Bug in sub_rs_set_config
Posted: Wed May 19, 2010 5:14 pm
Hello
There is a bug in sub_rs_set_config(sub_handle hndl, int config, int baud).
The manual says:
The actual baudrate may slightly differ from the desired as actual baudrate is an integer quotient from dividing the 16MHz reference clock.
But it does not differ "SLIGHTLY".
It differs EXTREMELY !
I need a strange baudrate for my application of 186000 Baud.
I set this baudrate and the result is that the SUB-20 works with 200000 Baud instead.
This is completely useless as the deviation is so much that I get read errors.
If the SUB-20 would behave as the manual says, it would not deviate so much.
If you divide 16 MHz by the integer 86 the result is 186046 Baud which is very near to the desired value.
But 200000 is a deviation of 7,5 %!
Why is it not possible to divide 16 MHz by 86 ?
Either there is a bug in the DLL (or the controller) or the manual has to be corrected.
I hope that there is a bug because such an extreme deviation is unacceptable for me!
Pablo
There is a bug in sub_rs_set_config(sub_handle hndl, int config, int baud).
The manual says:
The actual baudrate may slightly differ from the desired as actual baudrate is an integer quotient from dividing the 16MHz reference clock.
But it does not differ "SLIGHTLY".
It differs EXTREMELY !
I need a strange baudrate for my application of 186000 Baud.
I set this baudrate and the result is that the SUB-20 works with 200000 Baud instead.
This is completely useless as the deviation is so much that I get read errors.
If the SUB-20 would behave as the manual says, it would not deviate so much.
If you divide 16 MHz by the integer 86 the result is 186046 Baud which is very near to the desired value.
But 200000 is a deviation of 7,5 %!
Why is it not possible to divide 16 MHz by 86 ?
Either there is a bug in the DLL (or the controller) or the manual has to be corrected.
I hope that there is a bug because such an extreme deviation is unacceptable for me!
Pablo