Trouble with the PIC16F688

jreinhart · Joined: 07 Jul 2009 Posts: 3

I'm having trouble getting this device to run the way I want.
My first issue is the clock speed doesn't seem to be setting up correctly. That is, I'm attempting to use the internal 8MHz clock. When I toggle a pin in the main loop, it toggles at only 284kHz.

My second issue is I can't seem to debug with this chip. I get:

"Could not start target: The target was not halted after reset. Check the target oscillator and MCLR."

I'm using the same ICD connections I've used for several PIC designs. No cap on MCLR line and 47k pullup. I've used the diagnostics tool in ccsload to make sure voltage levels on the ICD lines are correct and everything checks out.

Here the code I'm trying to run:

FvM · Joined: 27 Aug 2008 Posts: 2337 Location: Germany

PIC16F688 doesn't provide built-in debugging capabilities. An ICD2 debugging header (Microchip Part# AC162052) is required.

I don't understand the observed timing problems. I'm using PIC16F688 with INTRC oscillator, and the timing is correct, including UART baud rates. You possibly need to set OSCON accordingly.

Guest · Posted: Tue Jul 07, 2009 7:54 am

Look at the .LST file.

Comment out the other feature setup lines of code.
You are not using fast_io so there can be TRIS code around the toggle
statement that takes more cycles - plus ICD overhead.

So the freq you see may be just about right - given the situation.
Evaluate the .LST file to see for yourself.

jreinhart · Joined: 07 Jul 2009 Posts: 3

It's unfortunate that I can't debug with this chip. I wasn't aware. This will make it tricky to work with and test the A/D.

Well, based on other posts I've seen, I had already tried manually setting the osccon register and that didn't change anything.

I should also add that I've tried setup_oscillator(OSC_4MHZ) to see if I could even affect the clock. The toggle speed remained constant at 284kHz.

I looked at the .lst file and the toggle loop is about 4 instructions long. The fact that the observed toggle rate is so much slower than expected, (like 28 times slower), plus the fact that I can't seem to affect the rate with setup_oscillator() leads me believe the clock isn't being setup properly.

I'm stuck.

Ttelmah · Guest

Worth remembering though, that 'output_toggle', takes several instructions. The tris bit for the requirted bit is cleared, then the latch is read, bit modified, finally the new bit is output. It will be longer than the four instructions you quote.

jreinhart · Joined: 07 Jul 2009 Posts: 3

Ttelmah,

Thanks, you're right. I was thrown off by the fact that I didn't seem able to affect the system clock with setup_oscillator(). I was able to slow it down by manually setting osccon. Seems to be some problem with setup_oscillator for this chip and my compiler version. So, short of not being able to debug, it seems things are working. Thanks again.