PDP-11/70 panel behaviour

[Joerg Hoppe]

Hi all,

In february 2018 Oscar (Vermeulen) and me where allowed do make hour-long tests with "Miss Piggy" .
That's the PDP-11/70 for exhibit in LCM Seattle, a former Microsoft server.

Goal was to make an exact emulation of the front panel operation, necessary for BlinkenBone and the PiDP11.
In fact many open questions and weird effects could not be deduced from the manuals, you had to see them.

In the test run, documentation and about 45 videos were generated.

All this stuff was barely public, but now I was told this material would be of historic importance.

Really?

Anyhow: now there's an article http://www.retrocmp.com/stories/pdp-11-70-panel-research
which explains the settings and links to youtube playlist
https://www.youtube.com/embed/videoseries?list=PL9ECYK0f539oZh3rIz-eEPAPpM6Lq2P77

thanks to Oscar Vermeulen and Steven Falco for nudging me in that direction,
also many thanks to Mike Hill, who friendly (and mercyless) pointed to all the emulation errors.

Joerg

 

[fritzm]

Kudos for taking the time to get this right!

HALT, DATA knob to BUS REG, then observe data lights following the switches is a very good/early test commonly done to verify that the microcode is actually running.

Another thing I have seen mishandled in emulators is the PROG PHY feature for verifying address mappings while halted:

- Set ADDRS knob to PAR set for which you wish to verify a mapping
- Toggle in a _virtual_ address and hit LOAD ADRS. Now hit EXAM to run a bus cycle; note that USER/SUPER/KERNEL and DATA lights now update to reflect which PAR set was used for that cycle (will match ADDRS knob setting)
- Now flip ADDRS knob to PROG PHY and address lights will change to show the physical address to which the bus cycle was actually mapped

My /45 is currently in running condition; please let me know if you'd like me to try any panel experiments for the /45 and I'm happy to help out if so!

 

[Joerg Hoppe]

Hi fritzm,

...
My /45 is currently in running condition; please let me know if you'd like me to try any panel experiments for the /45 and I'm happy to help out if so!

I'd really like to add the '45 to the pool of BlinkenBone panels, even got several requests for a BlinkenCape+BlinkenBoard+11/45 adapter.
However, I don't have a '45 panel in my collection for development ... and who is giving it away in 2021?

best regards,
Joerg

 

[fritzm]

I'd really like to add the '45 to the pool of BlinkenBone panels, even got several requests for a BlinkenCape+BlinkenBoard+11/45 adapter.
However, I don't have a '45 panel in my collection for development ... and who is giving it away in 2021?

Ah, well, I'd hesitate to lend it internationally since it would be pretty difficult to replace if it were to become lost. But I'd be happy to lend it to you for some time _locally_, for example if you ever ended up out here for a VCF West or some such (I am in Oakland, California). Or, more than happy to assist on this end with it if you felt like shipping any loaners this way -- I am well enough equipped here with a logic analyzer, etc. and familiar enough with Linux and embedded systems and with the /45 HW to possibly be useful to you.

Of course, if we do get past this global plague and people can travel again, I think there are other collectors in this forum much closer to you with /45s who I would think might also be talked into this...

cheers,
--FritzM.

 

[daver2]

I still haven’t retired yet (I did offer, but the company I work for wouldn’t let me go)...

I still can’t find the time at the moment to get my 11/45 running. One day I will retire though and get it going...

Anyhow, I did a lot of work with LOGISIM and the 11/45. As a large majority of the panel control is embedded within the 11/45 microcode I would be willing to either make this available, or to spend a bit of time tweaking things so that we can see how the panel operation parts of the microcode actually operates. I did find a number of disparities and/or omissions between the operation described within the documentation and the hardware/microcode.

Let me know if this may be of help?

Nice to hear from everyone again. I may even try and make it to a VCF event once all this is over... Stay safe in the meantime!

Dave

 

[Terry Kennedy]

BTW, LSSM has an operable 11/70 and if anyone needs to do further experimentation to resolve undefined front panel behavior, I'm sure Dave would be happy to run a few sequences or make the machine available to someone who wanted to visit New Kensington, PA (near Pittsburgh) to run them themselves.

 

[Joerg Hoppe]

Hi Fritz, Dave and all,

Ah, well, I'd hesitate to lend it internationally since it would be pretty difficult to replace if it were to become lost. But I'd be happy to lend it to you for some time _locally_, for example if you ever ended up out here for a VCF West or some such (I am in Oakland, California). Or, more than happy to assist on this end with it if you felt like shipping any loaners this way -- I am well enough equipped here with a logic analyzer, etc. and familiar enough with Linux and embedded systems and with the /45 HW to possibly be useful to you.

Of course, if we do get past this global plague and people can travel again, I think there are other collectors in this forum much closer to you with /45s who I would think might also be talked into this...

cheers,
--FritzM.

Thank you for your support.

To be honest, I'm hoping to get a 11/45 panel (even without bezel, even without acryl front) not just because
its needed for an BlinkenBone implementation.

Wanna have it permamently to complete my gallery!

kind regards,
Joerg