Micro won't boot correctly after programming

SteveM · Sep 17th 2008, 9:16pm

Hello,

I'm using two SEGGER J-link units in a test fixture in our manufacturing plant to program NXP2134 devices on seperate PCBs as part of our manufacturing process.
We've been trying to upgrade the J-Link/J-Flash software from V3.80a to V3.90d, and after we did this, can't get the device to communicate with our test software after release the reset on our micro.
I'm able to switch between the old and new versions, and both report erasing, programming and verifying the device, but when we allow the product to boot afterward, we get no response after programming with the newer version unless I remove and then reapply power.

Is there a problem that I'm discovering with the software, or something else I should be aware of?

thanks in advance,
Steve

SEGGER - Alex · Sep 23rd 2008, 5:01pm

Hello SteveM,

We have tested this:
1. Loading a Program into the target flash via J-Flash and select "Start application"
--> Program runs as expected
2. After downloading the Application via J-Flash, start J-Link commander, pull reset to low, release reset
--> Program runs as expected

Testing environment:
J-Flash/J-Link commander V3.90d
J-Flash project: LPC2138.jflash
IAR LPC2138 Eval board (Chip: LPC2138FBD64), should make no difference to LPC2134
J-Link hardware version V6

So we do not see any problems.

Best regards,
Alex

SteveM · Sep 24th 2008, 7:18pm

Alex,
Thanks for your reply. Here's my setup:
I have an ST-2134 on a board that I test as part of the manufacturing process.
The main microprocessor is held reset after the board is powered up and I make DC measurements.
The Reset line is toggled low then high using non-Segger hardware. The JLink RESET line isn't wired to to 2134 RESET Line.
I call the JFlash exe using the 2138 job, and load our .bin file
the job executes using the -auto, exit parameters
the JFlash program reports that the micro is programmed and verified, and I can see this when I check manually
When the JFlash completes successfully, my test sequence releases the Main Microprocessor's reset, and allow the board to boot.
I never go into JLink to set the ARM reset high, because it is already high after being toggled.

This has worked for the past year or so, and now it's not, with V3.90d. Using V3.90, the 2134 progams correctly, but it doesn't appear to boot afterward.
I can switch back to the V3.80a software that I've been using till now, and have the program and boot.

Does the JLink need to control the RESET Line directly?
If I do, do I need to go into JLink to set the RESET Line high afterward?
Can I toggle my external RESET line after programming and have the 2134 boot properly?

Steve

SEGGER - Alex · Sep 26th 2008, 12:16pm

Hello Steve,

I think you mean NXP-2134 (you wrote ST-2134, but as far as I know such a chip does not exist) in you last post.

We have analyzed this and we have found the problem:
In J-Flash V3.90d the breakpoints, set during J-Flash session, are not cleared and the CPU is not restarted after
closing J-Flash and the DLL (via the "-exit" command). When now a reset is performed (while the CPU is halted),
the CPU will not start properly after the RESET pin is released (this is a bug on some ARM7TDMI-S cores).

This problem will be fixed in the next beta version of the J-Link software (V3.91n) which can be downloaded from our website:
segger.com/download_jlink_beta.html

Best regards,
Alex

SteveM · Sep 26th 2008, 1:37pm

Alex,

Thanks for the reply.
I must have had ST on the brain the other day.
I'll look at the latest version and try it out.
Again, thanks.
Steve

Share