87536 25-MAY 05:54 Telecom (6809)
9600
From: JEVESTAL To: ALL
I just recently got a 14.4 Fax modem, a Best Data Smart One 1442FX. Does
anybody here have one too?
My CoCo 3 is an unmodified 6809 machine, I do not have a 6309 installed
yet. I am usisng SACIA (Randy’s high speed patches installed) with /t2
xmoded to par=12 and xtp=88. I use CoNect single RS-232 port (compat.
with Tandy rs-232 pak). The is in slot one of my mpi (upgraded for CoCo 3).
I am using ed. 9 clock for the BB XT-RTC. My modem is set as follows:
Compression off and RTS/CTS hardware flow control (AT&K3%C0%M0&D2X4).
I can’t get reliable 9600 baud access at all. Calling Delphi on
the 9600 Tymnet node yeilds dropped characters (average 1 or 2 per
screen), I get the same results from local boards. I use
Supercomm At 9600, Zmodem times out and Ymodem yeilds less than 450 CPS.
This is on both Delphi and local boards. At 4800 baud the modem seems
reliable if I do not do any multi-tasking (at 2400 I can multitask
any program and still download without errors).
Any sugestions on how I can get more reliable 9600 baud? Please
do not respond “Get a 6309 with NitroOS9 (or Powerboost). I plan
to do that one of the days but can’t afford it now.
Thanks,
Jim
======================== InfoXpress 01.01.00 OS-9/6809 ======================
| Narnia BBS: 11pm-8am PDT serving CoCo OS-9 users
—-|—- StG network: sysop@Narnia “Exclusively OS-9″
| Delphi: JEVestal [at] delphi [dot] com
Marysville, CA InterNet: JEVestal [at] narnia [dot] wa [dot] com
| or : JEVestal [at] citrus [dot] sac [dot] ca [dot] us
(916) 743-2617 Voice: 8am-11pm PDT :1 Corinthians 1:18 & Romans 1:16
=============================================================================
Jim Vestal: Assistant editor of The International OS-9 Underground,
“Magazine dedicated to OS-9/OSK Users Everywhere
87560 26-MAY 01:35 Telecom (6809)
RE: 9600 (Re: Msg 87536)
From: MITHELEN To: JEVESTAL
You first step is to verify that hardware handshakeing is taking place. Get
a RS-232 line status indicator. If you don’t see the RTS lite going off
and on along with the RD light, then something is wrong with your modem
setup, or the RS-232 port/setup.
87584 27-MAY 23:13 Telecom (6809)
RE: 9600 (Re: Msg 87536)
From: TEDJAEGER To: JEVESTAL
Just letting you know that I have the same modem though it is hooked to
my MM1. I have not yet achieved any 9600 baud delphi connections but havent
tried real hard yet. Initially I was getting 9600 CONNECT on my screen,
then garbage echoing, and then nada. I will be piddling with this over
the next month so lets exchange notes, huh?
Bests
—TedJaeger
87601 28-MAY 18:28 Telecom (6809)
RE: 9600 (Re: Msg 87584)
From: JEVESTAL To: TEDJAEGER
> Just letting you know that I have the same modem though it is hooked to
> my MM1. I have not yet achieved any 9600 baud delphi connections but
> havent tried real hard yet. Initially I was getting 9600 CONNECT on my
> screen, then garbage echoing, and then nada. I will be piddling with this
> over the next month so lets exchange notes, huh?
Have you been able to call out to local PC boards at 9600 or 14.4?
Unless the port is real bad there is no reason a 68K based system
can’t do high speed connects. What is your modem settings? Do a
AT&V to list them to your terminal. From I hear both S-term and
TasCom (OSterm) should work fine.
Jim
87563 26-MAY 08:43 Telecom (6809)
RE: 9600 (Re: Msg 87560)
From: JEVESTAL To: MITHELEN
> You first step is to verify that hardware handshakeing is taking place.
> Get a RS-232 line status indicator. If you don’t see the RTS lite going
> off and on along with the RD light, then something is wrong with your
> modem setup, or the RS-232 port/setup.
I don’t have a line status indicator, if they are cheap (Radio Shack?)
then I may pick one up. In the meantime assuming that handshaking is
taking place then what?
By the way I have tried downloading to a ramdisk too, but ymodem at
9600 yeilds about 450 or less cps.
Is PROTOCOL:LAPM (V-42 LAP-M Error Correction) the same as handshaking?
Usually my modem responds with this message.
Jim
87564 26-MAY 16:31 Telecom (6809)
RE: 9600 (Re: Msg 87563)
From: ISC To: JEVESTAL
Jim,
I thought I would jump in here because I want to learn all I can to facilitate
my own move to a higher speed modem when it comes. “Handshaking” is the term
used for the exchange of signals which takes place during a transmission
between two modems. You can monitor this process by purchasing an RS-232
“mini-tester” from Radio Shack for only a few bucks (I am sure mine was less
than $5). The “mini-tester” plugs in to your RS-232 output plug on one end
and to your RS-232 output cable on the other end. It has 7 indicator lights
for TD – transmit data, RD – receive data, RTS – ready to send, CTS – clear
to send, DSR – Data set ready, DTR – data terminal ready and CD – carrier
detect. By watching what happens during transmission, you can monitor
when the signals are being sent to and from your modem.
Error checking protocols are methods of checking the data stream for various
patterns to detect or correct errors caused by line noise or interference.
The RS-232 signals must be sent in the correct sequence so that both modems
will continue to transmit and receive “in sync” with each other. On the
other hand there are several error checking and correction protocols, so
you must know which are compatible with your system and the host system
to which you are calling.
I hope that this has not confused you more, but I am sure that our colleagues
here will add to the discussion and that we will both learn more.
Bill
87568 26-MAY 19:59 Telecom (6809)
RE: 9600 (Re: Msg 87564)
From: MITHELEN To: JEVESTAL
Assumeing you are getting soe kind of RTS/CTS handshakeing, and the CoCo
is indeed getting the signals. Then, you might have to fine tune some of
the “S” register settings that deal with intewrnal buffers and transition times
in the modem.
Get the tester first, they are cheep, definately under 10$, and help solve
many problems (I just had to use mine to figure out why my MAI terminal
was not handshakeing properly… Seems it likes to use DTR instead of RTS
to signal that t is bust, and to not send anything to it)
Also, don’t expect to get massivly high throughput rates until you are
running in 6309 mode. What you want to acheive first are error free receives.
Then you might want to do some hacks to fix the CTS problem so you
can get error free transmissions.
Another side note… Saw a CoCo connected at 38.4K baud to my Unix box
at the fest and have no problem handleing a zmodem receive (only got 984 CPS)
but, it didn’t loose any characters or interupts, and we even loaded the
coco with a few other tings to do…this was using a serail port
with a 16550 UART. It was in Mark Marlette’s (?SP) machine (Quite an
impessive machine!)
–
Paul Jerkatis – SandV BBS (708)352-0948: Chicago Area OS-9 Users Group
87574 27-MAY 01:00 Telecom (6809)
RE: 9600 (Re: Msg 87564)
From: JEVESTAL To: ISC
> I thought I would jump in here because I want to learn all I can to
> facilitate
> my own move to a higher speed modem when it comes. “Handshaking” is the
> term used for the exchange of signals which takes place during a
> transmission between two modems. You can monitor this process by
Thanks for the input.
Jim
87575 27-MAY 01:01 Telecom (6809)
RE: 9600 (Re: Msg 87568)
From: JEVESTAL To: MITHELEN
> Assumeing you are getting soe kind of RTS/CTS handshakeing, and the CoCo
> is indeed getting the signals. Then, you might have to fine tune some of
> the “S” register settings that deal with intewrnal buffers and transition
> times in the modem.
> Get the tester first, they are cheep, definately under 10$, and help
> solve many problems (I just had to use mine to figure out why my MAI
> terminal was not handshakeing properly… Seems it likes to use DTR
> instead of RTS to signal that t is bust, and to not send anything to it)
> Also, don’t expect to get massivly high throughput rates until you are
> running in 6309 mode. What you want to acheive first are error free
> receives. Then you might want to do some hacks to fix the CTS problem so
> you can get error free transmissions.
Thanks Paul, I will pick up one of those testers soon.
> Another side note… Saw a CoCo connected at 38.4K baud to my Unix box
> at the fest and have no problem handleing a zmodem receive (only got 984
> CPS) but, it didn’t loose any characters or interupts, and we even loaded
> the coco with a few other tings to do…this was using a serail port
> with a 16550 UART. It was in Mark Marlette’s (?SP) machine (Quite an
> impessive machine!)
How can the average CoCo user get one of these 16550 ports? There’s
several of us that wonts one!
Jim
87602 28-MAY 18:30 Telecom (6809)
RE: 9600 (Re: Msg 87568)
From: JEVESTAL To: MITHELEN
> Assumeing you are getting soe kind of RTS/CTS handshakeing, and the CoCo
> is indeed getting the signals. Then, you might have to fine tune some of
> the “S” register settings that deal with intewrnal buffers and transition
> times in the modem.
Which S register settings in particular? I’ve scanned my modem’s manual
and didn’t see any that I think could be changed that would help except
for maybe S26?
> Get the tester first, they are cheep, definately under 10$, and help
> solve many problems (I just had to use mine to figure out why my MAI
> terminal was not handshakeing properly… Seems it likes to use DTR
> instead of RTS to signal that t is bust, and to not send anything to it)
I got one of the testers yesterday, Radio Shack must have risen there
prices there are no $15 not under $10. It is a neat little toy. It
seems that RTS/CTS handshaking is working properly. I noticed that
the RTS light goes red only during streaming protocols such as
zmodem. I guess the ymodem protocol itself handles the flow control.
> Also, don’t expect to get massivly high throughput rates until you are
> running in 6309 mode. What you want to acheive first are error free
> receives. Then you might want to do some hacks to fix the CTS problem so
> you can get error free transmissions.
It will be a while before I can get a 6309 installed, not anytime soon.
I would like to hear from other people who have claimed to have gotten
reliable communication at 9600 baud from the CoCo without the 6309.
I still say that it doesn’t look it the CoCo can do it. Please
someone show me that I am wrong by telling me how MY CoCo can do it.
Jim
======================== InfoXpress 01.01.00 OS-9/6809 ======================
| Narnia BBS: 11pm-8am PDT serving CoCo OS-9 users
—-|—- StG network: sysop@Narnia “Exclusively OS-9″
| Delphi: JEVestal [at] delphi [dot] com
Marysville, CA InterNet: JEVestal [at] narnia [dot] wa [dot] com
| or : JEVestal [at] citrus [dot] sac [dot] ca [dot] us
(916) 743-2617 Voice: 8am-11pm PDT :1 Corinthians 1:18 & Romans 1:16
=============================================================================
Jim Vestal: Assistant editor of The International OS-9 Underground,
“Magazine dedicated to OS-9/OSK Users Everywhere
87577 27-MAY 01:59 Telecom (6809)
RE: 9600 (Re: Msg 87575)
From: MITHELEN To: JEVESTAL
WELP… I’m not sure if the serial port board he had was of his own design,
or, if it was a redesigned version of a prototype CoCoIO board
( it actually had 2 16550′s on it) He also had his own home
brew 2 meg upgrade, and IBM keyboard interface.
87604 28-MAY 19:31 Telecom (6809)
RE: 9600 (Re: Msg 87602)
From: MITHELEN To: JEVESTAL
The coco can indead do it running in 6809 mode… I’ve done it myself when
SandV was running on a CoCo. That was a long time ago though. The reason that
YModem and Xmodem seem they don’t require the RTS/CTS handshakeing is that
because they are blocked protocals, the receive buffer is must likely
large enough (in SAcia) to hold an entire 1K packet, so the system never
gets to the point that it needs to halt the transmitter. I can’t recall your
original setup, but, do make sure you have the new Edition 9 clock installed
with SAcia…
Try playing with the S49 (ASB Buffer size lower limit) and S50 (ASB
buffer size upper limit)
87609 28-MAY 20:03 Telecom (6809)
RE: 9600 (Re: Msg 87602)
From: DBREEDING To: JEVESTAL
> I would like to hear from other people who have claimed to have gotten
> reliable communication at 9600 baud from the CoCo without the 6309.
>
> I still say that it doesn’t look it the CoCo can do it. Please
> someone show me that I am wrong by telling me how MY CoCo can do it.
I’ll relate some of my experiences, don’t take word for gospel, as I didn’t
do some of the tweaking Paul spoke of. My setup: Coco 3, 6309 w/Powerboost,
MPI w/ Tandy RS-232 Pak – it has 6551 ACIA, not 6551A, have one but didn’t
install it. I have an edition 9 (?) clock – thought it was supposed to be
edition 7. It’s for Disto clock, anyway. My modem is Hayes Optima 2400.
I have my /t2 “xtp” set to 5 “par”= 2.
I have supercomm 2.2, also OSTerm. In logging on to local BBS w/40 MHZ PC,
I would lose some characters especially when a big bunch of stuff came in to
display. On a ZModem d/l, I would get a few errors (that’s with error-
correction turned on in modem). I think on one file about 100K, I got 3 or
4 errors. This was all at 9600, and the 6551 might be too weak for this
rate. I began logging on @ 4800 and it’s beautiful.
I notice with error-corr & h/w handshake on, my “rd” light blinks rapidly
with CIS & Delphi. On the local BBS, the light blinks at oh, maybe 1/2 that
rate. However, the BBS has compression turned on, with CIS & Delphi comp.
is off, but err-corr is on.
This is only my experience. I wonder, speaking to Paul, are your successful
9600 connections with another coco? I can see where a CoCo-CoCo connection
would be quite successful, since I would think neither would be supplying
the data faster than the other could receive. But can a Coco keep up with
a much faster computer at 9600? I’m thinking that by the time the coco
detects a buffer-almost-full condition and shuts flow off, that the buffer
might be overflowed? Could this be?
— David Breeding –
CompuServe : 72330,2051
Delphi : DBREEDING
*** Sent via CoCo-InfoXpress V1.01 ***
^^^^ ^^^^^^^^^^
87608 28-MAY 19:37 Telecom (6809)
RE: 9600 (Re: Msg 87601)
From: MITHELEN To: TEDJAEGER
I Take it you are using /T3 or /T4? If not, that is most likely your
:| P4
problem… Also, check the jumpers on the paddle board. On the P4 jumper
block, you should have it jumpered as follows:
(where ‘|’ is a jumper, and yopu are looking at the board from the front)
Also, make sure you are the TYPE set to 80, or have RTS/CTS flow control
enabled from the term program (if it provoids that option)
87622 29-MAY 00:36 Telecom (6809)
RE: 9600 (Re: Msg 87609)
From: MITHELEN To: DBREEDING
My experiances were between a CoCo, and a 486/33 Unix box. Also, I was using
a true 9600 baud modem (ie, v.32), but, that is besides the point. One of the
newer version of SAcia may be needed to get the reliability I had. I can’t
really give too much more help, cause, I no longer have a CoCo, and it has
been well over a year since I used that setup.
–
Paul
87657 31-MAY 00:06 Telecom (6809)
RE: 9600 (Re: Msg 87622)
From: DBREEDING To: MITHELEN
> My experiances were between a CoCo, and a 486/33 Unix box. Also, I was
> using a true 9600 baud modem (ie, v.32), but, that is besides the point.
OK, that answers my question. I didn’t pursue the project all that far. I
did have SAcia. My modem was a 2400 Hayes using v.42bis. I think maybe the
screen output was overtaxing my system or something – well I did get errors
on zmodem downloading so I dunno. I had everything I was told I needed -
except didn’t do the DSR/? switching (think it was DSR) that was recommended
(believe this was developed by you, wasn’t it?).
> had. I can’t really give too much more help, cause, I no longer have a
> CoCo, and it has been well over a year since I used that setup.
Well, I suppose it’s a moot question for me, too. Just received my
System 5 Fri the 28th. Soon as I get my serial ports rewired for straight
RS-232, (or get the internal modem), I’ll probably be using the OSK sys
for all telecom.
— David Breeding –
CompuServe : 72330,2051
Delphi : DBREEDING
*** Sent via CoCo-InfoXpress V1.01 ***
^^^^ ^^^^^^^^^^
87662 31-MAY 02:20 Telecom (6809)
RE: 9600 (Re: Msg 87657)
From: MITHELEN To: DBREEDING
The DSR/CTS swamp hack was actally thought up, and implemented first
by someone else (who’s name eludes me) that was from Minesota…. I added
one or two fixes to that hack (it tended to loose the real status for some
reason on occasions, I forget… it’s been a while) and the actual
“fix ” (ie assembly) code came from someone else (Randy I think) I just
peiced together all the hacks, and put them together in one source
file, and archive…
Also… the DSR/CTS hack only effected TRANSMISSION, not Receptiuon.
Anyhow.. like you said, it is a moot point for you and myself now.. although
I’m sure others still want to get this high speed thing worked out…
87695 1-JUN 22:10 Telecom (6809)
RE: 9600 (Re: Msg 87662)
From: JEVESTAL To: MITHELEN (NR)
> The DSR/CTS swamp hack was actally thought up, and implemented first
> by someone else (who’s name eludes me) that was from Minesota…. I added
> one or two fixes to that hack (it tended to loose the real status for
> some reason on occasions, I forget… it’s been a while) and the actual
> “fix ” (ie assembly) code came from someone else (Randy I think) I just
> peiced together all the hacks, and put them together in one source
> file, and archive…
> Also… the DSR/CTS hack only effected TRANSMISSION, not Receptiuon.
>
> Anyhow.. like you said, it is a moot point for you and myself now..
> although I’m sure others still want to get this high speed thing worked
> out…
I have read the patch doc file, but can’t use it since I can’t modify
my cable, hopeuflly my friend Tim will help set up a modified cable
for us to test.
I uploaded a fire for the first time at 4800 baud got a blazing 315 CPS
using the latest SZ! Why so slow, moving the file to ramdisk did
not help much, so it’s not the hard drive slowing down the tranfer.
Unlike Zmodem receive at 4800 baud (which yields about 400 CPS) I noticed
that the SD light on my modem was not on constrantly using zmodem. Ymodem
batch yeilded closer to the expected 400+ CPS…. why is zmodem so slow?
I would like the users on my board to be able to download using SZ but
I can’t recomend a 300 CPS transfer rate… that is way too slow.
Any sugestions… if any CoCo user gets better result WITHOUT a 6309
could you please send me a copy of your os9boot file, and a copy of
your modem settings?
Jim “It can’t be done on a CoCo” Vestal
87755 4-JUN 02:05 Telecom (6809)
RE: 9600 (Re: Msg 87695)
From: MITHELEN To: JEVESTAL
Welp… Zmodem was originally written on big, fast computers with
plenty of CPU cycles to spare, and unforch, the CoCo version has not
had many optimizations done. Sendind is esciacially bad, because the
SACia doesn’t have any kind of send buffer, like it does for receive, so
it gobbles up lots of CPU. Teh old, original, ACIAPAK actually had a send
buffer. It would be great if some hacker got around to hacking it into
SACia. I’ve mentioned it to many of the local CoCo hackers, and it
seems no one has the time to do it.
Hopefully, some more assembly optimization will get done on the
CoCo version of Zmodem one of these days, and it will bring the
throughput up to reality… even with one of the “mythical” CoCoIO type
ports with the high speed 16550 uarts, zmodem as it is tops out just
under 1000 CPS on receiveing… didn’t do any sending tests, but I
imagine there isn’t much improvement there.
–
Paul
87784 5-JUN 21:45 Telecom (6809)
RE: 9600 (Re: Msg 87755)
From: JEVESTAL To: MITHELEN
> Welp… Zmodem was originally written on big, fast computers with
> plenty of CPU cycles to spare, and unforch, the CoCo version has not
> had many optimizations done. Sendind is esciacially bad, because the
> SACia doesn’t have any kind of send buffer, like it does for receive, so
> it gobbles up lots of CPU. Teh old, original, ACIAPAK actually had a send
> buffer. It would be great if some hacker got around to hacking it into
> SACia. I’ve mentioned it to many of the local CoCo hackers, and it
> seems no one has the time to do it.
> Hopefully, some more assembly optimization will get done on the
> CoCo version of Zmodem one of these days, and it will bring the
> throughput up to reality… even with one of the “mythical” CoCoIO type
> ports with the high speed 16550 uarts, zmodem as it is tops out just
> under 1000 CPS on receiveing… didn’t do any sending tests, but I
> imagine there isn’t much improvement there.
But ymodem (on Supercomm) sends files with an average speed of 400+CPS
(at 4800)… I thought Zmodem was supposed to be superior to ymodem.
Jim
87794 6-JUN 18:23 Telecom (6809)
RE: 9600 (Re: Msg 87784)
From: MITHELEN To: JEVESTAL
Sure, it isa superior _protocal_, but the current implementation on
the CoCo is not written as effeciently as existing CoCo ymodem routines…
Whats needed is for someone to write Zmodem routines for the CoCo from scratch,
that way it can be written cleanly from the start.
87796 6-JUN 21:02 Telecom (6809)
RE: 9600 (Re: Msg 87784)
From: JEJONES To: JEVESTAL
> But ymodem (on Supercomm) sends files with an average speed of 400+CPS
> (at 4800)… I thought Zmodem was supposed to be superior to ymodem.
Well…yes, but:
1. zmodem is a “sliding windows” protocol, which means that the receiver
has to be able to write the packet it just received to disk while the
next packet is coming in.
2. Like the original poster said, the author of rz/sz wrote it for Unix,
and it doesn’t do things one would definitely want to do under OS-9,
like notice how many bytes are ready to read and get them all at once
to minimize system call overhead. It would take some work to modify
or maintain rz/sz–it has many conditional compilation directives to
work on various flavors of Unix. A couple of places in the source
appear unaffected by the “structured programming” movement of the
late 60s and early 70s.
Opinions herein are solely those of their respective authors.
Clipper Chip: Big Brother Inside
87806 7-JUN 00:01 Telecom (6809)
RE: 9600 (Re: Msg 87796)
From: MITHELEN To: JEJONES
One correction James, Zmodem DOES check to see how many characters are
pending, and reads what is available, up to what it needs.
88187 24-JUN 21:18 Telecom (6809)
RE: 9600 (Re: Msg 87796)
From: JEVESTAL To: JEJONES
> > But ymodem (on Supercomm) sends files with an average speed of 400+CPS
> > (at 4800)… I thought Zmodem was supposed to be superior to ymodem.
>
> Well…yes, but:
>
> 1. zmodem is a “sliding windows” protocol, which means that the receiver
> has to be able to write the packet it just received to disk while the
> next packet is coming in.
>
> 2. Like the original poster said, the author of rz/sz wrote it for Unix,
> and it doesn’t do things one would definitely want to do under OS-9,
> like notice how many bytes are ready to read and get them all at once
> to minimize system call overhead. It would take some work to modify
> or maintain rz/sz–it has many conditional compilation directives to
> work on various flavors of Unix. A couple of places in the source
> appear unaffected by the “structured programming” movement of the
> late 60s and early 70s.
Thanks James. Like I told Paul in my reply to his message to me I’m
planning to send my spare CoCo to have the 6309 installed so that if
I get NitroOS9 I should have no problems use 9600 baud.
Later,
Jim
88186 24-JUN 21:17 Telecom (6809)
RE: 9600 (Re: Msg 87794)
From: JEVESTAL To: MITHELEN (NR)
> Sure, it isa superior _protocal_, but the current implementation on
> the CoCo is not written as effeciently as existing CoCo ymodem
> routines… Whats needed is for someone to write Zmodem routines for the
> CoCo from scratch,
> that way it can be written cleanly from the start.
Thanks Paul. I’ve decided that I’m going to send my spare CoCo off to have
the 6309 installed and then get NitroOS9 and that should allow me to
use 9600 baud without errors (I hope). Until then I’m satified with 4800
baud, even if uploads are only at 300 cps.
Jim