This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

Can replace a new MCU at execution board?

Date : 28/02/2008

To Keil Suport and LPC2103/MCB2103 user:

Im a user that use the MCU of LPC2103, test with MCB2103 Evaluation Board, developed with Keil Software, communicate with ULINK USB-JTAG Adapter. For now occur that "JTAG Communication Failure" matter.

Confirm that MCB2103 have power from PC USB Port, jumper J9 for JTAG comfirm have on, J1 for ISP have off.

Besides it, I found that when I open the window that Project - Options - Debug - ULINK ARM Debugger Settings, also cannot automatic detect the LPC2103 device. Just appear that "JTAG Communication Failure" in JTAG Device Chain windonw.

I already follow the FAQ from dear compamy website (http://www.keil.com/support/docs/2935.htm), follow the instrution that

=====================================================
-- The power supply of the ULINK driver might be incorrect. Open ULINK and set the jumper VCC to 3.3V. For more information refer to the Getting Started User's Guide, JTAG Debugging - ULINK: Interface Schematic.

->I try already, fail..

=====================================================

-- The ARM device may execute some code that interrupts communication with the Embedded ICE. This happens when the device executes random code. Some devices have also problems with the JTAG interface, when the device goes into IDLE mode. To recover such devices, you may erase the complete program using serial program utilities.

-> I try use with the Philips LPC-2000 Flash Utility v2.2.3, and change the jumper J1(ISP) to on, jumper J9(JTAG) to off, with the RS-232 extension cable that I do, (1-1, 2(PC RX)-2(MCB2103 232,TX), 3(PC TX)-3(MCB2103 232, RX), 4-4, 5-5, 6-6, 7-7, 8-8, 9-9 ) connect to MCB2103 UART-0 comport with 9600BPS, 8, N ,1. The result that also "Cannot communicate with test board!" after push the Read Device ID.

=====================================================

-- When ARM devices work with a slow clock (below 6MHz) the ULINK may be too fast by default. Reduce the MAX JTAG Clock in the dialog Project - Options - Debug - ULINK ARM Debugger Settings.

-> All the clock speed i try already, also popup the window that "JTAG Communication Failure" when i want download the program to Flash.

=====================================================

Actually this problem occured after i make a mistake in experimen. I thought that the battery pin (use for tight the battery long pin on pcb board, I have not use the battery, so I use the plastic seal the battery without contact with this pin)is ground, so I use the osciloscope get this pin as ground and try get other CPU pin signal...

So if i have the ability to solder out this one LPC 2103 CPU from the MCB 2103 execution board and solder a new LPC 2103 CPU, is it ok??

Parents
  • Did the oscilloscope have any other ground reference besides this pin, i.e. did the oscilloscope short this pin ground or to any other potential?

    I normally work in an environment where the test platform and the oscilloscope are galvanically isolated from each other, so if I connect a ground cable from the scope to the target, no current will flow. This is required to be able to measure with any signal as reference, without the need for using the A-B mode of the scope.

    Have you checked if the crystal on the board oscillates? Of course this test doesn't mean anything if the processor defaults to an internal RC oscillator until a program specifically reconfigures it to use an external crystal.

    Are there any I/O pins on the chip that should switch to specific logic states when the processor is turned on, and before any program changes them to outputs and forces them to high or low state? Are there any indications that the processor do power up?

    Note that unless you have decent equipment and a bit of experience, it is far cheaper/quicker to buy a new board than to spend the time and replace the chip. I assume that it is a LQFP48 chip on the board, since it has a VBAT signal.

Reply
  • Did the oscilloscope have any other ground reference besides this pin, i.e. did the oscilloscope short this pin ground or to any other potential?

    I normally work in an environment where the test platform and the oscilloscope are galvanically isolated from each other, so if I connect a ground cable from the scope to the target, no current will flow. This is required to be able to measure with any signal as reference, without the need for using the A-B mode of the scope.

    Have you checked if the crystal on the board oscillates? Of course this test doesn't mean anything if the processor defaults to an internal RC oscillator until a program specifically reconfigures it to use an external crystal.

    Are there any I/O pins on the chip that should switch to specific logic states when the processor is turned on, and before any program changes them to outputs and forces them to high or low state? Are there any indications that the processor do power up?

    Note that unless you have decent equipment and a bit of experience, it is far cheaper/quicker to buy a new board than to spend the time and replace the chip. I assume that it is a LQFP48 chip on the board, since it has a VBAT signal.

Children