7-bit ASR-33 TTY operation

[Francis Massen]

What changes are needed to accept a login to the Raspbian console as well to use SIMH using only capital letters?
My ASR-33 TTY (driven through ttyUSB0 by a USB-to-Serial converter at 110bd) refuses the login, as it does not find the small "i" letters in the user name? I read the discussion about the "naked board driving" and the "set console..." command in the SIMH config. Is there not one single switch to make work using only 7bit capital letters acceptable? If I use a laptop with TeraTERM set for 110 baud everything is ok (even if configured to 7 bit and 2 stop bits), as small letters seem acceptable...

[Warren Young]

On Monday, March 26, 2018 at 1:26:46 AM UTC-6, Francis Massen wrote:

What changes are needed to accept a login to the Raspbian console as well to use SIMH using only capital letters?

On old Unix systems, an all-caps user name at the login prompt was taken as a hint to the kernel's TTY driver code to assume a 7-bit all-caps terminal, so it would fold uppercase to lowercase.

However, I suspect this was never implemented on Linux, its lifetime having never overlapped that of 7-bit uppercase-only terminals. It was a dirty hack at the time, so why replicate it in a world where full 7-bit ASCII had been surpassed once already by ISO 8859 and such, and was about to be surpassed again by Unicode?

I've just tested an uppercase login on a FreeBSD system, and it doesn't work any more there, even though its kernel TTY code does in some way date back to the days of 7-bit uppercase-only terminals. They must have removed the feature somewhere along the way.

 

My ASR-33 TTY (driven through ttyUSB0 by a USB-to-Serial converter at 110bd)

[snip]

I read the discussion about the "naked board driving" and the "set console..." command in the SIMH config.

...and? Did you try it and fail, or have you just rejected the idea, and if so, why?

I've written up the method in the PiDP-8/I project wiki. The only downside of it that I can see is that you need a separate Linux console to do some things, but lacking 7-bit uppercase-only support in the Linux kernel at the TTY driver level, I don't see an easy way around it.

Well, I suppose someone could build a hardware "lowercaser" for RS-232, but that seems rather heavy-handed.

[david richards]

Hi Warren,

I am having problems connecting by serial USB in the way which I mentioned previously.

( set console SERIAL etc)

This is odd because It did work but now I can only see characters coming into my terminal but nothing seems to get into simh from it.

I think I need to check out how logging works to see if this will help identify the problem somehow.

Something I have discovered is that once a ttix or console is opened then it not not closed until a reboot occurs (or possibly a restart of simh, I'm not sure)

This looks like the cause of an error which occurs when the same mapping is re done on a subsequent simulation boot.

I haven't found a way to close a ttix or serial connection yet.

I can see they appear to stay pen after a simulation shuts down by using the show ttix, or console functions

I'm trying to set up a demo to boot from tapes and get a working console on a serial monitor.

I'm having all sorts of problems, I expect it was harder on a real machine but for different reasons.

Kind regards, David.

[Warren Young]

On Monday, March 26, 2018 at 3:53:30 PM UTC-6, david richards wrote:

I am having problems connecting by serial USB in the way which I mentioned previously.

( set console SERIAL etc)

That wasn't clear from your original post, because you talked about the "Raspbian Console", which implies the "console=" setting in /boot/cmdline.txt, which in turn suggests that /dev/ttyUSB0 is being used by the Linux console driver, not by SIMH.

So to be clear, the USB-to-serial adapter is only talking to SIMH here, not being used as the Linux console?

 

This is odd because It did work but now I can only see characters coming into my terminal but nothing seems to get into simh from it.

I suspect you haven't set the serial parameters correctly for the terminal, per section 3.14 in the main SIMH manual. Try this:

    set console serial=ttyUSB0;110-7M1

I'm guessing on that latter bit. It's based on past readings about how DEC would configure the ASR33s they re-sold as PDP terminals to use 7 real data bits and 1 "mark" parity bit (M) which was sometimes interpreted instead as 8N1 with the high bit always set.

You could also try setting it to 8N1, but then configure the PDP-8 TTI driver to mode 7P, which clears the high bit before delivering the character to the simulated PDP-8. See section 2.4.3 in the SIMH PDP-8 supplement.

If neither of those works, some RTFMing in the ASR33 and SIMH manuals may bring you to the correct parameters.

Please report when you get this working. You're probably not the only one who wants to try this.

[Rick Murphy]

Not to detract from Warren's attempt to help here, but you should also be aware that there's a bunch of USB serial adapters that can't do 110 baud. Also, you want two stop bits, not one. IIRC (7M2).

All of the "recent" devices talk multiples of 150 baud - 300, 600, 1200, 9600, etc. so the manufacturers simplify the clock generators to be multiples of 150 bits per second.

Test on the Pi to make sure that your adapter actually works at 110 baud, 2 stop bits, etc.

     -Rick

--
You received this message because you are subscribed to the Google Groups "PiDP-8" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pidp-8+unsubscribe@googlegroups.com.
To post to this group, send email to pid...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/pidp-8/8d6a93b9-c071-48a6-bd1a-43819af1ba22%40googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--

Rick Murphy, CISSP-ISSAP, K1MU/4, Annandale VA USA

[Francis Massen]

Rick, I am using a Prolific USB-to-Serial which works fine at 110, 2 stop bits. Driving the PI_ZeroW through it's only USB port using such a cable and at the RS232 end a null modem cable and an old laptop with real RS232 connector and TeraTerm as the communication software works absolutely fine. The problem is when I replace this laptop by a vintage ASR33 TTY I get the login /password messages, but the login is refused ("Incorrect login"). So in my opinion the problem really boils down to uppercase/lowercase, and not to any hardware...

Francis

[david richards]

Francis, could you change the password to be all upper case and get around the problem that way? David.

Edit;

I was under the impression that the user name was insensitive to case, I see this is not the case. I'm sorry.

[Francis Massen]

Hi Warren, thank you for your answer.

 I would suggest a possible easier solution: as the simulated OS8 runs ok with uppercase only, a possible solution without using the "set console..." in SIMH would be to autologin as user pidp8i, and than autostart  OS8 (which is done by typing the pidp8i command). I do not need to run the raspbian console, OS8 is all I ask for

I tried the autologin by changing the ExecStart line of /etc/systemd/system/serial...@ttyUSB0.service to:

ExecStart = -/sbin/agetty --autologin pidp8i 110 vt100

which runs fine with an old terminal (with real RS232 port) and Teraterm (110 bit/s, 7 bit) I guess agetty figures out what the number of stop bits and the parity should be.

As my Linux and systemd competences are rather on the low side, I was unable to define the autostart of pidp8i (which I guess must be called by an ExecStart = ~/pidp8i/bin/pidp8i) in a new service like startpidp8i.service located in /etc/systemd/system/

I would really appreciate your help for this ....

Francis

[Warren Young]

On Tuesday, March 27, 2018 at 1:59:01 PM UTC-6, Francis Massen wrote:

ExecStart = -/sbin/agetty --autologin pidp8i 110 vt100

I've filed this idea, with some refinements, as a new feature request.

We've already got an equivalent method lined out on the wiki, however. Since this scheme is a redundant way to do something I don't recommend in the first place, I've set the priority to "Low." That means I'm unlikely to implement this at all.

I suggest that you get past your current systemd limitations and finish the implementation yourself. If you provide it as a patch, I may include it as a disabled-by-default feature in the next release, as an alternative to the current method lined out in the wiki.

[Warren Young]

On Tuesday, March 27, 2018 at 9:35:11 AM UTC-6, david richards wrote:

Francis, could you change the password to be all upper case and get around the problem that way? David.

Edit;

I was under the impression that the user name was insensitive to case, I see this is not the case. I'm sorry.

Even if you made the user name and password all-uppercase, then you'd need a way to attach to the GNU screen session hosting the simulator, so now you'd need something like an all-caps uppercase alias for the `pidp8i` command.

No, the "right way," insofar as there is there is such a thing in 2018, to use an all-uppercase terminal with Linux would be to reimplement that old TTY layer hack from the early days of Unix. That at least would get you a way to do all-lowercase things on the Linux side.

Of course, then you run into problems *again* if you wanted to give a command involving, say, /dev/ttyUSB0. That's not the same thing as /dev/ttyusb0. Linux doesn't have a large number of commands and files using mixed case, but they're common enough. The PiDP-8/I software distribution includes several.

There are ways to get a case-insensitive filesystem on Linux, but that's a battle all by itself.

I'll stick with my recommendation that you should separate the Linux console from the simulator, sidestepping all of these problems.

[Warren Young]

On Tuesday, March 27, 2018 at 6:24:48 AM UTC-6, Francis Massen wrote:

The problem is when I replace this laptop by a vintage ASR33 TTY I get the login /password messages, but the login is refused ("Incorrect login").

It is as I originally thought: you have the Linux console on the serial line already, so that you need to go through that to get to the simulator.

Instead, you should disable the Raspbian serial console mechanism, which will then free up /dev/ttyUSB0 to be used by the simulator.

As you have it now, when the simulator comes up near the end of the boot sequence, ttyUSB0 is already grabbed by the kernel, so your "set console serial" directive cannot be obeyed.

If you really need a Linux serial console as well, I'd suggest adding a second USB-to-serial adapter. (Yes, I realize this means adding a hub; an un-powered hub may suffice here.) Then let Linux grab ttyUSB0 for itself, leaving ttyUSB1 for the ASR33.

[Francis Massen]

Thanks to all for good hints...I have everything working now in the simplest way:

1. autologin added to the "agetty..." command

2. defined alias in uppercase ("PDP as an alias for pidp8i")

Now when I switch on everything, the TTY shows the rasbian console messages and finishes with the usual prompt.

I just type "PDP" on the TTY and am taken by SIMH into OS/8, which accepts uppercase characters

[Warren Young]

On Sunday, April 8, 2018 at 2:28:44 PM UTC-6, Francis Massen wrote:

Thanks to all for good hints...I have everything working now in the simplest way:

1. autologin added to the "agetty..." command

Is the agetty command identical to what you posted before, or did you have to modify it to work with the 7-bit + mark parity of the ASR-33?

2. defined alias in uppercase ("PDP as an alias for pidp8i")

I've modified the script so that it lowercases its first argument, so you can now give commands like "PDP RESTART" or "PDP STOP".

 

Now when I switch on everything, the TTY shows the rasbian console messages and finishes with the usual prompt.

Let's be clear here: The Linux console is still /dev/ttyUSB0? How many pages of output do you get on the terminal before you get to a Linux command prompt?

If this is still the case, I'd suggest disabling the Raspbian serial console, if only to save paper. I can think of at three alternatives for the Linux console just offhand:

1. HDMI monitor + USB keyboard
2. Second serial port
3. SSH

The last option is only of use after the system boots, but it's what I do all the time with my PiDP-8/I. I've never had to take option 1 or 2.

[Francis Massen]

1. agetty login: yes the command is the same as posted before

2. Yes the Linux console is /dev/ttyUSB0

Raspbian/Debian text at startup: yes I get quite some lines of the usual console messages, but IMHO this is quite nice from a vintage point of view (even if half a page of paper is consumed)

3. Thanks for the script modifications... are they included in the "ready-to-use" SD image ?

I have assembled the TTY+PdDP8/i on a moving stand to be displayed at our Computarium, so what really matters is that people can hear the rattling of the TTY, see OS/2 starting and all LEDs of the kit blinking, load BASIC and run a few lines of code.

4. Are paper rolls with guiding holes in the format needed by the TTY ASR-33 still available? Here in Europe I did not find anything...

Many thx for all answers and your good work!

Francis

[William Cattey]

Normally an ASR33 is friction feed, so you should not need the guiding holes for tractor feed.

Indeed, I've never seen an ASR33 with the necesessary sprockets for tractor fed paper.

[clasystems]

I've been told there is a tractor feed option for the 33.  We had it standard on a KSR-35

cjl

[AndyB]

Hi Francis, I have an off-topic question  -hope that's ok!

I recently got my hands on a working ASR-33.  I've spent the last week or so cleaning it up, lubricating, greasing etc.  I do have a Black Box current loop converter. I was wondering if you could share how you connect your Teletype to your serial connection?  I have a long cable coming out the teletype with 6 wires (red/orange/black/blue/green/yellow).  Reading around I notice just four wires used for 20ma loop. If I was to hook these up to my current loop converter I am concerned that I am likely to break something due to my total ignorance! 

If you have any tips or a configuration you have successfully used that you can share I'd be most grateful,

thanks in advance

Andy

On Monday, March 26, 2018 at 3:26:46 AM UTC-4, Francis Massen wrote:

[CLASystems]

Mentioning of ASR terminals.  Anyone intending to actually read paper-tapes?  That requires the reader-run modification on the teletype and support for it on several levels.  Otherwise, it's just a cute way to print slowly and inauthetically on the whole.

The DEC modifications cause the reader to NOT run even if enabled with a tape in the reader.  Power clear or execution of 6032 or 6036 releases the reader for one character.  That way, the software controls the reader done originaly in an era where many other vendors had problems.  It became a standard common "stunt box" option, but you need the entire change of control for iit to work.

In programs such as FOCAL, 1969 without modifications AND having interrupts enabled, this could be a problem because you cannot software prevent the next interrupt; clearing the interrupt causes yet another character to be read, etc.  This problem was addressed in the PDP-8/E and there are some software patches around that fix the problem [but only on the KL8 family of interfaces and newer.  Since the PiDP-8 emulates the PDP-8/E it is fair game to implement it.  Thus, even FOCAL can work properly with the small software patch, etc.

The change is the new instruction 6030 which is JUST clear the keyboard flag.  If the flag is clear but the reader-run was not set, the reader stops.  When the program is again ready to accept a character, the 6032 is executed normally and that sets the reader-run, etc. as is normal on the older machines, etc.

cjl

On Tue, Apr 24, 2018 at 8:37 AM, AndyB <lordha...@gmail.com> wrote:

Hi Francis, I have an off-topic question so I thought I'd take the liberty of contacting you directly -hope that's ok!

I recently got my hands on a working ASR-33.  I've spent the last week or so cleaning it up, lubricating, greasing etc.  I do have a Black Box current loop converter. I was wondering if you could share how you connect your Teletype to your serial connection?  I have a long cable coming out the teletype with 6 wires (red/orange/black/blue/green/yellow).  If I was to hook these up to my current loop converter I am concerned that I am likely to break something due to my total ignorance! 

If you have any tips or a configuration you have successfully used that you can share I'd be most grateful,

thanks in advance

Andy Barnett

On Monday, March 26, 2018 at 3:26:46 AM UTC-4, Francis Massen wrote:

What changes are needed to accept a login to the Raspbian console as well to use SIMH using only capital letters?
My ASR-33 TTY (driven through ttyUSB0 by a USB-to-Serial converter at 110bd) refuses the login, as it does not find the small "i" letters in the user name? I read the discussion about the "naked board driving" and the "set console..." command in the SIMH config. Is there not one single switch to make work using only 7bit capital letters acceptable? If I use a laptop with TeraTERM set for 110 baud everything is ok (even if configured to 7 bit and 2 stop bits), as small letters seem acceptable...

--
You received this message because you are subscribed to a topic in the Google Groups "PiDP-8" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/pidp-8/XQyKu7zSSVA/unsubscribe.
To unsubscribe from this group and all its topics, send an email to pidp-8+unsubscribe@googlegroups.com.

To post to this group, send email to pid...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/pidp-8/2dfa0276-1f12-41fe-983b-759671ca46c5%40googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--

"In the future, OS/2 will be on everyone's desktop"

Bill Gates, 1992

[Francis Massen]

Reply to Andy: how to connect the TTY to a current loop converter

Sorry Andy, but I have no good answer: what is sure is that you need only 4 wires: the first couple being the Read_from_TTY and the second being the Write_to_the_TTY. The last is actually a current loop driving the "magnet" inside the TTY. In my PiDP8i setup I use a 3-transistor board given to me, but I do not know the exact wiring.

On a second ASR-33 I use a RS/20mA loop converter built as a kit. Look here for the scheme:

http://computarium.lcd.lu/COMPUTERMUSEUM/files_and_photos/ASR-33/Horter_V24TTY_for_ASR33.jpg

( just look at http://computarium.lcd.lu/works/index.html  and go down to May2011)

The letters at the right side (3,4,6,7) correspond to the terminal strip inside the TTY (not the Molex connector). You notice that one loop (R+) gets the positive current (20mA) and that the other (T-) the GND.

I do not think that you can do much damage by goofing: best would be to find the couple of wires that stop the "chatter" of the TTY when no current loop is connected. This tells you that these two are ok.

------------------------------------------ 

On Monday, March 26, 2018 at 7:26:46 AM UTC, Francis Massen wrote:

[Francis Massen]

My Teleprint model of the ASR-33 has a tractor: I now use normal metric fan-fold paper (A4) with sprocket guides and cut down the right side at 203mm from the very left border (left to the left holed strip). Thus the paper is still guided by the sprockets at the left, and the remaining friction is enough to avoid it going skewed.

On Monday, March 26, 2018 at 7:26:46 AM UTC, Francis Massen wrote:

[AndyB]

thank you !

However I sadly report that my ebay Teletype optimism has been dashed somewhat...  

I finally powered up yesterday and unfortunately it just chatters in LINE or LOCAL.  I've read various posts (the "clothespeg solution" etc.) for chattering.  Keyboard is also "solid."  I have manuals so I am working through.  I probably put the keyboard back wrong. Frankly I am petrified to do a full disassemble (I did enough to get to the mail oil/grease points) but the machine looks very intact and clean.  

Here it is, chattering it's little heart out in LOCAL mode.  https://www.youtube.com/watch?v=MEgqg-TTpVA  

If anyone has a diagnostics reference they can share I'd appreciate it.  I have manuals that show assembly/disassembly/theory of operations etc but I guess I need an Idiots Guide as my enthusiasm far outranks my technical skills :-)

cheers!

Andy

[Neil Higgins]

Andy, fear not.

I completely disassembled and reassembled a Model 15 teletype, and it wasn't as bad as you might think. Even though an ASR33 is a 7-bit device (as opposed to 5 bits for a Model 15), I found as follows:
a) You learn as you go - what does what, and how it works
b) The parts are pretty much self-describing and hard to mix up
c) If you look at each part carefully as you go, you'll quickly spot the problem(s). For my Model 15, it was a tiny lip worn into a lever in the start-bit release mechanism
d) These things are "built like brick out-houses" with very robust parts, tapped holes (no self-tappers in those days), etc. You can play all day and not damage anything
e) With the Model 15, even though there were a few tweak points (e.g. beautifully machined eccentric collets), the thing seemed to work regardless of how I set it up

One warning though: It's best to disconnect and remove the motor first. That way, there's nothing that can hurt you - just turn the shafts by hand.

Work on the keyboard or the printer, not both - the only thing they share is motive power.

It's easy to design an RS-232 or TTL to 10mA current source. It can be done with one transistor, one Zener diode and one resistor: Google "current source".

The thing that most people don't realise is that you need at least 50Vdc to drive the 10mA loop (the more the merrier, until potentially lethal at about 120Vdc) to get quick enough rise and fall times for the printer decoder solenoid. If that solenoid is not energised, the printer will just keep churning and doing nothing - the quiescent state is a "mark" (10mA) and the start bit is a "space" (0mA).

Dive in and have a go.

[AndyB]

Neil- thanks for the inspiring words. I will be diving in! Have to say that the mechanisms are a lot of fun to simply gaze at.  So much diversity and variation.  Every single system in the machine is quite an engineering feat in it's own right. Fairly hopeful I won't electrocute myself or other family members :-)

[Neil Higgins]

I’d be more worried about fingers and hair being ripped out. Take that motor out!

And if you are playing in a.c. circuits, an isolation transformer is a good investment.

[Jack Rubin]

And don't disassemble the '33 keyboard unless you have a really good reason to do so and know what you're doing. Lots of little springs are inside, waiting to jump out and disappear!

[Audin Malmin]

The 33 will chatter if there isn't any loop current. Mine came in the same state: it would chatter in both modes. The local loop supply was broken due to a bad capacitor and I didn't have an external supply hooked up.

Also of note is that the 33 doesn't need a high voltage supply because it has a transistor magnet driver. So 5 volts is enough to drive the typing unit. A somewhat higher voltage (I'm using 9v) is needed for the keyboard half as the contacts aren't really good enough for lower voltage.

[Neil Higgins]

Yes, it’s a 20mA loop, not 10mA as I said above. Accounts of the drive voltage vary. If 5V works, well and good!

[AndyB]

thx Neil for guidance, also about the isolation transformer.  As I continue to divert us yet further from the main purpose of this group, I wonder can you clarify something for me? I read an Article on hackaday a while back about those transformers.  And perhaps now is the time to get one (Death By Teletype is after all a little un-glamorous) but when I do some Amazoning these transformers call out clean AC etc but not all mention safety.  Do you have a particular brand/model you'd care to recommend?  

thanks!

andy

On Friday, April 27, 2018 at 4:47:34 PM UTC-4, Neil Higgins wrote:

[AndyB]

Hi Audin- my capacitors look visually intact and not bulging, although I cannot quite see all of them easily without disassembly. Do you happen to recall which one was the culprit?

thanks, Andy

On Saturday, April 28, 2018 at 2:17:47 PM UTC-4, Audin Malmin wrote:

[AndyB]

Yeah you've got me convinced. I'm definitely not going to do that.  Got enough things disappearing around me as it is :-)

[Neil Higgins]

Andy: Not sure where you live (I’m in Australia, so brands and house voltages my be different).

The key features you will require are:
1. The right input (house) voltage and output (TTY) voltage
2. An adequate V.A (power) rating
3. An UNEARTHED (floating) output side, and most importantly
4. GALVANIC ISOLATION between the two sides of about 1000V, i.e. NOT an autotransformer (the latter are cheaper but unsuitable)

According to Wikipedia (the source of all knowledge), “This machine requires less than 4 amperes at 115 VAC 60 Hz,” so ypu’ll need a 460V.A (say 500V.A) or bigger tranny.

It should also preferably have pre-wired input cord and an output socket, so there is exposed a.c. voltage.

In the U.S. a company like Jameco should have them.

Cheers, Neil

[Neil Higgins]

Correction: “ so there is NO exposed a.c. voltage.”

[AndyB]

awesome thx Neil.  

[Bob P]

Has anyone successfully connected an ASR-33 to a PiDP as a console via a USB port?  I see the discussion on uppercase versus lower case from last March, and I understand that may preclude using the Teletype as a console.  What about as a line printer?  Has the line printer output function been implemented in OS/8?  If so, can you tell me how to set it up?  Using the Teletype as a printer only would avoid the upper/lower case problem.