Path: utzoo!attcan!uunet!wuarchive!zaphod.mps.ohio-state.edu!mips!daver!wombat!george
From: george@wombat.UUCP (George Scolaro)
Newsgroups: sci.electronics
Subject: Re: Need pointer to CMOS Uart
Message-ID: <458@wombat.UUCP>
Date: 19 Mar 90 16:19:39 GMT
References: <53558@microsoft.UUCP> <34002@prls.UUCP> <1990Mar16.055156.19874@utzoo.uucp> <445@wombat.UUCP> <1990Mar19.014241.5473@utzoo.uucp>
Reply-To: george@wombat.UUCP (George Scolaro)
Organization: Assn. for the prevention of Polar Bears and Kangaroos
Lines: 37

In article <1990Mar19.014241.5473@utzoo.uucp> henry@utzoo.uucp (Henry Spencer) writes:
>In article <445@wombat.UUCP> george@wombat.UUCP (George Scolaro) writes:
>>...The only
>>problem with the part (not a bug as such) is that it lacks sufficient
>>handshake signals...  Signetics gave the chip a bucket of general purpose
>>inputs, but not enough general purpose outputs, so you need an external
>>latch to take care of the dtr...
>
>This is actually the right decision if you are short of pins.  Inputs

But they weren't short of pins. They have 16 pins for external clocks, 8 for
the receivers and 8 for the transmitters (note these pins can also be
outputs, the part is very programmable). Now, the normal user of this part
will want full modem control and the same baud rate on each receiver -
transmitter, so freeing up 8 outputs for dtr. In output mode the 16 pins
become TxRDY and RxRDY/FFULL, which don't seem to useful since they can be
read from within the part via registers. Their only use is to (optionally)
wire-or them to generate RxRDY or TxRDY interrupts (in addition to the 4
separate interrupts that come from each pair of channels - like the 2681
DUART). It just seems to me that a slightly different selection (or
superset) of options would have made the part a single chip solution rather
than needing a separate latch and hence decode circuit plus pcb routing
etc...

Note: The part has programmable baud rate control for each of the channels
internally. The reason for the external baud rate control is for some of
the higher rates (such as 115kbaud). Also, Signetics have rev-ed the part
from the original version  (this is the 2698B), and in doing so changed the
assignments of some of the input/output functions, but

so close to perfect....

regards,

-- 
George Scolaro
george@wombat.bungi.com                [37 20 51 N / 122 03 07 W]