UART problem in dsPIC30F4011 ?

small_chick · Joined: 17 Jul 2012 Posts: 53

I'd like to send only one character into Hyperterminal at each time writing into U2TXREG. But when I used the following code, it gave me 8 similar characters. I can't explain why this happened like that. I hope someone can help me ! i use a hardware crystal 10MHz.

temtronic · Posted: Fri Nov 23, 2012 2:32 pm

Generally speaking...manually setting up registers is not a 'good idea'.
99% of 'help' here uses the C compiler functions that CCS has designed/tested and supplies. They are known to work.

I, for one, have no idea if your assignments/values are correct and there probably aren't too many using the 30F series either.

Honestly, I'd try one of the examples CCS supplies for 'serial' communications, again, they are known to work and would be a good starting point/

hth
jay

small_chick · Joined: 17 Jul 2012 Posts: 53

i think this problem is not easy to explain ! i've tried to use supported function by CCS and realize this interesting thing.
Compare two kind of coding following:
*1st case:

gpsmikey · Joined: 16 Nov 2010 Posts: 588 Location: Kirkland, WA

Seems very odd - this is actual hardware you are running on and that is the entire code you show in your example ? What version of the compiler are you running ?

A couple of things come to mind to try assuming the above is true:
in the end "while" loop, add code to the loop at the end and see if the compiler does it differently ( while(1) {x=1;} ) or something like that that does nothing, but force the compiler to deal with it differently. You also should check the erata sheet for that chip and see if there are any known issues. I certainly don't see anything in your code that could cause it to go around the code multiple times (which is what it looks like it is doing). It almost sounds like (from your example) a flag bit in the uart is not getting cleared properly. I think I better go make coffee ... brain is fuzzy.

mikey
_________________
mikey
-- you can't have too many gadgets or too much disk space !
old engineering saying: 1+1 = 3 for sufficiently large values of 1 or small values of 3

Ttelmah · Joined: 11 Mar 2010 Posts: 19539

I'd say, either the watchdog is getting enabled, and the chip is therefore restarting, or you have a glitch on the power supply, and the chip is doing a brownout restart. Either would explain what you are seeing.
The brownout in particular, does have a wide range in the data sheet. If I remember correctly BORV42, can actually trigger as high as 4.7v on some chips. I had a batch a while ago, that triggered well above the rated figure....

Best Wishes

small_chick · Joined: 17 Jul 2012 Posts: 53

Ttelmah · Joined: 11 Mar 2010 Posts: 19539

Just because it is disabled in the code you post, doesn't mean it is disabled in the chip. Depending on what programmer you are using, the fuse might not be being erased, or might be being set by a 'setup' option somewhere. The only way to be 'sure' is to read the fuses back from the chip, and verify the corresponding bit.
I really would suspect you have got spikes on your power rail though. The processor _must_ have decoupling with really good high frequency behaviour (this means something like ceramic, or polyester capacitors, _not_ electrolytic), right close to the chip. I'd be suspicious you are actually getting tiny spikes as things happen...

Best Wishes