[WriteLog] serial interface PCI expansion boards
Dave W7DPW
w7dpw@attbi.com
Wed, 10 Jul 2002 14:59:18 -0700
The UART takes the 5-bit parallel Baudot code and converts it into a serial
stream and its start(1) and stop(1.5) bits on transmit and conversely on
Receive, it converts the 5-bit serial code, along with its start(1) and stop
bits(1.5), and converts to parallel 5-bit code. The transmitted code
appears on the TxD line. The stop bit is not always 1.5 but in anycase is
determined by the LineControlRegister(LCR) Bit-2.
For Information on UART look at the following WebPage:
http://www.byterunner.com/16550.html
The Line Control Register (LDR) determines the data format . WordLength
Bit-0,1
Additional Information at:
http://www.byterunner.com/why.html
Dave W7DPW
----- Original Message -----
From: "Steve Dyer" <sdyer@interlogue.com>
To: "Joe Subich, K4IK" <k4ik@subich.com>; "'Ron Stordahl'"
<ronald.stordahl@digikey.com>; <n2tk@arrl.net>; <aa5au@bellsouth.net>;
"'Writelog (E-mail)'" <writelog@contesting.com>
Sent: Wednesday, July 10, 2002 2:06 PM
Subject: RE: [WriteLog] serial interface PCI expansion boards
> The 5 bit baudot code data is NOT on the TXD line. The TXD line only
carries
> mark/space signaling which represents baudot characters.
>
> A fine point I realize. Look at the FSK interfaces in the WL or MMTTY help
> files. They do not do any baudot translations, just isolation/level
> conversion. I don't know of any rig that understands baudot input. All FSK
> inputs are for mark/space.
>
> Both RttyWrite and MMTTY are bit banging. When transmitting, letters and
> numbers are translated to baudot representation, then translated to on/off
> signals on the TXD line.
>
> If there are any timing problems with PC generated CW or RTTY, look for
> processes hogging the machine or upgrade to a faster CPU. Most all the
> multi-tasking issues have been ironed out by a combination of much faster
> hardware and Microsoft improving the process scheduler.
>
> Best 73,
>
> Steve, W1SRD
>
> -----Original Message-----
> From: writelog-admin@contesting.com
> [mailto:writelog-admin@contesting.com]On Behalf Of Joe Subich, K4IK
> Sent: Wednesday, July 10, 2002 11:40 AM
> To: 'Ron Stordahl'; 'Steve Dyer'; n2tk@arrl.net; 'Writelog (E-mail)'
> Subject: RE: [WriteLog] serial interface PCI expansion boards
>
>
>
> If the five bit data is on the TXD line, it is using the 5-bit
> capability of the UART. If the output is on the RTS or DTR,
> then the CPU is doing the bit banging. Far better to have the
> UART handle the timing.
>
> 73,
>
> ... Joe, K4IK
>
>
> > -----Original Message-----
> > From: writelog-admin@contesting.com
> > [mailto:writelog-admin@contesting.com] On Behalf Of Ron Stordahl
> > Sent: Wednesday, July 10, 2002 1:55 PM
> > To: Steve Dyer; Ron Stordahl; n2tk@arrl.net; Writelog (E-mail)
> > Subject: Re: [WriteLog] serial interface PCI expansion boards
> >
> >
> > Interesting.....this is surprising to me, I would have
> > thought it would use
> > the UART in 5 bit mode. The reason is that I would not think
> > you could
> > trust the timing in a multitasking machine to to generate
> > consistent data,
> > while the UART would do so by design...and it is also
> > buffered for a few or
> > more characters on the send side too. Keep in mind that
> > systems that try to
> > generate CW by toggling DTR do a inconsistent job, and
> > generatting rtty
> > would require even more precision (timing accuracy).
> >
> > Are you sure about this Steve? TXD is data from the UART, I
> > don't think a
> > program can toggle it (could be wrong). RTS on the other
> > hand would be
> > under driver control, however W2K and XP might not allow a
> > user task to do
> > this with the built in driver. Speculation here Ill admit.
> >
> > On the PK900, no, of course not, its standard 8 bit no parity
> > on the serial
> > line.
> >
> > Ron N5IN
> >
> >
> > ----- Original Message -----
> > From: "Steve Dyer" <sdyer@interlogue.com>
> > To: "Ron Stordahl" <ronald.stordahl@mailhub.digikey.com>;
> > <n2tk@arrl.net>;
> > "Writelog (E-mail)" <writelog@contesting.com>
> > Sent: Wednesday, July 10, 2002 12:25 PM
> > Subject: RE: [WriteLog] serial interface PCI expansion boards
> >
> >
> > > Direct FSK does not use 5 bit. MMTTY, WL, etc. toggle a
> > single serial line
> > > (TXD or RTS) to generate FSK.
> > >
> > > I don't know if the PK900 actually communicates with the
> > serial using 5
> > bit
> > > mode, but I would be very suprised. I assume all TNC's use
> > standard ASCII
> > > characters in 8 bit mode. The MFJ unit I had did it this way.
> > >
> > > You are correct, don't get or use ISA. One additional point
> > on ISA is the
> > > Plug and Play enumerator cannot probe ISA slots fully.
> > >
> > > Steve, W1SRD
> > >
> > > -----Original Message-----
> > > From: writelog-admin@contesting.com
> > > [mailto:writelog-admin@contesting.com]On Behalf Of Ron Stordahl
> > > Sent: Wednesday, July 10, 2002 7:22 AM
> > > To: n2tk@arrl.net; Writelog (E-mail)
> > > Subject: Re: [WriteLog] serial interface PCI expansion boards
> > >
> > >
> > > I have two of the Byterunner PCI-410HSP-9's. They are PCI
> > cards, use a
> > > single IRQ and have 4 serial and 1 parallel ports. Just
> > for a test I put
> > > both of them in a single PC so that I had a total of 10 COM
> > ports (the
> > > mainboard has 2 of its own as is common) and 3 parallel
> > ports (1 on the
> > > mainboard). I tested all of the ports, they all worked. I
> > did not test
> > in
> > > 5 bit mode for direct FSK as I am not running soundcard
> > fsk, rather I am
> > > using a PK900 for that.
> > >
> > > I did ask the guy at Byterunner this card supported 5 bit
> > mode and he said
> > > it does....however I would test it myself before I believed it.
> > >
> > > The additional ports (8 coms and 2 parallel ports) use a
> > single IRQ (=9).
> > >
> > > You might want to check http://byterunner.com and call
> > their technical
> > > service line, you will probably talk with 'Sean' (he has
> > answered every
> > > time I have phoned--could be a one man operaton) and put
> > him on the spot
> > on
> > > the 5 bit issue. Ask him to guarantee it will work or
> > money back and test
> > > it yourself and let us know one way or the other.
> > >
> > > Ron N5IN
> > >
> > > PS--Don't buy ISA cards....they are dying!!!
> > >
> > >
> > >
> > > ----- Original Message -----
> > > From: "N2TK" <tony.kaz@verizon.net>
> > > To: "Writelog (E-mail)" <writelog@contesting.com>
> > > Sent: Wednesday, July 10, 2002 7:16 AM
> > > Subject: [WriteLog] serial interface PCI expansion boards
> > >
> > >
> > > > Time to add more serial ports. I just have the 2 that are on the
> > > > motherboard. Which one to buy? I see serial boards
> > offered by Boca,
> > SIIG,
> > > > Lava, WinComm, Rocketport, etc. It seems they all use one
> > IRQ. Is there
> > > any
> > > > problems with using one IRQ? If not, I can save a few IRQ's by
> > installing
> > > a
> > > > 4 port serial PCI card and disable the onboard serial ports. If
> > > performance
> > > > is better by using individual IRQ's, then I can get a 2
> > port serial card
> > > and
> > > > use these ports along with the 2 ports on the motherboard.
> > > > Is there a preferred PCI board(s) and/or ones to stay away from?
> > > > Tnx
> > > > Tony
> > > > N2TK
> > > >
> > > > _______________________________________________
> > > > WriteLog mailing list
> > > > WriteLog@contesting.com
> > > > http://lists.contesting.com/mailman/listinfo/writelog
> > > >
> > >
> > > _______________________________________________
> > > WriteLog mailing list
> > > WriteLog@contesting.com
> > > http://lists.contesting.com/mailman/listinfo/writelog
> > >
> > >
> > >
> >
> > _______________________________________________
> > WriteLog mailing list
> > WriteLog@contesting.com
> > http://lists.contesting.com/mailman/listinfo/writelog
> >
>
> _______________________________________________
> WriteLog mailing list
> WriteLog@contesting.com
> http://lists.contesting.com/mailman/listinfo/writelog
>
>
> _______________________________________________
> WriteLog mailing list
> WriteLog@contesting.com
> http://lists.contesting.com/mailman/listinfo/writelog