[OT] Looking for advice on debugging the JTAG on LPC2129

Development Boards MicroMod
1

Colleagues,

This is an off-topic question, however I know there hardware designers in this group, who use LPC2000 chips.

I’ve built a board (link to the schematic below), found a ULINK2 debugger and downloaded an evaluation version of uVision. As the next step, I’m trying to verify that the JTAG interface works. I’m trying to download the “Blinky” project that comes with uVision. I do realize that the code itself might not work properly, because it’s written for a different chip. But if I could download it, that would show that the JTAG works.

Unfortunately, I’m getting error messages, when I try to communicate to the LPC2129 through JTAG. Here’s what’s happening:

  1. When the ULINK2 is connected to the laptop and to the powered target board, the LEDs on the ULINK2 look like this: USB - solid red all the time, COM – off, RUN – off.

  2. When I try to download, I get the error messages: 1st message box “JTAG communication failure”. COM LED is solid blue.

  3. The moment I click OK on that one, the other one comes up “Error: flash download failed – target DLL has been canceled”. COM LED turns off.

Could anyone give me some tips for debugging this? Could anyone point me to a tutorial?

If anyone is curious to look at my schematic, here’s a link to it http://www.prolifictec.com/filehost/arm … _rev01.pdf pages 1 and 2 show the JTAG connector and the LPC2119. R1 – R8 and R11 are loaded. J7 is installed . The clock is clocking. At the moment, the majority of the remaining circuit is not populated (except for the necessary components providing core and I/O power).

Any insight would be appreciated!

Thanks,

  • Nick
2

The JTAG connections on the schematic look OK, except that 100k might be too high for the resistors.

Make sure that the leads on the LPC2129 are properly soldered, especially the ones used by the JTAG I/F. They can sometimes look as if they are OK, but aren’t in fact making contact.

Leon

3

I’ve changed the resistors to 10k (like Olimex), but still get the same errors.

4

UPDATE:

Two (2) significant bugs were found:

  1. +1.8V and +3.3V are swapped with respect to the pins of the chip. Brain fart. Embarrassing.

  2. The clock should have the core levels (0 to +1.8V), not the I/O levels (0 to +3.3V).

An updated schematic is here: http://www.prolifictec.com/filehost/arm … _rev01.pdf.