UniBone Kit Arrived!

[Kevin McQuiggin]

Hi All:

Just FYI, my UniBone kit arrived, so the build is in progress.  

I will post comments as I build, test and then install the board.

Thanks to Joerg for the 2025-02 board update!

Kevin

[Ethan Dicks]

What are the updates for 2025-02? I looked back on the mailing list
archive and didn't see an announcement around Feb 2025.

(I have older boards, probably from the first or second run. Curious
what the change are).

Cheers,

-ethan

[Joerg Hoppe]

Hi,

> What are the updates for 2025-02? I looked back on the mailing list
> archive and didn't see an announcement around Feb 2025.
>
> (I have older boards, probably from the first or second run. Curious
> what the change are).

All boards are functionally identically.
PCB layout changes over the years:
- introduction of the fixed Beaglebone spacers.
- gradually replaced "Through the hole" parts by SMTs,
- removed the I2C EEPROM address jumpers.
- Make the "BUS-ON" LED controllable by software,
- Moved LEDs around,
- cleanup the NPG/BG jumpers
- added a 2x5 pin plug for the status-LEDs and mode-switches
- alternate sockets for both AM26S10 and 8641 bus drivers.

kind regards,

Joerg

[Kevin McQuiggin]

Hi All:

Here’s a progress report on my Unibone.  I’m a new user and this is a work in progress.

I got the kit together with just about no problems.  The hardest part of the build is soldering of the two little circuit boards that hold the male headers that the BBB mounts on, as alluded to by Joerg on his retrocmp web site.

My Unibone passed the initial standalone tests, but I am still working on the “Acceptance Test” portion of the project.  The problems I am seeing are not with the Unibone itself, but with the old DD11-DK backplane that I am using for loopback testing.  My backplane has some issues, likely because it has been out of service for many years, so I am working to clean it.  “De-oxit Gold” has been deployed, and I am going to more closely inspect the wirewrap side of the backplane.

I am now getting great results except for the Unibus SACK signal, and I expect that this is due to some dirt in a connector or otherwise bad in the backplane.  

I should add that Joerg has been VERY helpful in answering my questions as I went into the "testing on a backplane” phase.  I have a systems software/hardware background but have not done low level bus testing in a DEC/Unibus environment.  

Joerg’s “demo.sh” application is comprehensive has all the options that you need to do complex testing.  It does not have a fancy user interface but everything is there for a technically knowledgable user.

You may need to deploy a logic analyzer (mine is a small Saleaea USB device), but there’s a lot that you can do via the testing components of the application itself.  The “TL” (test Unibus lines) sub-application has many helpful options, so you need to read the menus and look at all of the sub-commands to see all the tests you can do.  The retrocmp web site just covers the basics.  Do not underestimate the time you’ll need to get your backplane kinks worked out.

I am working to get my backplane in order, and only then will move on to testing a memory board (core set, in my case), emulating missing memory, and lastly booting RT11 and XXDP.  I have another backplane I can test with but want to get the current one working first if at all possible.

My impetus for getting a Unibone was my sick 11/05 that is functional except for a small portion of core.  XXDP will run but I have had no success in getting RT11 or other (bigger) packages running.  pdp-11 BASIC does run, however, at least until it hits a memory address around 40000, my flaky region.

It’s a fun project-in-progress, and I just thought I’d post a status report.

Kevin McQuiggin, in Canada

[Frits Jalvingh]

Hi Kevin,

Nice writeup! I found that having an Unibone that had a bit of trouble was not too bad; it actually helped with understanding the thing. There is really a lot to learn about it; Joerg made an amazing tool for us all that can do so many things...

Have fun with your project!

Regards,

Frits, from the Netherlands and France ;)

Sent with Proton Mail secure email.

--
You received this message because you are subscribed to the Google Groups "UniBone" group.
To unsubscribe from this group and stop receiving emails from it, send an email to unibone+u...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/unibone/e4d87601-5966-426c-937f-4d2b93c9cf76n%40googlegroups.com.

[Kevin McQuiggin]

Hi All:

A short update as promised.  The Unibone is now fully functional in my pdp-11/05.  The system has booted from an emulated RX01, an emulated RL02, and most recently from my physical RX11/RX01 floppy drive.  

My 32KB core set (16K words) has a small number of intermittent problems between address (octal) 40000 and 40100, but I was able to use the Unibone to emulate memory above 40000, and this sidestepped the bad core and gave me a fully functional address space.  Top address via emulation is 157776 (octal), the full 28 KW that is supported by the 11/05.

I then took a second approach and simply removed the four boards in the core set (core mat, driver, sense/inhibit, and the M8293 timing board) and then set up Unibone to emulate the entire memory address range 00000 through 157776.  This worked as well so I am a very happy user.

I had a few more issues with the backplane (apparent stuck bits when addresses were accessed from the M9312 monitor), but with some more Deoxit and use of a two-finger extender card on the backplane to exercise all the pin contacts, I got good connections from the Unibone fingers to the backplane's C - F slots, and memory read correctly from both the Unibone side (in the 'D' utility) and on the console.  

As an example, I used the 'TA' command in the 'D' utility to deposit the relevant address into every byte in memory.  Bit 7 would "stick", so that if I viewed location 00000 using the M9312 console, it did not contain 00000 but rather 00100.  Location 4000 would apparently contain 04100 rather than 04000.  Viewing these locations within the Unibone utility would show the correct contents.  

I thought that clearly this was likely a backplane connection issue.  I swabbed Deoxit on both sides of the two fingers of the extender card and worked it into all the slots of the entire backplane.  This took awhile, but it seems to have cleaned all the individual connectors.

After this all was well and (fingers crossed) will continue to be nominal.

Further comments: for new users, it takes a bit of time to get used to the demo.sh application and its sub-menus and commands, but you can learn very much by just reading the .sh files provided to see how the command interpreter and the sub-utilities work.  

My 11/05 has been idle, in terms of booting into and running an OS, for at least 25 years so it is fantastic that the machine is functional again.  I rescued it from the dumpster (actually, the pallet that was heading to the dumpster) at my university around 1995 or 2000, and it sat in my basement until about 6 months ago when I finally decided to try to get it running fully again. 

I had previously toggled in some little programs and had them run, so I knew that machine was basically functional, but never had success with trying to get it to run an OS like RT11.  Nice to have it running a real OS now, but it did take a lot of work to get it to this point.

Now I need to think about how to use the machine.  I would like to integrate it into some of my other hobbies somehow, just for the novelty effect.  I am an amateur radio operator (VE7ZD).  For example, it would be fun to interface the 11/05 to my transceiver or with a popular digital mode and make international contacts using it.  Odd and anachronistic, but fun nonetheless.

Kevin (VE7ZD)

[Mark Matlock]

Kevin,

   Thanks for the update on your PDP-11/05. The UniBone really makes a great tool to work with the 11/05. 

   I recently ran across a RSX11M V4.2 unmapped baseline distribution that runs in 28KW. Back in the day I worked a bit with RSX11S on diskless 11/23s loading 11S via DECnet over a DUV-11 connected to an 11/44 with a DV11. Running RSX11M in 28KW from the RL01 distribution was interesting. 

   All of this unmapped RSX executive manages to fit into 14KW leaving a 14KW partition for user programs. I wondered if it was even possible to build a task with such a minimal system. The only editor was EDI which actually was not that difficult to use. Macro-11 compiled a simple program and amazingly TKB linked it. To get more disk space I used RSX running on simh to BRU the RL01 image to a RL02 and that booted fine. It only had two RL drives sysgenned so making them bigger is important.

  On the topic of doing something ham radio related with your 11/05, you might be interested in Lee Gleason’s Morse code generating program for the PiDP-11/70.

  I think it could be adapted to RT-11 without too much trouble.

Best,

Mark

To view this discussion visit https://groups.google.com/d/msgid/unibone/82125f15-5769-4c3a-97da-96f506609266n%40googlegroups.com.

[Henk Gooijen]

Hi Kevin,
if you can build a little hardware, like an (A)FSK tone decoder, you could write an RTTY decoder (and encoder, if you add an AFSK generator like the XR2206). Might be a lot of fun, because you can add all kind of "goodies", like UnshiftOnSpace, mark/space inversion, bit inversion, and some storage for prepared messages (RYRYRY, QBF, CQ, etc) options.

If you build a tone decoder based on filters, you can send the analog mark and space filter output amplitudes to an AR11, and write code to display the tuning lissajoux circle/elipses on an oscilloscoop (set to X-Y display operation).

Sending morse code is of course possible using a digital I/O option to connect a key or even a paddle and a digital output (one to key the transceiver, one to generate a tone output to listen to, but decoding (sloppy keyed) morse code is not trivial. The human's ear and brain is still very good at that, but sound card interfaces and DSP have come a long way in the past 10 years or so.

Writing a logbook with a database behind it would be an other option ... When you are an active HAM, you could add rotator control. For example, based on a heard prefix rotate your antenna in that direction (probably not an option on shortwave, hi).
Henk, PA8PDP

[Kevin McQuiggin]

Hi Henk:

Thanks for the reply and sorry for the delay.  Busy week here!

I have an old RTTY modem that I built when I was about 20.  It uses an XR2211, I recall, and worked quite well on my PCjr in those days.  My rig was a Yaesu FT-101Z.  I wrote the software in Pascal, which was the popular language for the PC set in those days, mostly because Turbo Pascal was cheap: about C$25 versus the “real” compilers from IBM et cetera that were on the order of C$200 or more.  

Morse code usage would also be fun, but I am not really active on that mode these days other than occasional use.  I’m fully into modern digital modes and active on 6m and 23 cm EME these days.

I disagree re CW decoding, it is not that hard and one can build an adaptive decoder that tracks CW input as code speed changes, and that can compensate for all but the very worst “fists”.  A lot of CW decoder developers focus on absolute timing and adjustment, but what works best is a fuzzy method that uses the average difference in timing between a “dit’ and a “dah”.  

Absolute timing decoders are hard, but the adaptive approach can track CW across widely varying inputs from 5 WPM to 30-35 WPM automatically.  My decoders (first wrote one about 1980, latest a couple of years ago for the article below) track input CW speed changes and catch up within about 8 to 16 code elements.  I wrote a two part article about the process for an online amateur radio magazine at https://archive.org/details/SARC_Communicator_2024-05/page/22/mode/2up.  

Back to the 11/05, it would not be too hard to write an adaptive decoder for the machine.  THAT might be a fun project!  I could use the GPIO pins on the Unibone, mapped to a UNIBUS address as per Jorge’s example at RETROCMP, to gain access to the input CW stream.

73 (best wishes, to those non-hams on the list)

Kevin VE7ZD (also KN7Q)

[Kevin McQuiggin]

Hi Mark:

Thanks for the reply, sorry for my delay in response.  I checked out RSX11M for my 11/05 and just found out that the common versions (RSX11M for example) won’t run in 28 KW.  

Would you have a pointer to the V4.3 version that *will* run in 28 KW, or the smaller RSX11S version?  I would like to give RSX11 a try.  If so, thank you!

I read Lee’s blog post as well, fun project!!

Kevin

[Mark Matlock]

Kevin,

   The common versions of RSX11M required mapped (18 bit+) PDP-11s. The version of RSX11M you need is for unmapped (28KW) PDP-11s. 

    Attached is a zip file has the Baseline Unmapped RSX11M V4.0 

    In the .zip is the RSX11M V4.0 unmapped distribution on RL01s, but I’ve transferred the bootable RSX11M system disk image to a RL02 to provide more free disk space. There is also a sim-RSX28K.ini in case you want to test it with Open Simh.

   Below is an example boot from Open Simh. I was amazed that I could create a Macro-11 program with EDI, then MAC and TKB it in the 14KW user memory (PAR14K) to create a RSX task (HELLO.TSK).

Best,

Mark

sim> b rl0

  RSX-11M V4.0 BL32   28.K (BASELINE)

>RED DL:=SY:

>RED DL:=LB:

>MOU DL:RSXM32

>@DL:[1,2]STARTUP

>* PLEASE ENTER TIME AND DATE (HR:MN DD-MMM-YY) [S D:"8:00 12-JUL-86"]: 

>TIM 8:00 12-JUL-86

>* ENTER LINE WIDTH OF THIS TERMINAL [D D:132.]: 

>SET /BUF=TI:132.

>INS $PIP

>INS $EDI

>INS $MAC

>INS $TKB

>MOU DL1:/OVR/PUB/VI

Volume Information

Class:       Files-11

Device:      DL01

Volume label:EXCPRV

Pack serial: 00000010123

Owner:       [1,1]

Protection:  [RWCD,RWCD,RWCD,RWCD]

Default:     [RWED,RWED,RWED,R]

Processor:   F11ACP

>@ <EOF>

>DEV

DL0: Mounted Loaded Type=RL02

DL1: Mounted Loaded Type=RL01

TT0:

NL0:

TI0:

CO0:    TT0:

CL0:    TT0:

LB0:    DL0:

SY0:    DL0:

>pip /fr

DL0: has 13412. blocks free, 7068. blocks used out of 20480.

Largest contiguous space = 13412. blocks

1349. file headers are free, 187. headers used out of 1536.

>pip dl1:/fr

DL1: has 1088. blocks free, 9152. blocks used out of 10240.

Largest contiguous space = 1088. blocks

1291. file headers are free, 245. headers used out of 1536.

>PAR

LDR    023300 023164 000000 MAIN TASK

PAR14K 037744 040000 070000 MAIN TASK

GEN    037710 040000 040000 SUB  TASK

SYSPAR 037654 130000 012000 MAIN TASK

DSKPAR 037620 142000 003000 MAIN TASK

SECPAR 037564 145000 004200 MAIN TASK

>SET /POOL

POOL=400:02382.:02478.

>TAS

. LDR.        LDR    248. 000000 LB0:-00000000 FIXED

TKTN   04.8   SYSPAR 248. 012000 LB0:-00005403 

...DMO 23.26  GEN    160. 040000 LB0:-00002721 

MCR... 2.02   SYSPAR 160. 012000 LB0:-00004335 

...MOU 2503   GEN    160. 040000 LB0:-00004511 

F11ACP M0401  SYSPAR 149. 012000 LB0:-00003070 CHECKPOINTED

...INS 06     GEN    100. 040000 LB0:-00003776 

...SAV 2.18   GEN    100. 040000 LB0:-00004754 

...EDI M11.04 GEN     65. 040000 LB0:-00002761 

...AT. 02     PAR14K  64. 070000 LB0:-00003510 

...LOA 03.3   GEN     50. 040000 LB0:-00004144 

...UNL 3.0    GEN     50. 040000 LB0:-00005423 

...PIP M1344  GEN     50. 040000 LB0:-00004603 

...MAC M1200  PAR14K  50. 070000 LB0:-00004203 

...TKB M40.02 PAR14K  50. 070000 LB0:-00005114 

>ATL

. LDR. 023332  LDR    023300 023164-023164  Pri - 248.  Dpri - 248.

   Status:  -CHK FXD STP PRV NSD

   TI - CO0:  IOC - 0.  BIO - 0.  Eflg - 000001 000000  PS - 000000

   PC - 021702 Regs 0-6 021632 023164 000000 036746 037654 036714 023256

MCR... 036714  SYSPAR 037654 130000-142000  Pri - 160.  Dpri - 160.

   Status:  CKD -PMD PRV CLI NSD CAL

   TI - TT0:  IOC - 0.  BIO - 0.  Eflg - 000001 040000  PS - 000001

   PC - 136410 Regs 0-6 000000 130476 133260 130432 177777 130412 130362

F11ACP 037234  SYSPAR 037654 130000-142000  Pri - 149.  Dpri - 149.

   Status:  OUT CKP STP ACP -PMD PRV NSD CAL

   TI - CO0:  IOC - 0.  BIO - 0.  Eflg - 000002 040002

>PIP /LI

Directory DL0:[200,200]

12-JUL-86 08:02

BLDLAINIT.CMD;1     14.        22-JAN-82 10:19

SGNBLDDRV.CMD;1     19.        22-JAN-82 10:19

SGNKLAB.CMD;1       70.        22-JAN-82 10:19

SGNSTAND.CMD;1      4.         22-JAN-82 10:19

SYSGEN.CMD;1        85.        22-JAN-82 10:19

SYSGEN2.CMD;1       79.        22-JAN-82 10:19

SYSGEN3.CMD;1       52.        22-JAN-82 10:19

CLR.TSK;2           4.      C  10-JUL-86 20:03

HELLO.MAC;4         1.         20-JUN-86 17:01

HELLO.OBJ;5         1.         20-JUN-86 17:02

HELLO.LST;3         2.         20-JUN-86 17:02

HELLO.TSK;1         4.      C  20-JUN-86 17:02

CLR.MAC;3           1.         20-JUN-86 17:05

CLR.OBJ;1           1.         20-JUN-86 17:17

HELLO.MAP;1         3.         20-JUN-86 17:02

Total of 340./340. blocks in 15. files

>

[Kevin McQuiggin]

Hi Mark:

THANK YOU!

Kevin

<rsx28k.zip>

To view this discussion visit https://groups.google.com/d/msgid/unibone/a40200e8-1779-44c9-8847-cd20af8f756fn%40googlegroups.com.

[Kevin McQuiggin]

Hi Mark:

Small point, but the .ini script attaches rl1 to a file that wasn’t in the zip archive:  'attach rl1 excprv.rl01’

Could you please post/email excprv.rl01 so I can give the system a try?

Thanks,

Kevin