Comments inline.
--
Rick Altherr
"He said he hadn't had a byte in three days. I had a short, so I split
it with him."
On Nov 20, 2007, at 6:28 AM, "John" <[EMAIL PROTECTED]> wrote:
Many thanks for all the replies.
My project is rolling along again for the moment.
I went to a night class in C programming a few years ago
but seam to have forgotten a few things.
I am reading out values to a computer screen via serial link
in the manner of largedemo by replacing "new" with whatever
I want to read out. Even with error declared as unsigned uint16_t
I had the same positive and negative readout as with error declared
as int_16t. I will have to think why.
If you are using printf(), the format string includes a description of
the type. I would bet that you are using %d which means signed integer
of size int. %u means unsigned integer of size int. I would need to
check what int16_t is on avr, but if it is not int (could be short for
example) then you would also need a length modifier. I suggest
reading the printf() manpage.
As for the volatile declaration I just used the idea from largedemo.
Some of the variables there are declared as volatile. Looking up
the meaning of volatile in C programming I am given :-
"not stored in machine register"
but where it is stored and what it means in terms of avr programming
I really dont know.
It means exactly what it says. The value should not be cached in
registers when it is used for multiple instructions. Normally a C
compiler will optimize a sequence of instructions to leave a
particular variable on a register since accessing a register is faster
than other types of access. Volatile is a hint to the compiler that
the variable is actually contained in a device that may change value
independently of your thread of execution, so the compiler should
access the variables permanent storage location everytime the value is
referenced to avoid operating on stale values.
At the moment I am using two potentiometers to simulate real word
inputs. I am buffering the internal 2.56V reference to supply the
potentiometers and using same internal reference for adc.
It is atmega8 with 10 bit converssion so my potentiometers read
out 0-1023 and I dont have the overflow type of problem just at the
moment.
If you are using the ADC, you won't run into the overflow problem.
Just something to be aware of.
I am working towards a controller like PID descibed on Atmels site
alternatively a fuzzy one.
John
_______________________________________________
AVR-chat mailing list
AVR-chat@nongnu.org
http://lists.nongnu.org/mailman/listinfo/avr-chat
_______________________________________________
AVR-chat mailing list
AVR-chat@nongnu.org
http://lists.nongnu.org/mailman/listinfo/avr-chat
