• Please review our updated Terms and Rules here

PDP-11/05 restoration blog

Unfortunately my PDP-11/05 started behaving strangely and increasingly failed. First Joe's new BASIC played up and forgot how to do TAB(). Then the diags failed after running for about one hour. Finally I couldn't even deposit 177777 into the first 000100 memory locations. When I examined them I got back random garbage.
The problem turned out to be core memory related. By that time the system was running and powered for several hours in a room with 20 degrees C but with the side lid off. There is enough air movement but the boards got warm - I estimate about 35 - 40 degrees C.

I have a spare set of G231, G110 and H214 boards to narrow down the cause and it turns out that the original G231 developed some nasty fault.
With the replacement G231 the system is back operational, but I will now have to fix the original G231 to maintain a complete working set of memory board spares.

It is interesting that several times now I had a fully working 11/05 which then failed in various ways.
I hope that after a few more such faults the system eventually stabilises.

As an insurance policy I would love to find spare 11/05 CPU boards. Specifically:
  • M7260 – Processor Data Path and UART
  • M7261 – Processor Control Logic and Microprogram
Please let me know via DM if you can spare these two boards (working or not) and would consider selling them.
 
thunter0512 said:
It is interesting that several times now I had a fully working 11/05 which then failed in various ways.
I hope that after a few more such faults the system eventually stabilizes.
Click to expand...
I have observed this behavior with all of the DEC systems that I have restored. After some runtime hours the failures decreased significantly. When running just on Saturdays for demonstrations the transistor/diode systems run for maybe 6 months between failures, and the integrated circuit systems run for years between failures.
 
thunter0512 said:
Unfortunately my PDP-11/05 started behaving strangely and increasingly failed. First Joe's new BASIC played up and forgot how to do TAB(). Then the diags failed after running for about one hour. Finally I couldn't even deposit 177777 into the first 000100 memory locations. When I examined them I got back random garbage.
The problem turned out to be core memory related. By that time the system was running and powered for several hours in a room with 20 degrees C but with the side lid off. There is enough air movement but the boards got warm - I estimate about 35 - 40 degrees C.

I have a spare set of G231, G110 and H214 boards to narrow down the cause and it turns out that the original G231 developed some nasty fault.
With the replacement G231 the system is back operational, but I will now have to fix the original G231 to maintain a complete working set of memory board spares.

It is interesting that several times now I had a fully working 11/05 which then failed in various ways.
I hope that after a few more such faults the system eventually stabilises.

As an insurance policy I would love to find spare 11/05 CPU boards. Specifically:
  • M7260 – Processor Data Path and UART
  • M7261 – Processor Control Logic and Microprogram
Please let me know via DM if you can spare these two boards (working or not) and would consider selling them.
Click to expand...

I find it interesting that your machine has experienced several separate failures. My /05 has run with minimal issue since I purchased it and repaired a few broken capacitors and transistors. I have not had the machine degrade or fail, however I have not put extensive runtime hours on the machine at this point. I will report back if I observe a similar phenomenon as the machine gets more time running real programs. My /05 was effectively brand new inside despite showing some external battle scars. I suspect the good condition of my boards may be a contributing factor to my /05's continued operation.

I wish I had a spare set to consider getting to you. I am also unfortunately also on the exact opposite side of the globe and shipping would be prohibitively expensive even if I did have a spare set.

My /05 runs at just about the same temp. I believe the cover significantly ducts the air and keep the chips in the path of the airflow. I hope the situation will be a little better with the lid on. I have experienced excessive heating on /44 systems with the cover off.
 
cullyrichard said:
I find it interesting that your machine has experienced several separate failures. My /05 has run with minimal issue since I purchased it and repaired a few broken capacitors and transistors. I have not had the machine degrade or fail, however I have not put extensive runtime hours on the machine at this point. I will report back if I observe a similar phenomenon as the machine gets more time running real programs. ...
Click to expand...

My earlier repairs & restorations of a LAB-8/e, PDP-8/e and DG Nova 2 stabilised relatively quickly. These required between 2 and 3 repair cycles and have since been running reliably.

The PDP-11/05 has been exposed to similarly bad conditions as the other 3 machines, just for longer. It also appears to have been exposed to harsher conditions during its productive life and has also been butchered by someone prior to me receiving it.

cullyrichard said:
... My /05 was effectively brand new inside despite showing some external battle scars. I suspect the good condition of my boards may be a contributing factor to my /05's continued operation.
Click to expand...

You are lucky man. My boards are very far from "brand new". Scrap is probably a better description. :)

Oh well - there is not much fun if a machine just works. :)

cullyrichard said:
I wish I had a spare set to consider getting to you. I am also unfortunately also on the exact opposite side of the globe and shipping would be prohibitively expensive even if I did have a spare set.
...
Click to expand...

Shipping does not have to be a problem as I have a US based shipping address.
 
m_thompson said:
I have observed this behavior with all of the DEC systems that I have restored. After some runtime hours the failures decreased significantly. When running just on Saturdays for demonstrations the transistor/diode systems run for maybe 6 months between failures, and the integrated circuit systems run for years between failures.
Click to expand...

Same here. I'm going through that with the PDP-11/34a, which was "completely functional" after the initial cleanup and power system rebuild. It's since had a failed latch in the MMU and a failed register file in the SPM. Totally normal for 50+ year old equipment :p Kinda like trying to daily a vintage car -- you eventually get to where you don't drive for three days, and fix for two!

Make sure you're hammering on it with MAINDEC and/or XXDP.
 
jlang said:
That tape is basic for the heath kit H-11. Heathkit will sue (and has) anyone posting their stuff.
I don't want to poke that bear. I never posted the disassembly anywhere only the edited and commented version that matched the LPTBA listing.
Click to expand...
just submit any heathkit stuff to libgen, they can sue all they want but the russians don't give a shit about copyright
 
For the past month I have been working on other stuff - mostly repair of a "new" RK8-E board set for my PDP-8/e. I have also finally built the Retro Chip Tester mostly to be able to test some of the more obscure Signetics ICs used in my DEC machines. Most of the more standard SN74xx series I can nicely test with my Minipro TL866ii Plus made by XGecu.

After cleaning all other distractions off my work bench, today I powered on my PDP-11/05 and via Pdp11GUI loaded Joe Lang's disassembled & reassembled version of BASIC. After starting it stopped after the first instruction. I then verified the core contents via Pdp11GUI and it is mostly garbage. This 11/05 was fully working one month ago. It was stored inside my house at a stable 21 - 24 degrees Celsius with very low humidity.

I am surprised how this machine develops new faults a short time after repair of previous faults. It works stable for a while and then fails without any environmental factors contributing to the fault. The 240V AC mains supply is stable. The power supply has very stable power rails with minimal ripple. Fans supply plenty of air movement over the boards. There is no obvious overheating of any components.

This is quite different from my LAB-8/e and PDP-8/e, both of which have been working reliably ever since I fixed them.

Oh well, back to debugging.
 
Are you perhaps up against backplane troubles? The systems I've had which changed behavior on movement or a few weeks of down time have almost always ended up with backplane issues -- mostly corrosion/oxidization in the card edge connectors.
 
jlang said:
That tape is basic for the heath kit H-11. Heathkit will sue (and has) anyone posting their stuff.
I don't want to poke that bear. I never posted the disassembly anywhere only the edited and commented version that matched the LPTBA listing.
It will only run on the LSI-11 so it's not much use to you.
Click to expand...
Apparently, Heathkit has risen from the dead.

Heathkit® FAQ

Heathkit Frequently Asked Questions
www.heathkit.com
 
jlang said:
copyright and trademark.
Click to expand...
this, they own the heathkit trademark and all associated IP and sell access to it, so they have a vested interest in making sure the filthy peasants who aren't willing to fork over $25 for a manual that should be free never get it
 
I like the concept of new Heathkits, but none of the ones I saw on their site made me want to buy them right away. It’s a shame that they are being uncooperative about people sharing vintage manuals for products that they could not market again even if they wanted to because of component obsolescence.
 
NF6X said:
I like the concept of new Heathkits, but none of the ones I saw on their site made me want to buy them right away. It’s a shame that they are being uncooperative about people sharing vintage manuals for products that they could not market again even if they wanted to because of component obsolescence.
Click to expand...
IMO the new Heathkits are NOT the value-proposition that the old ones were. The old ones came with detailed assembly instructions for a multitude of parts (and typically) lots of wiring, a good theory-of-operations, and often other theory-primer-like documentation. Current technology is a bunch of black-boxes with beaucoup pins. I don't see much value-added beyond the brand name over purchasing the same capability (kit or pre-assembled) on Alibaba or Amazon. I seriously doubt that "Heathkit" tests and burns-in any of their products/components after import from foreign sources before slapping on the logo and marking up the price.

I can understand their business model, but IMO the market has shifted radically from the glory-days of Heathkit when assembling your own TV was cheaper than paying American Union labor to do it for you. And you had the benefit of education for free. The labor question is highly inverted these days with automated assembly on foreign shores. The modern Heathkit label doesn't say anything more than many defunct labels that turned into resellers of (e.g.) Japanese imports in the 70's. IMO using it is rather an affront to history! Anyway ...

I can't judge whether the new Heathkits are the "gateway drug" for the modern generation that they were in the 60's. But I have my doubts as there are manifold competitors in the same space today, easily accessible in your mailbox. Heathkit was essentially unique in their heyday. We had a local store and window-shopping there was ... special.

So aggressively protecting their brand is the major portion of their business model, IMO. They haven't anything else that is "special" to sell.
 
thunter0512 said:
Thanks for all the valuable information about Heathkit but I am struggling to see any relevance to my PDP-11/05 restoration and debugging which is the subject of this thread. ;)
Click to expand...
Sorry I think that's my fault. I brought up the heathkit HT-11 basic paper tape and down the rabbit hole we went.
 
You must log in or register to reply here.
Back
Top