Many Small Issues Adding Up

[D-Man]

Hi,
I'm experiencing several issues that I'm starting to wonder if they are all related somehow.
I've been searching the forum and haven't found topics that have lead me down a path to solve them. Tried lots of suggestions, though.
I’m using the internal osc for the first time on a project and can’t help to think this is my issue.

Compiler 4.119
PIC16F1937 44-pin TQFP

1) Debug doesn’t work. Getting the “Could not start target: target was halted after reset Check…” Done all of the checks. It’s working any other time, just not for the debug.

2) Can’t get pin A7 (OSC1) to work as an output. I’m using pins A6 and A7 and the internal osc because I’m using just about every pin on the chip. To get around A7 not working I jumpered it to E1 and was able to continue writing code. Yes, right now E1 is electrically tied to A7. It will be challenging to cut the trace so I haven’t done it yet. Pin A6 (OSC2) is working fine as an output. A7 isn’t completely dead, as I can get a very poor square wave out of it, maybe 1vp-p and not very smooth.

3) Have to cycle the power after each time loading the program. Is this normal??? All of my previous projects I used an external osc and never had to cycle the power after each load of the program. Of course, debug worked then too.

4) #int_timer1 with a function that follows appears to make the device unresponsive. This is the strangest and the icing on the cake. I have never had a problem with interrupts, especially timers. If I comment out or remove the “#INT_TIMER1”, leaving the function, everything starts working again, except for the interrupt.

Any ideas/suggestions?

I don’t feel it’s the code, as I’m using pieces of known working code from previous projects.

I know my compiler is out of date, could this be my issue? I am able to compile without errors and what I do have written, works. Although I have noticed some strange behaviors that could be related to pins A6 & A7.

Thanks for your help in advance,

Dave

[asmboy]

[PCM programmer]

[D-Man]

asmboy,
I hear ya. I struggled with this post, but came to the conclusion that it must be program related and not hardware related and I needed help.

PCM programmer,
I'm using the ICD-U64. Also new for me on this project is the Tag-Connect. Not sure why that would matter but thought I'd mention it.

I had found/read your post on the #int_timer1. Good point, in my haste to cut and paste, I didn't put any effort into the verifying the timing. I assumed it would be "good enough" to start and I'd get the exact timing down later. I believe T1 divide by 4 or 8 and running 8MHz INTRC_IO. At my day job right now, will check when I get home.

Also read the part about A6 & A7. Initially I didn't do anything. I let the compiler do its job. After reading the post I tried the fixed_IO, with still no luck.

A little background, I've been using CCS compilers since 2004. Always had good luck with reading posts and solving my problems on my own. Although I don't do programming at my day job any more, I still give it a try at home.
I feel skill level is probably an advanced-beginner to intermediate.
This is my first post.

It could be a few days before I can get back to this and I'll put some code up. Please stand by...

Thanks,

Dave

[Ttelmah]

The critical point was INTRC_IO.

Fixed IO only applied because the poster wanted to perform byte wide output on a port, while keeping B0/B1 as input.

Best Wishes

[D-Man]

Ok, got these problems fixed.
1, 2, & 3 were fixed by removing the ECH fuse. I used the PIC Wizard thinking it would help speed up the start of my project. I also cleaned out a few other fuses that I decided weren’t needed.
Anyone have any comments on why the PIC Wizard would insert fuses that shouldn’t apply?
Here’s the original. I commented out the fuses for this post and deleted them in my program. Take note of the NOBROWNOUT fuse. Then near the bottom it sets the BORV19, I changed that too.

[Ttelmah]

Key word throughout, is 'think'.

The Wizard does not 'think'.
You need to 'think'.

The Wizard will set basically all fuses, even if you are selecting others that make these 'pointless'. That doesn't actually matter at all, except in terms of wasted lines. The Wizard can setup the fuses perfectly sensibly, but _you_ need to do the thought when using it.
There are a couple of faults it perpetually has. One classic is 'setup_spi(SPI_SS_DISABLED);', where the tickbox to 'disable slave select' sets this, where the user probably actually wants to disable the SPI, not the slave select (setup_SPI(FALSE);). Generally, using the Wizard is actually likely to cost time, rather than save it, once you have any experience with the compiler at all, while for a new user, things like the above, can be very dangerous.

It is an 'accident waiting to happen', unless used with care.

Best Wishes