atn@cory.Berkeley.EDU (Alan Nishioka) (02/20/89)
I am trying to build a digital data link, as an experiment, which uses
variable length packets (from 1 to about 10,000 bits in length).
All of the UART's that I could find only work with up to 9 bit characters.
Are there any UART's which work with longer bit strings?
I thought of using Ethernet chips, which should allow the required lengths,
but they are more complicated to control, and are over-kill for my project.
Are there any good books or references that deal with building a UART?
Transmitting is not as large a problem as receiving. I could use FIFO's
for a giant shift register, but I don't know about all the subtle receiving
end considerations (timing recovery, clock skew...).
I greatly appreciate any and all help.
-----------------------------------------------------------------------------
Alan Nishioka atn@cory.berkeley.edu ...!ucbvax!atn%corylogajan@ns.UUCP (John Logajan x3118) (02/21/89)
Alan Nishioka asks: > I am trying to build a digital data link, as an experiment, which uses > variable length packets (from 1 to about 10,000 bits in length). This sort of thing is almost always done more easily in software than in hardware (also much more cheaply.) Assuming your running your "UART" at conventional speeds, you could easily write machine language (and many higher level languages) which could easily assemble your bit stream in memory. Hardware requirements are then pratically ziltch (beyond the computer), you only need a single bit (such as one bit on a parallel input port -- many printer ports have bi-directional data lines which would be sufficient for your project.) -- - John M. Logajan @ Network Systems; 7600 Boone Ave; Brooklyn Park, MN 55428 - - ...rutgers!umn-cs!ns!logajan / logajan@ns.network.com / john@logajan.mn.org -
torkil@psivax.UUCP (Torkil Hammer) (02/21/89)
In article <10108@pasteur.Berkeley.EDU> atn@cory.Berkeley.EDU (Alan Nishioka) writes:
#
# I am trying to build a digital data link, as an experiment, which uses
#variable length packets (from 1 to about 10,000 bits in length).
#
# All of the UART's that I could find only work with up to 9 bit characters.
#
Since this is an experimental project, you can put an embedded controller
to work. My choice is Intel 8751. It has a conventional 8/9 bit UART
which you can use for communication with the main CPU, and 4 kbytes of
eprom which you use for the external uart control. In other words,
program your own uart in software and use an 8751 i/o port for
the hardware link, and use the built-in uart to shuffle data to your IBM
PC or whatever, using its RS232 port.
It is doable, and not that hard. Good luck.
torkilhenry@utzoo.uucp (Henry Spencer) (02/22/89)
In article <10108@pasteur.Berkeley.EDU> atn@cory.Berkeley.EDU (Alan Nishioka) writes: > I am trying to build a digital data link, as an experiment, which uses >variable length packets (from 1 to about 10,000 bits in length). > > All of the UART's that I could find only work with up to 9 bit characters. > > Are there any UART's which work with longer bit strings? You don't want a UART at all. The A in UART stands for "asynchronous", which refers to the technique which sends start and stop bits for every character of data to resynchronize receiver and transmitter. If you want to send a whole bunch of bits without start and stop bits, that's *synchronous* transmission, so you want a USRT or USART (which does either). These come in various forms; there are lots of them about. Note that you will need pretty good clock generation, since the receiver clock has to match the transmitter clock to within half a bit time over the time needed to send 10,000 bits. This means either carefully-aligned clocks or some sort of transmission scheme that sends clock information with the data. It may be less trouble to just send 10,000 bits as 1250 8-bit bytes using a UART. Apart from some inefficiency (20% of the bits on the wire are start and stop bits), is there a reason why this isn't adequate? -- The Earth is our mother; | Henry Spencer at U of Toronto Zoology our nine months are up. | uunet!attcan!utzoo!henry henry@zoo.toronto.edu