Re: [M100] Model 200 issues with Serial Wi-Fi Adapter

[B 9]

Also, if you want to prevent the lockups, you could loop back DTR to DSR as
described earlier in this thread.

—b9

On Thu, Dec 22, 2022 at 10:10 AM B 9  wrote:

> Thanks for doing that direct test from your Tandy 200 to 102. That
> confirms that the Tandy 200's TELCOM program (and possibly anything that
> calls the ROM to read the serial port) is different. I hadn't known that
> before. I think it must be a bug because even a program that uses DSR to
> find out whether another device is connected shouldn't hang and require
> Shift-Break.
>
> Are there any ROM experts here who know what the difference is? Maybe
> it'll be possible to patch the Tandy 200's ROM.
>
> By the way, XON/XOFF is pretty much required. As Jim said, TELCOM does not
> support hardware handshaking.
>
> —b9
>
>
>
> On Thu, Dec 22, 2022 at 1:14 AM Cedric Amand  wrote:
>
>> I'm super interested in those serial problems, but I can't fully grasp
>> John's phrase below, would you please clarify and elaborate a bit ?
>>
>> > 3 wire is all the model 100 / 200 actually uses unless you're using
>> HTERM
>> > or you like the godlike simplicity of the cable check lockup, or you
>> cannot disable the cable check in software.
>>
>> I often have lockups on my M200, I'm also under the impression (but I'm
>> no expert yet) that in some way it's serial (in this case null modem)
>> implementation niside TERM differs from the one of the M102. If I do a
>> direct null modem (full cable) between a M102 and a M200, with TERM, the
>> M200 will systematically lockup when the M102 hangs up ( EXIT/F8 ) whereas
>> the reverse is not true.
>> This with both ends using (or supposed to use) hardware handshake (and
>> xon/xoff disabled)
>> This is pure observation and might be specific to my setup, of course.
>>
>> However if you could please elaborate a bit about how the M10x/M200 uses
>> the serial lines, that would be much appreciated
>>
>> I "think" from my hours of fiddling that null modem serial on the
>> M10x/M200 only works reliably with xon/xoff and the hardware only handshake
>> (at least in TERM) is flaky on the receiving end. I have yet to try HTERM.
>> (It's for sure on my todo)
>>
>>
>

Re: [M100] Model 200 issues with Serial Wi-Fi Adapter

[B 9]

Thanks for doing that direct test from your Tandy 200 to 102. That confirms
that the Tandy 200's TELCOM program (and possibly anything that calls the
ROM to read the serial port) is different. I hadn't known that before. I
think it must be a bug because even a program that uses DSR to find out
whether another device is connected shouldn't hang and require Shift-Break.

Are there any ROM experts here who know what the difference is? Maybe it'll
be possible to patch the Tandy 200's ROM.

By the way, XON/XOFF is pretty much required. As Jim said, TELCOM does not
support hardware handshaking.

—b9



On Thu, Dec 22, 2022 at 1:14 AM Cedric Amand  wrote:

> I'm super interested in those serial problems, but I can't fully grasp
> John's phrase below, would you please clarify and elaborate a bit ?
>
> > 3 wire is all the model 100 / 200 actually uses unless you're using HTERM
> > or you like the godlike simplicity of the cable check lockup, or you
> cannot disable the cable check in software.
>
> I often have lockups on my M200, I'm also under the impression (but I'm no
> expert yet) that in some way it's serial (in this case null modem)
> implementation niside TERM differs from the one of the M102. If I do a
> direct null modem (full cable) between a M102 and a M200, with TERM, the
> M200 will systematically lockup when the M102 hangs up ( EXIT/F8 ) whereas
> the reverse is not true.
> This with both ends using (or supposed to use) hardware handshake (and
> xon/xoff disabled)
> This is pure observation and might be specific to my setup, of course.
>
> However if you could please elaborate a bit about how the M10x/M200 uses
> the serial lines, that would be much appreciated
>
> I "think" from my hours of fiddling that null modem serial on the
> M10x/M200 only works reliably with xon/xoff and the hardware only handshake
> (at least in TERM) is flaky on the receiving end. I have yet to try HTERM.
> (It's for sure on my todo)
>
>

Re: [M100] Model 200 issues with Serial Wi-Fi Adapter

[Jim Anderson]

> -Original Message-
> 
> direct null modem (full cable) between a M102 and a M200, with TERM, the
> M200 will systematically lockup when the M102 hangs up ( EXIT/F8 )
> whereas the reverse is not true.

That completely tracks, because when the T102 (or M100) hangs up, it drops DTR 
(and RTS) which your null cable is sending to the T200 as DSR and CTS.  T200 
sees the loss of DSR and CTS, and TELCOM freezes.  The reverse is not true 
because the T102 (and M100) do not care whether or not they get those signals.

> This with both ends using (or supposed to use) hardware handshake (and
> xon/xoff disabled)

As a side note, TELCOM does not use hardware flow control (aside from locking 
up on the T200), you can only (E)nable or (D)isable software flow control in 
the STAT string.  If you disable software flow control, it doesn't do any flow 
control whatsoever.







jim

Re: [M100] Model 200 issues with Serial Wi-Fi Adapter

[Cedric Amand]

I'm super interested in those serial problems, but I can't fully grasp John's 
phrase below, would you please clarify and elaborate a bit ? > 3 wire is all 
the model 100 / 200 actually uses unless you're using HTERM > or you like the 
godlike simplicity of the cable check lockup, or you cannot disable the cable 
check in software. I often have lockups on my M200, I'm also under the 
impression (but I'm no expert yet) that in some way it's serial (in this case 
null modem) implementation niside TERM differs from the one of the M102. If I 
do a direct null modem (full cable) between a M102 and a M200, with TERM, the 
M200 will systematically lockup when the M102 hangs up ( EXIT/F8 ) whereas the 
reverse is not true. This with both ends using (or supposed to use) hardware 
handshake (and xon/xoff disabled) This is pure observation and might be 
specific to my setup, of course. However if you could please elaborate a bit 
about how the M10x/M200 uses the serial lines, that would be much appreciated

 I "think" from my hours of fiddling that null modem serial on the M10x/M200 
only works reliably with xon/xoff and the hardware only handshake (at least in 
TERM) is flaky on the receiving end. I have yet to try HTERM. (It's for sure on 
my todo)

Re: [M100] Model 200 issues with Serial Wi-Fi Adapter

[John R. Hogerhuis]

On Wed, Dec 21, 2022 at 10:34 PM B 9  wrote:

> On Wed, Dec 21, 2022 at 5:10 PM John R. Hogerhuis 
> wrote:
>
>> Given that it wasn't actually using flow control I see no reason to make
>> a simple 3 wire cable fail. It's punishingly opinionated. Sort of like
>> Frank Lloyd Wright building interior walls at an angle because he didn't
>> like people hanging pictures in the houses he made for them.
>>
>
> Heh! I heard they hung pictures anyway. And, of course, I agree that
> TELCOM shouldn't have cared about DSR unless it had a mighty good reason.
>
> Let's see if I got this right,
>
>Model 100: Ignores DSR
>Tandy 200: Becomes punishingly opinionated about DSR
>Tandy 102: Goes back to ignoring DSR
>
> Maybe someone at Microsoft was attempting to add flow control to the Tandy
> 200 ROM and got it partially done before they had to ship.
>
> Does anyone know how to check the DSR register on the UART from BASIC? I
> see something about I/O ports in the service manual
> ,
> but I'm not clear what port I should be reading or how.
>

https://bitchin100.com/wiki/index.php?title=Model_100_Serial_Interface

There's a I/O map at the end of the page. Did you mean Model 100/T102 or
Tandy 200? There is an article on the wiki but I think it was copied from
the Model 100 one as a starting point. Probably needs work.

 -- John.

Re: [M100] Model 200 issues with Serial Wi-Fi Adapter

[B 9]

On Wed, Dec 21, 2022 at 5:10 PM John R. Hogerhuis  wrote:

> Given that it wasn't actually using flow control I see no reason to make a
> simple 3 wire cable fail. It's punishingly opinionated. Sort of like Frank
> Lloyd Wright building interior walls at an angle because he didn't like
> people hanging pictures in the houses he made for them.
>

Heh! I heard they hung pictures anyway. And, of course, I agree that TELCOM
shouldn't have cared about DSR unless it had a mighty good reason.

Let's see if I got this right,

   Model 100: Ignores DSR
   Tandy 200: Becomes punishingly opinionated about DSR
   Tandy 102: Goes back to ignoring DSR

Maybe someone at Microsoft was attempting to add flow control to the Tandy
200 ROM and got it partially done before they had to ship.

Does anyone know how to check the DSR register on the UART from BASIC? I
see something about I/O ports in the service manual
,
but I'm not clear what port I should be reading or how.

Thanks!

—b9

Re: [M100] Model 200 issues with Serial Wi-Fi Adapter

[Joseph Colson III]

This did the trick.   Thank you.

Jim 
Anderson
 Wed, 21 Dec 2022 11:50:55 
-0800
I've run into this before with the 200 - I don't remember if it was CTS or DSR
or both that it wants to see asserted, but it definitely does behave
differently from the 100 and 102.  If you don't have one or both of those lines
high, the 200 appears to freeze up if you attempt serial communication using
TELCOM (ctrl-Break recovers).  The 100 and 102 (at least in TELCOM) don't seem
to care.

There's no real harm in jumpering both of these signals high if your wifi modem
isn't actively asserting either of them.  In a DB9 you would jumper 7 (RTS) to
8 (CTS) and 4 (DTR) to 6 (DSR).  In a DB25 those would correspond with 4 to 5
and 20 to 6.


From: Joseph Colson III 
Sent: Wednesday, December 21, 2022 10:29 AM
To: m100@lists.bitchin100.com
Cc: joecols...@outlook.com
Subject: Model 200 issues with Serial Wi-Fi Adapter

I have been using a seral Wi-Fi modem with my Model 102 without much fuss for 
the past few days.   I switched it over to the Model 200 and no matter what I 
did I could not get it to work.   I searched the boards and found plenty of 
discussion on flow control on the WiFI232 modem.   While I don't understand it 
completely, the Model100/102 does not require it while the Model 200 does.  The 
Wifi232 has solderable jumpers to fix this issue.   The board I'm using does 
not have this, so I was considering jumping pins on the DB9 side of the device, 
however I'm not completely sure what I'm supposed to JUMP.I don't know if 
all that made much sense.   Hopefully, someone can decipher what am asking.

Any Help would be appreciated,

Joe

Re: [M100] Model 200 issues with Serial Wi-Fi Adapter

[John R. Hogerhuis]

B9 - 3 wire is all the model 100 / 200 actually uses unless you're using
HTERM or you like the godlike simplicity of the cable check lockup, or you
cannot disable the cable check in software. Given that it wasn't actually
using flow control I see no reason to make a simple 3 wire cable fail. It's
punishingly opinionated. Sort of like Frank Lloyd Wright building interior
walls at an angle because he didn't like people hanging pictures in the
houses he made for them.

Generally prefer full null cables since they work with everything.

Except... this embedded wimodem seems to be a different beast.

It appears it doesn't set DSR high, and it doesn't implement the jumpers it
is supposed to have to set it. So proper through wiring doesn't work. So
you have to defeat it in the connector. And modifying the t200 rom is
probably not an option either.

>
-- John.

Re: [M100] Model 200 issues with Serial Wi-Fi Adapter

[B 9]

CTS/RTS is the current standard hardware flow control, so I'd be surprised
if any modern dongle didn't handle it as it's necessary for high speed
connections. I suggest *not* jumpering those together. I believe there is
some software for the Model 100 that can use them for hardware handshaking.
(HTERM?)

DSR/DTR is, if I recall correctly, a higher-level part of the handshake
("is there a device connected at all" rather than "is the device ready to
receive"). In my memory DSR/DTR was mainly used for flow control by some
terminals and printers in the 1980s. While I could imagine a new device
failing to assert DSR, that'd be shockingly poor design. More likely, as
John¹ stated, you've got a cable that leaves those wires out. Check for
continuity on the DSR and DTR pins and get a new cable if that's the
problem.

If your cable is fine, then Jim's loopback may be the only solution.
Looping  DTR back to DSR makes it so that when the Tandy 200 checks DSR to
see if a modem is plugged in, it will hear its own signal from DTR
announcing its presence.

By the way, when people say "jumpering the wires together", don't literally
do that. You can end up with two devices fighting to set the same wire to
different voltages. It won't blow anything up, but it'll heat up and waste
electricity. What you actually want are the pins on one end jumpered
together, but not going across the cable to the other device, the first
picture in this ASCII art drawing:

A loopback cable  Normal null modem   Normal RS232C cable
DSR <--\/--> DSR  DSR <--\ /--- DSR   DSR (DTE) <- DSR (DCE)
)  (  X
DTR ---/\--- DTR  DTR ---/ \--> DTR   DTR (DTE) -> DTR (DCE)

While unlikely, the problem could also be Carrier Detect, Pin 1. Some
devices want to see CD asserted before they are willing to talk. You can
fake it by wiring CD to DSR (on the same side of the cable), so that any
signal getting sent to the DSR port also goes to CD.

—b9

_
¹ John, you were doing 3-wire RS-232!? I was amazed when I saw that on
Radio-Shack's CoCo, but I thought that was just Radio-Shack being stingy.




On Wed, Dec 21, 2022 at 12:09 PM John R. Hogerhuis  wrote:

>
>
> On Wed, Dec 21, 2022 at 11:51 AM Jim Anderson  wrote:
>
>> I’ve run into this before with the 200 – I don’t remember if it was CTS
>> or DSR or both that it wants to see asserted, but it definitely does behave
>> differently from the 100 and 102.  If you don’t have one or both of those
>> lines high, the 200 appears to freeze up if you attempt serial
>> communication using TELCOM (ctrl-Break recovers).  The 100 and 102 (at
>> least in TELCOM) don’t seem to care.
>>
>>
>>
>> There’s no real harm in jumpering both of these signals high if your wifi
>> modem isn’t actively asserting either of them.  In a DB9 you would jumper 7
>> (RTS) to 8 (CTS) and 4 (DTR) to 6 (DSR).  In a DB25 those would correspond
>> with 4 to 5 and 20 to 6.
>>
>>
>>
>>
>>
>
> Agree... one of my few Model T pet peeves... this is the "cable check"...
> uh... feature. Don't know what they were thinking. A given cable either
> works or it doesn't... in what universe is it a feature to do a sanity
> check on lines that aren't actually involved in communication on the Model
> T? And then if there is an issue, don't give a clear error message, just
> lock up.
>
> Just weird, I don't know what the engineers were thinking. TS-DOS, TEENY
> do it as well unless you're running a version with the feature disabled.
>
> Maybe Radio Shack and the 3rd party vendors wanted people to buy better or
> proprietary cables (from them)?  Since the main result of the feature is
> that 3-wire cables don't work without looping back.
>
> -- John.
>

Re: [M100] Model 200 issues with Serial Wi-Fi Adapter

[John R. Hogerhuis]

On Wed, Dec 21, 2022 at 11:51 AM Jim Anderson  wrote:

> I’ve run into this before with the 200 – I don’t remember if it was CTS or
> DSR or both that it wants to see asserted, but it definitely does behave
> differently from the 100 and 102.  If you don’t have one or both of those
> lines high, the 200 appears to freeze up if you attempt serial
> communication using TELCOM (ctrl-Break recovers).  The 100 and 102 (at
> least in TELCOM) don’t seem to care.
>
>
>
> There’s no real harm in jumpering both of these signals high if your wifi
> modem isn’t actively asserting either of them.  In a DB9 you would jumper 7
> (RTS) to 8 (CTS) and 4 (DTR) to 6 (DSR).  In a DB25 those would correspond
> with 4 to 5 and 20 to 6.
>
>
>
>
>

Agree... one of my few Model T pet peeves... this is the "cable check"...
uh... feature. Don't know what they were thinking. A given cable either
works or it doesn't... in what universe is it a feature to do a sanity
check on lines that aren't actually involved in communication on the Model
T? And then if there is an issue, don't give a clear error message, just
lock up.

Just weird, I don't know what the engineers were thinking. TS-DOS, TEENY do
it as well unless you're running a version with the feature disabled.

Maybe Radio Shack and the 3rd party vendors wanted people to buy better or
proprietary cables (from them)?  Since the main result of the feature is
that 3-wire cables don't work without looping back.

-- John.

Re: [M100] Model 200 issues with Serial Wi-Fi Adapter

[Jim Anderson]

I’ve run into this before with the 200 – I don’t remember if it was CTS or DSR 
or both that it wants to see asserted, but it definitely does behave 
differently from the 100 and 102.  If you don’t have one or both of those lines 
high, the 200 appears to freeze up if you attempt serial communication using 
TELCOM (ctrl-Break recovers).  The 100 and 102 (at least in TELCOM) don’t seem 
to care.

There’s no real harm in jumpering both of these signals high if your wifi modem 
isn’t actively asserting either of them.  In a DB9 you would jumper 7 (RTS) to 
8 (CTS) and 4 (DTR) to 6 (DSR).  In a DB25 those would correspond with 4 to 5 
and 20 to 6.







jim

From: M100  On Behalf Of B 9
Sent: Wednesday, December 21, 2022 11:06
To: m...@bitchin100.com; Joseph Colson III 
Subject: Re: [M100] Model 200 issues with Serial Wi-Fi Adapter

CAUTION External Sender: Do not click links or open attachments unless you 
recognize the sender and know the content is safe.

That is odd. Handshaking and Carrier Detect can cause peculiar problems, but 
I'm wondering if maybe something else is wrong with the serial port.

I have a Tandy 200 and my belief is that the the serial port is the same as the 
102 (which came later). The only difference I know of is that one needs to add 
the extra letters "NN" to TELCOM's `STAT` command. (For example, `STAT 98N1ENN` 
instead of `STAT 98N1E`). If you forget the extra letters, TELCOM will silently 
ignore the setting.

[image.png]
Does your Tandy 200's serial port work when connecting to anything else? For 
example, to your Tandy 102 via a null modem?

—b9

On Wed, Dec 21, 2022 at 8:29 AM Joseph Colson III 
mailto:joecols...@outlook.com>> wrote:
I have been using a seral Wi-Fi modem with my Model 102 without much fuss for 
the past few days.   I switched it over to the Model 200 and no matter what I 
did I could not get it to work.   I searched the boards and found plenty of 
discussion on flow control on the WiFI232 modem.   While I don’t understand it 
completely, the Model100/102 does not require it while the Model 200 does.  The 
Wifi232 has solderable jumpers to fix this issue.   The board I’m using does 
not have this, so I was considering jumping pins on the DB9 side of the device, 
however I’m not completely sure what I’m supposed to JUMP.I don’t know if 
all that made much sense.   Hopefully, someone can decipher what am asking.

Any Help would be appreciated,

Joe

Re: [M100] Model 200 issues with Serial Wi-Fi Adapter

[B 9]

That is odd. Handshaking and Carrier Detect can cause peculiar problems,
but I'm wondering if maybe something else is wrong with the serial port.

I have a Tandy 200 and my belief is that the the serial port is the same as
the 102 (which came later). The only difference I know of is that one needs
to add the extra letters "NN" to TELCOM's `STAT` command. (For example, `STAT
98N1ENN` instead of `STAT 98N1E`). If you forget the extra letters, TELCOM
will silently ignore the setting.

[image: image.png]
Does your Tandy 200's serial port work when connecting to anything else?
For example, to your Tandy 102 via a null modem?

—b9

On Wed, Dec 21, 2022 at 8:29 AM Joseph Colson III 
wrote:

> I have been using a seral Wi-Fi modem with my Model 102 without much fuss
> for the past few days.   I switched it over to the Model 200 and no matter
> what I did I could not get it to work.   I searched the boards and found
> plenty of discussion on flow control on the WiFI232 modem.   While I don’t
> understand it completely, the Model100/102 does not require it while the
> Model 200 does.  The Wifi232 has solderable jumpers to fix this issue.
> The board I’m using does not have this, so I was considering jumping pins
> on the DB9 side of the device, however I’m not completely sure what I’m
> supposed to JUMP.I don’t know if all that made much sense.   Hopefully,
> someone can decipher what am asking.
>
>
>
> Any Help would be appreciated,
>
>
>
> Joe
>
>
>

[M100] Model 200 issues with Serial Wi-Fi Adapter

[Joseph Colson III]

I have been using a seral Wi-Fi modem with my Model 102 without much fuss for 
the past few days.   I switched it over to the Model 200 and no matter what I 
did I could not get it to work.   I searched the boards and found plenty of 
discussion on flow control on the WiFI232 modem.   While I don't understand it 
completely, the Model100/102 does not require it while the Model 200 does.  The 
Wifi232 has solderable jumpers to fix this issue.   The board I'm using does 
not have this, so I was considering jumping pins on the DB9 side of the device, 
however I'm not completely sure what I'm supposed to JUMP.I don't know if 
all that made much sense.   Hopefully, someone can decipher what am asking.

Any Help would be appreciated,

Joe