Pi 4B considerations.

[clasystems]

Apparently, in early production (and the poor initial availability) is the 4B Rasberry Pi, with an expected price of about US $120 for the complete "kit" and $55 bare-bones.

Will this be a relevant factor for any of the models?  (especially PiDP10)

cjl

[AndyB]

Fingers crossed the new Pi will allow me to easily add more than 4 serial terminals... I have many perfectly good serial-to-USB hubs that work fine on the PiDP8 and the PiDP11 ….but I can only connect only one four-port hub at a time (if I add more than one, it only recognizes only one of them.) I read extensively about it but had to give up in the end, something to do with the problems of enumeration.

My PiDP10 mainframe will need many users!! :-)

cheers

andy

[Tom Lake]

I have eight serial ports on a Pi 3 Model B+ based PiDP-11 so it's not a limitation of the Raspberry Pi itself. I use two USB to four serial port adapters. Maybe the SIMH implementation of the PDP-8 is faulty.

Tom L

[Warren Hardy]

Tom,

What adapters are you using? Do you have configure the pi to recognize the adapters or are they plug and play

[Tom Lake]

I use Lava 4-232 serial links. They are plug and play as far as the Pi is concerned. Of course you have to add the lines in the SIMH configuration file.

https://lavaports.com/products/usb-serial-links/usb-serial-link-4-232-db9-cabled/

;

; Add extra DZ11 terminal lines

;

; Linux: ser1=/dev/ttyUSB0 ser2=/dev/ttyUSB1 ser3=/dev/ttyUSB2 ser4=/dev/ttyUSB3

;            ser5=/dev/ttyUSB4 ser6=/dev/ttyUSB5 ser7=/dev/ttyUSB6 ser8=/dev/ttyUSB7

;

; Windows: ser1=COM1: ser2=COM2: ser3=COM3: ser4=COM4:

;                  ser5=COM5: ser6=COM6: ser7=COM7: ser8=COM8:

;

attach dz -V line=0,connect=ser1

attach dz -V line=1,connect=ser2

attach dz -V line=2,connect=ser3

attach dz -V line=3,connect=ser4

attach dz -V line=4,connect=ser5

attach dz -V line=5,connect=ser6

attach dz -V line=6,connect=ser7

attach dz -V line=7,connect=ser8

On the PiDP-8 change dz to be whatever multiplexer the PDP 8 uses.

Tom L

[AndyB]

Hey Tom - no other drivers required for the 2x lava adapters you used?  

cheers

Andy

[Tom Lake]

Nope. I plugged them into the USB ports, applied power to the Pi and there they were. The drivers are already in Raspbian, I guess. Anyway it was as simple as it could get.

Tom L

[AndyB]

Looks like these are now "legacy" products and not readily available... 

It's a weird thing: I have 2x startech 4-port USBs and 2x Keyspan 4-porters.  All of them work fine (plug and play) on PidP8 and PidP11... but only one at a time!

cheers

andy

[Tom Lake]

I have used other brands before as well as the Lava. Are they powered? If not, how much power do your units draw? Maybe together they tax the Pi power supply. Could you try a higher wattage power supply at the same voltage? When you plug in both at the same time, what do you see in /dev? Only the first four?

Tom L

[Tom Lake]

Here's a place that still says they have one

http://nextwarehouse.com/manufacturer/?2406_listall_page2

Maybe it would work with one of your other units.

Here's a place that sells an 8-port unit. That might be better.

https://www.usbgear.com/pages/GM-U28RS232.html?pk_campaign=googleshop&pk_kwd=&gclid=Cj0KCQjw6cHoBRDdARIsADiTTzYhAETxVjIWQboI3r0lMlkJIMqAXCGuZbS95NVo016MxHU9eL_1_8EaAhhSEALw_wcB

Tom L

On Monday, June 24, 2019 at 10:15:12 PM UTC-4, AndyB wrote:

[Tom Lake]

The StarTech ICUSB2328I 8-port USB to RS-232 adapter specifically says it's compatible with Linux.

https://www.startech.com/Cards-Adapters/Serial-Cards-Adapters/8-Port-Industrial-Wall-Mountable-USB-Serial-Adapter-Hub~ICUSB2328I

One way or another I'm going to find one that works for you!

Tom L

On Monday, June 24, 2019 at 1:11:00 PM UTC-4, clasystems wrote:

[AndyB]

Thanks Tom!  When I connect two of my 4-port hubs today (on PidP8 or PiDP11),  the /dev/ USB TTY's seem to load in Linux, but then fail in some way.. As you can see they all do seem to exist (at some level) in /dev/.   I don't really understand the info below, but I get similar results if I swap the order of the hubs around, or mix and match my startech and keyspan hubs, or if I try to use a regular USB hub to give me more individual USB serials. Net result: never more than 4 will work!   I read that recompiling the kernel or doing some work with GRUB is required which is way outside my skillset.  I'm pretty good at following instructions (tweaking SIMH settings and the various scripts like start.com) as needed, but when it comes to Linux I tend to stay in my plug-and-play comfort zone :-)

cheers

Andy

[AndyB]

This apparently works out-of-the-box Linux.. cheap too!

https://www.amazon.com/dp/B01NCYUYSV/ref=psdc_3015401011_t3_B009AT5TB2#customerReviews

[Digby R.S. Tarvin]

There may be other pertinent messages in the system log that are just not being selected by the ttyUSB filter. 

I'd redirect the output of dmesg into a file and then search for the ttyUSB error messages in an editor and post a snapshot of the messages surrounding one of the ftdi_set_termios errors.

I have seem similar things, for example, if enabling all of the ports would exceed the available USB power budget. (if that happens to be the problem, a powered hub or USB to serial with its own power suppyl as per AndyBs suggestion might be the next thing to try.

DigbyT

On Wed, 26 Jun 2019 at 17:24, AndyB <lordha...@gmail.com> wrote:

Thanks Tom!  When I connect two of my 4-port hubs today (on PidP8 or PiDP11),  the /dev/ USB TTY's seem to load in Linux, but then fail in some way.. As you can see they all do seem to exist (at some level) in /dev/.   I don't really understand the info below, but I get similar results if I swap the order of the hubs around, or mix and match my startech and keyspan hubs, or if I try to use a regular USB hub to give me more individual USB serials. Net result: never more than 4 will work!   I read that recompiling the kernel or doing some work with GRUB is required which is way outside my skillset.  I'm pretty good at following instructions (tweaking SIMH settings and the various scripts like start.com) as needed, but when it comes to Linux I tend to stay in my plug-and-play comfort zone :-)

cheers

Andy

 

On Monday, June 24, 2019 at 10:57:22 PM UTC-4, Tom Lake wrote:

The StarTech ICUSB2328I 8-port USB to RS-232 adapter specifically says it's compatible with Linux.

https://www.startech.com/Cards-Adapters/Serial-Cards-Adapters/8-Port-Industrial-Wall-Mountable-USB-Serial-Adapter-Hub~ICUSB2328I

One way or another I'm going to find one that works for you!

Tom L

--
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+un...@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-8/448eb212-1f2d-411a-b5f9-cc68713174a6%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[AndyB]

Hey Digby, regarding the power supply -  I have a (recent) Cana kit Pi adapter.  I also have power supplies for the serial hubs (same symptoms powered or unpowered)

Will see if I can figure out how to get messages out of the system log.

cheers

Andy

On Wednesday, June 26, 2019 at 1:00:34 PM UTC-4, Digby R.S. Tarvin wrote:

There may be other pertinent messages in the system log that are just not being selected by the ttyUSB filter. 

I'd redirect the output of dmesg into a file and then search for the ttyUSB error messages in an editor and post a snapshot of the messages surrounding one of the ftdi_set_termios errors.

I have seem similar things, for example, if enabling all of the ports would exceed the available USB power budget. (if that happens to be the problem, a powered hub or USB to serial with its own power suppyl as per AndyBs suggestion might be the next thing to try.

DigbyT

On Wed, 26 Jun 2019 at 17:24, AndyB <lordha...@gmail.com> wrote:

Thanks Tom!  When I connect two of my 4-port hubs today (on PidP8 or PiDP11),  the /dev/ USB TTY's seem to load in Linux, but then fail in some way.. As you can see they all do seem to exist (at some level) in /dev/.   I don't really understand the info below, but I get similar results if I swap the order of the hubs around, or mix and match my startech and keyspan hubs, or if I try to use a regular USB hub to give me more individual USB serials. Net result: never more than 4 will work!   I read that recompiling the kernel or doing some work with GRUB is required which is way outside my skillset.  I'm pretty good at following instructions (tweaking SIMH settings and the various scripts like start.com) as needed, but when it comes to Linux I tend to stay in my plug-and-play comfort zone :-)

cheers

Andy

 
On Monday, June 24, 2019 at 10:57:22 PM UTC-4, Tom Lake wrote:

The StarTech ICUSB2328I 8-port USB to RS-232 adapter specifically says it's compatible with Linux.

https://www.startech.com/Cards-Adapters/Serial-Cards-Adapters/8-Port-Industrial-Wall-Mountable-USB-Serial-Adapter-Hub~ICUSB2328I

One way or another I'm going to find one that works for you!

Tom L

--
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.

[Digby R.S. Tarvin]

On Wed, 26 Jun 2019 at 18:50, AndyB <lordha...@gmail.com> wrote:

Hey Digby, regarding the power supply -  I have a (recent) Cana kit Pi adapter.  I also have power supplies for the serial hubs (same symptoms powered or unpowered)

Will see if I can figure out how to get messages out of the system log.

I think that should be the next step.. No point talking about power adapters if the messages indicate a different issue..

Just do something like:

digbyt@PiDP11:/ $ dmesg >/tmp/messages

digbyt@PiDP11:/ $ grep -n ttyUSB /tmp/messages

1341:[609479.252189] usb 1-1.4: FTDI USB Serial Device converter now attached to ttyUSB0

digbyt@PiDP11:/ $ vi /tmp/messages -c 1341

ie create a file with all the messages in it, search for the same messages you searched for before but include line numbers of matches, then select the line of interest and open the editor, telling it to jump to the selected line...

DigbyT

 

[Steve Tockey]

Today's news about the Pi 4 and incorrectly designed USB-C. It doesn't seem like it's a major issue, and they claim it will be fixed in a future version, but maybe something to keep in mind.

https://arstechnica.com/gadgets/2019/07/raspberry-pi-4-uses-incorrect-usb-c-design-wont-work-with-some-chargers/

[John Mann]

Hi,

I tried a Pi 4B on a new PiDP-8 build.

Software-wise, as per https://tangentsoft.com/pidp8i/wiki?name=Home I had to start with Rasbian Buster, and then add pidp8i-v2019.04.25.zip

Then as per https://obsolescence.wixsite.com/obsolescence/2016-pidp-8-building-instructions I ran pidp8i-test.

The LEDs all lit up OK, but the switch test didn't work. 

- The default pattern was "0037 0037 0037". 

- None of the SR1..SR12 switches changed this pattern.

I temporarily borrowed a Pi 3B from my 3D printer setup, and the lights and switches tested OK with it.

So, now I'm thinking of buying a Raspberry Pi 3 Model A+ which has Quad core CPU, 512 MB, WiFi and one USB port.  

Doesn't have the extra USB and Ethernet ports the kit doesn't need.

Can still run the incandescent lamp simulator, and hopefully it won't be as hot in the enclosed case.

Thanks,

    John

[Warren Young]

On Saturday, August 3, 2019 at 1:27:54 AM UTC-6, John Mann wrote:

The LEDs all lit up OK, but the switch test didn't work. 

Please try the latest trunk per the standard update instructions, then report back whether it helps or not.

The latest change replaces some hard-coded logic in the GPIO common module with calls to routines provided by Broadcom, then redistributed by the Raspberry Pi Foundation in libraspberrypi-dev. The source code behind this doesn't appear to be materially different than what we had hard-coded, but maybe there's a fork of that library on Buster that solves the problem.

 

I temporarily borrowed a Pi 3B from my 3D printer setup, and the lights and switches tested OK with it.

Good test.

 

So, now I'm thinking of buying a Raspberry Pi 3 Model A+

I hope you keep trying with the Pi 4 until we solve this. Someone's gotta take the arrows in the back. :)

[Warren Young]

By the way, you might need to install libraspberrypi-dev on your system. It's installed on my binary OS images, but I don't know if the upstream Buster version includes it by default or not. I suspect it's pulled in as a dependency of the other build steps, so it isn't included by default. I've added a note about this to README.md just in case.

[Obsolescence]

Warren et al,

Apologies, I've been offline rather a lot this last week.

Regarding the Pi 4:

The problem with the Pi 4 is that it needs different code to activate its internal pull-ups (used for reading the switches by the PiDP circuit). Python's GPIO library was updated for this about a week ago, and thus the Python test program will work with Buster images updated since last week.

Fortunately the Python GPIO guys showed the code. See this link for the PiDP-11 (same thing in this case): https://groups.google.com/d/msg/pidp-11/tpSHLEK11ug/-6V0GNcpFgAJ

That is very helpful, as the datasheet for the new Broadcom chip in the Pi 4 is as yet unpublished AFAIK.

So the pull-up activation code in gpio.c needs an update for the Pi 4 as per the link above. I've no chance of doing it this week, but it's top of the list for the week after!

Kind regards,

Oscar.

[John Mann]

Warren,

> Please try the latest trunk per the standard update instructions, then report back whether it helps or not.

No, that didn't help.

I'm not sure if the Incandescent Lamp Simulator is working.

The build log says

"Driving PiDP-8/I front panel LEDs using incandescent lamp simulator."

But "pidp8i start" currently gives me a Pi load average of 0.04.

---

pi@pidp-8:~ $ pidp8i status

● pidp8i.service - PiDP-8/I Simulator

   Loaded: loaded (/home/pi/.local/share/systemd/user/pidp8i.service; enabled; vendor preset: enabled)

   Active: active (running) since Mon 2019-08-05 22:04:04 AEST; 4min 23s ago

     Docs: https://tangentsoft.com/pidp8i/doc/trunk/README.md#unit

 Main PID: 953 (screen)

   CGroup: /user.slice/user-1000.slice/us...@1000.service/pidp8i.service

           ├─953 SCREEN -dm -S pidp8i /opt/pidp8i/bin/pidp8i-sim /opt/pidp8i/share/boot/0.script

           └─957 /opt/pidp8i/bin/pidp8i-sim /opt/pidp8i/share/boot/0.script

Aug 05 22:04:04 pidp-8 systemd[511]: Starting PiDP-8/I Simulator...

Aug 05 22:04:04 pidp-8 pidp8i[938]: Booting from /opt/pidp8i/share/boot/0.script...

Aug 05 22:04:04 pidp-8 systemd[511]: Started PiDP-8/I Simulator.

---

Shouldn't there be another process using 100% of a core?

By the way, you might need to install libraspberrypi-dev on your system. It's installed on my binary OS images, but I don't know if the upstream Buster version includes it by default or not. I suspect it's pulled in as a dependency of the other build steps, so it isn't included by default. I've added a note about this to README.md just in case.

After installing Buster, one of the first things I did was an apt update; apt upgrade

Doing so updated a few packages, including

upgrade libraspberrypi-dev:armhf 1.20190709-1 to 1.20190718-1

I'll wait for Oscar to fix  gpio.c .

Thanks,

    John