"Reset vector not initialized"

[CBB]

If I run my code in the chip simulator, it gives me a warning of "Reset vector not initalized". In my code, I have the reset vector set to the start of the code, and If I look at the memory on the microcontroller, FFFE-FFFF is set to the start of the code.
Is there some possible conflict somewhere that I should look at?
Is this also the reason that the code will execute correctly when a debugger is being used, but not when there is no debugger present?

[CBB]

The code will run if I have a debugger attached to it. As in run perfectly fine and normal and exactly what I would see if I was using the software emulator.

If I do not have the debugger connected, the program runs very oddly (outputs incorrect values, doesnt respond properly to input, or possibly is comletely unresponsive...).

So, It is a case of everything works with a debugger, and nothing works without it.

[JimDon]

Well, it is hard to respond to that.
What mcu  part? What version of CW (not the ide do 'about' then 'installed products') ?

If you create a default project does that run?

What output are you referring to?

[CBB]

CodeWarrior Dev. Studio V6.2 (is this what you wanted?)

Yes. A default project will (or at least seems to) run, but it gives the same error message...

There is a SCI interface that we are using to output variables from memory. These variables end up being garbage, but the same garbage each time.

[JimDon]

---

CBB wrote:
CodeWarrior Dev. Studio V6.2 (is this what you wanted?)
Yes.

Yes. A default project will (or at least seems to) run, but it gives the same error message...
You mean in the simulator? If so, again, the simulator has issues.

There is a SCI interface that we are using to output variables from memory. These variables end up being garbage, but the same garbage each time.
So it sounds like a clock issue. What are you using for the clock? Which mcu chip is it? Is it a demo board or one of your own design? Did you use PE or write the code or sample code?

---

[CBB]

No, this is the debugger that interfaces with the chip (BDM), not the simulator. But suprisingly, it's somehow being helpful this time.

And sorry, i should have included the information on this earlier. It is a MC9S08DZ60 (32 pin) processor attached to the EVB9S08DZ60 evaluation board (which doesnt support 32 pins, so the reset and BKG pins are connected via wires, and a seperate power supply is being used, and nothing else is connected).

The clock is whatever is internal to the processor..

I took the settings for the SCI from the SCI example code for the evaluation board, but changed the baud rate.

[CBB]

I probably should include that the rest of the settings were made by PE.

[CBB]

It turn out that the problem seems to be in our power up circuit. The code runs as expected when the reset pin is used, but fails to run when the micro is powered off and then back on.

Does any one know where I should look for info on the power-up requirements for the micro? (filtering etc.)

Thank you for you help.

Chris B.

[JimDon]

Well, the data sheet is a good place to start. You can find it here.

This part does not require an external reset circuit, but since you have done some hacking, you might check the Reccomended Connections on page 30 and make sure things are properly connected and that your wires are connected as you think they are supposed to be. I say that from experience.

Also, the power suuply may not be fully shutting of so you might try leaving it off with a volt meter on it to be sure it has truly powered down. It seems after using the BDM to program, this is sometimes required.

[peg]

Hello Chris,

The first thing I would check for in this regard is that you have properly setup the LVI so it can hold off operation until power is good and stop it when it goes bad (off).

[edit] actually called LVD in this device.



Message Edited by peg on 2008-08-16 09:32 AM

[JimDon]

---

CBB wrote:
If I run my code in the chip simulator, it gives me a warning of "Reset vector not initalized". In my code, I have the reset vector set to the start of the code, and If I look at the memory on the microcontroller, FFFE-FFFF is set to the start of the code.
Is there some possible conflict somewhere that I should look at?
The simulator has issues.

Is this also the reason that the code will execute correctly when a debugger is being used, but not when there is no debugger present?
No. The simulator is not the same as the debugger. Usually timing - in the debugger things run with different timing.

---

 
What is it that works in the debugger but not without the debugger?