Just bought dev board, debugger won't run

pasty · Joined: 30 Mar 2016 Posts: 6

I just bought the 24EP256GP206 dev board package.
https://www.ccsinfo.com/product_info.php?products_id=PIC24H-kit
I can compile and run the exercise programs from the included booklet.
But the debugger won't run.
I called tech support, they didn't have an answer.

Any ideas?

Thanks,
Bob

Ttelmah · Joined: 11 Mar 2010 Posts: 19447

What development environment are you using?.
MPLAB or CCS IDE.

Assuming CCS.

First, have you selected 'debug', 'configure', and selected the right debugger?.

Have you told your code to debug?.
add:

#DEVICE ICD=TRUE

in the code (don't actually need to do this, but it avoids having to manually say 'yes' each time).

Then click 'enable debugger'.

What do you see at each stage?. Is the debugger listed?. Where does it fail?.

pasty · Joined: 30 Mar 2016 Posts: 6

Hi Telmah,
Thanks for your reply,

I am running the CCS IDE
The correct debugger is selected (ICD-USB). I am using the ICD-U64 (just bought it a few days ago).

I am aware of the #DEVICE ICD=TRUE

The debugger says "attaching", then "loading", then "busy", then a dialog opens saying "error - target not running"

I have noticed that there seems to be two way things are compiling.
1. If I select "compile" while the debugger window is open, then I get a compiler warning that says " warning - Bootloader disabled" and the target will not run. RAM% shows zero, ROM% shows zero after compile

2. If I select "compile" when the debugger window is closed, then select "build and run", the program runs. RAM% shows zero, ROM% shows 3%.
After hitting "build and run", I see "loading executive" then it loads a hex file then it runs OK.

In case 1, the target does not run.
In case 2, the target does run.

Something different seems to be happening in the compilation and with a bootloader in the two cases.

I am not finding any information about the bootloader. What is this doing exactly? Is this related to the problem?

Thanks,
Bob

pasty · Joined: 30 Mar 2016 Posts: 6

Apparently, the bootloader is a UART bootloader that is part of the devkit_24ep256gp206.h include file.

I went back to the first exercise program in the example booklet that does not use the above include file, it uses the 24ep256gp106.h include file (no bootloader included).

The debugger still will not start and still says "error - target not running"
So, the bootloader is not related to my issue.

I just got a message back from tech support saying they are working on the issue. They can't get the debugger to run on the 24EP dev board either.

I will report their findings when they are available,
Bob

Ttelmah · Joined: 11 Mar 2010 Posts: 19447

As a completely 'off the cuff' thing to try. Try setting up your program for the simplest possible clock mode. So FRC, at 7.37MHz. See if this then works.

There is a problem with all the chips in this family, that they can't wake up if PLL is being used, using an external oscillator that is outside the range of 3Mhz to 5.5Mhz. If this is the case, you have to wake using FRC, and only once started switch the divisors and enable the PLL. Now DEBUG in many cases behaves differently to 'real', with regard to clock switching, so might be giving problems. If the oscillator wasn't starting, then you would get the device not running error.

newguy · Joined: 24 Jun 2004 Posts: 1906

If Ttelmah is right, here's code that should get you up & running reliably. I had a similar issue but with a different chip.

http://www.ccsinfo.com/forum/viewtopic.php?p=159723&highlight=pll+internal#159723

Ttelmah · Joined: 11 Mar 2010 Posts: 19447

I'd test first with just FRC. The point is that what Newguy posts is a way of getting the chip working if it doesn't start. I'm looking at the possibility that the debugger can't handle the startup instead.

pasty · Joined: 30 Mar 2016 Posts: 6

Hi Telmah,
Thank you for the suggestion.
I am new to these processors and to programming micros in C.

What would the code be to try this?

Thanks,
Bob

Ttelmah · Joined: 11 Mar 2010 Posts: 19447

Something like:

pasty · Joined: 30 Mar 2016 Posts: 6

Hi Telmah,
I tried running the test code you sent.
I still get he same error "error - target programming not running" after launching the debugger.

Bob

Ttelmah · Joined: 11 Mar 2010 Posts: 19447

It was worth trying.

I was going through the list of things that would make the 'target not run'.

Power. However if CCS have tried as well, they should know how to power the board.
Clock.
Connections.

The only one of these that was in your control, was the clock. Hence worth trying....

If I had one of these in front of me, I'd sit down and play. Probably stick a logic analyser on the ICD connections and try to work out what is happening. Unfortunately looks as if you will have to wait and see what CCS say... :(

ljblinds · Joined: 13 May 2016 Posts: 2

I had to put the following into my code for a 24FJ128GA106....

ljblinds · Joined: 13 May 2016 Posts: 2

Open the Device Editor for your device. Expand Tools. You will see there are 3 sets of pins you can connect the ICD to on the "ICSP Pins" line. You need to tell the compiler which pins you are using. Probably the 2nd set, B6/B7 so I THINK that's why I had to put a 2 in there.

It's been several years and I'm just getting back into C programming for PICs so correct me if I'm wrong here. Question