SWTPC 6800

[Wayne Sudol]

Hi Brad. Troubleshooting baud rate and parity issues can be very
frustrating. Might i suggest using an simple rs-232c breakout box to see
what signals are being produced on the port?  Check if the txd and rxd
blink when you hit the carriage return on your pc? Then you'll know if the
garbage seen is really a baud issue or just some noise that's put on the
line when you power up the swtpc.

On Fri, Aug 5, 2016 at 10:04 AM, Brad H <vintagecomputer at bettercomputing.net
> wrote:

> I don't think I have the right baud rate at all... as I mentioned, my MSI
> 6800 has a very well labelled serial card and I set it to whatever baud
> rate
> I was going for, set my terminal up and it wouldn't work.  It was only by
> accident that I discovered if I set the terminal to 9600 baud it worked.
> Some of the other switches worked at the speeds labelled but I think 1200
> or
> whatever was 9600.  So I wouldn't be surprised at all if this thing was
> modified beyond.  I have the MP-S, and I have the previous owner's
> handwritten sticker with the baud rates for each of the 5 DIP switches..
> and
> so I've got it set at what should be 300, but I have tried other baud rates
> up and down the line without success so far.  Going up just produces longer
> strings of random characters.  Going down produces less.
>
> I am using a 25 pin femaile to 9 pin female null modem cable.  The 9 pin
> goes into my old Thinkpad 380 and the 25 pin I have jumpered.  I followed
> the guide here:  http://www.swtpc.com/mholley/HiTerm/MPS_to_DB25.pdf
>
> Now, I don't have a DB25 female connector I could solder wires to.  So what
> I did was, where required to loop pin 4 to 5, I just took some jumper wire,
> and went from pin 4 to 5 on the female 25 pin (at the computer end), 6 to
> 20, 6 to 8, and so on.  You can get two jumper wire pins into a single
> shaft
> so the connections are all good. But maybe it doesn't work that way, I know
> null modem cables are a bit different.  In any event, wired as such, the
> terminal got nothing from the computer at all.
>
> In the event, I ended up reducing to having the MP-S set up per the
> instructions, and then just had the TX RX and ground connected from the
> card
> to the female connector.  With that setup, I get a response, albeit
> garbled.
> I'm certain cabling is an issue here.  I get a bit dyslexic with pins and
> understanding how things change on a cable like a null modem.
>
> -----Original Message-----
> From: cctalk [mailto:cctalk-bounces at classiccmp.org] On Behalf Of Fred
> Cisin
> Sent: Friday, August 5, 2016 8:52 AM
> To: General Discussion: On-Topic and Off-Topic Posts <
> cctalk at classiccmp.org>
> Subject: RE: SWTPC 6800
>
> > but as with all problem solving, Conan Doyle had it right when he said
> > "after eliminating the impossible, whatever is left, however
> > improbable, must be the solution"
>
> That quote is often horribly abused.  Diagnosis by elimination is
> inherently
> faulty.  Beginners, and "experts", will often use it to justify ridiculous
> premises, "because I eliminated everything else", while not having truly
> eliminated some of the most basic possibilities.  "It must be the
> carburetor, because everything else is new" usually meant that they had not
> gapped the points properly, or had the wrong firing order.
>
> 37? years ago, it took me 3 weeks to get ink on paper with a serial
> printing
> terminal and a TRS80, because I made stupid assumptions because I thought
> that I had eliminated . . .
> and didn't even realize that there was a bad connection between the serial
> interface and the rest of the Expansion Interface.
>
>
> Why do you think that you have the right baud rate?
> It might be 9600, but if it was expecting a printing terminal, then it
> might
> be 300, 150, 134.5, 110, 75
>
> In spite of "standardization" (everybody has a unique one of their own),
> sometimes you run into some handshaking using signals besides DSR, DTR,
> RTS, CTS.     Such as CD (8), CD2 (12)!, RI(22)
> "misuse of the standard"?  absolutely.
> (using db25 pinout - BTW, a DE9 is not a DB9.  A real DB9 consists of a DB
> shell with 1 thru 8 and 20)
>
>
> It can be frustrating.  There is at least one documented fatality from the
> frustration of serial cabling.   Guy took a printer and computer to store
> to get it cabled; after 6 weeks without success, he shot the tech.  Joe
> Campbell was sure that was an urban legend, so he tracked down the case,
> and
> mentioned it in one of his books.
>
>
> I don't intend to ever do it again; I'm getting rid of my ARC Data Tek
> 9600 DTS-1.
>
>
> --
> Grumpy Ol' Fred                 cisin at xenosoft.com
>
>