50 baud is dead

[spellberg_robert]

Chuck Robey wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> 
> Warren Block wrote:
> 
>>On Sun, 13 Jul 2008, Daniel O'Connor wrote:
>>
>>
>>>On Sat, 12 Jul 2008, Christian Weisgerber wrote:
>>>
>>>>Daniel O'Connor <doconnor at gsoft.com.au> wrote:
>>>>
>>>>>>It turns out that uplcom(4) adapters don't support the required
>>>>>>speed of 50 baud anymore.
>>>>>
>>>>>You know, it might actually support it if you hack up the driver.
>>>>>The source says the PL2303X "can set any rate".
>>>>
>>>>The data sheet disagrees.
>>>>
>>>>"The flexible baud rate generator of PL-2303X could be programmed to
>>>> generate any rate between 75 bps and 6Mbps."
>>>
>>>I guess you're out of luck then :(
>>
>>Or just try it at 75 and see if there's enough slop to read the 50 baud
>>data.
> 
> There won't be, 50% is way too far off, I think (from experiments I did 30 years
> back) that about 12% is the limit, when I had access to the best regenerators.
> Regens don't even get used anymore.  I forget the actual limit I probed, but I
> know 1/2 over is way too far to go.
> 
>>-Warren Block * Rapid City, South Dakota USA



first, in very round numbers,
   you get about five per cent error in the bit_time
   for fifty per cent error in ten bits.
usually, there are three reads, all near the center of the bit_time.
so, ya gotta be kinda close.
two per cent leaves plenty of room for slop.

i have an inexpensive hardware solution.

if any one is interested, please respond off_list, unless
   you think it appropriate for -chat.

rob