-mem shows as connected but when viewing the data I have no CRADLE PHONE etc around 102000 area.

Thanks

Andy.

[g1dtf]

Just realised that this 689 is a 6006 (6CD) so the ram contents are different (no CRADLE PHONE). I have however hopefully found the code on the stack but in a totaly different location (around 1x100BE0) and has the FFFF above and 04 just after.

Andy.

[Dunker]

Just realised that this 689 is a 6006 (6CD) so the ram contents are different (no CRADLE PHONE). I have however hopefully found the code on the stack but in a totaly different location (around 1x100BE0) and has the FFFF above and 04 just after.

Andy.

That should be it, I have 100BE0 in my notes. As a foot note, thinking back, I'm sure I had a lot more reliability with an earlier firmware version than the one I'm on now although like I said, I don't have too much trouble now. I'll try the new version you have when I get a chance.

I was also wondering if anyone has the J-Link china clone? Wouldn't mind trying something with flash write capability.