Hello everybody!!! I'm using a phyCore XC167 with AC step controller and a ULINK for debugging. I use the internal ROM by jumpering EA. For over a week I could download my sw direkt with my ULINK. Debugging does not work that good. I had some unpredictiable jumps to adress 0xfffff. But the main problem is, that my ULINK now cancel download with the error message "device can not enter bypass mode. Connection to target missing or incorrect". First I tought, i have killed my controller but then I've tested it on the orginal Phytec RDK board, and it worked fine (same software)!!! My second module works on both boards but it is configured with the default values from Phytec (external ROM...). Could any internal or hardware config cause this problem??? It seems, that my controller does not start running on my target hardware because the module sinks nearby no current. I even could not flash in bootstrap mode!!! Please let me know, if anyone could help me with that Problem. Thanks!!! Martin
There are some critical settings that you should verify. Please check the information on: http://www.keil.com/support/docs/2667.htm
I know this support document. But I use a module from Phytec, and as I know they even develop for Infineon. So I think this is not the problem. Yesterday late in the evening I found the wrong config in my Keil. I defined a RAM area where I have some IO extenions and not realy a RAM chip. I think the linker placed some vars there. But I could only flash the new file when I placed the module on the RDK board from Phytec. After that all works fine even on my target hardware. Wich difference of the boards could cause such a problem. On an external start (EA = 0) I didn't have any problems with the exactly same software. And the problem with the jumps to adress 0xfffff while debugging are still there. Thanks for helping!!! Martin
I also see jumps to address 0xfffff using ULINK with the XC164CS. Makes debugging very difficult. Can anyone explain why this happens? -Walt
We have seen this problems when the CPU is incorrectly configured. It looks like you have still a small configuration issue. The behaviour that you see exists if you have mis-selected the BOOT configuration. Please check the EA pin and other startup options on your board.
We are using:
EA = 1 ALE = 0 RD = 1 WR = 1
The error message "device can not enter bypass mode. Connection to target missing or incorrect" can also occur, if the pulldown on TRST in the target circuit is too low. uLink only has a 10k Pullup (no push-pull drivers) to place a High-Level on TRST, so the external pulldown should be 100k. I also have the problem with the jumps to 0xfffff. Startup configuration is also EA = 1, ALE = 0, RD = 1, WR = 1. Mostly it occurs when I use singlestepping. If I then stop debug mode and start it again, it works normally. How can this problem be solved? Regards, Roland