PDP 11/60 versus PDP 11/84 retoration.

[lowen]

I have received two rather nice older DEC systems here, and was wanting to see some opinions on which would be the 'cooler' and which would be the more 'doable' restoration projects with an eye to a working museum environment.

System number one is a PDP 11/60 in a non-original but very cool 42U rack. Number two is an 11/84 in the standard DEC rackmount chassis, but not in a rack.

Both of these systems have Emulex SU3110401 SMD interface cards (the only reason I know they're SMD is because the three drives that came with the two systems are SMD: Control Data SABRE 9720 368MB 8-inch non-removable media drives). Both systems have a pair of MDB DR11-B-compatible DMA cards cabled to a Sun 4/360 VME system. The 11/60 programmer's console even has the Sun 4/360 badge on it.....

I know that there are folks here with 11/84s running, but does anyone here have a running 11/60? (A search of the forum for 11/60 turns up empty.)

If so, how difficult was it to get running? I do have a spare set of cards for the 11/60 CPU (M7872, M7873, M7874, M7875, M7876, and M7877) as well as a spare RAM board for the 11/60 (Standard Memories MM-147; has 8 rows of 18 4116 DRAM's on it (128KW/256KB)). The 11/60 has what appears to be three DL11-W compatible cards; one is an actual M7856 but the other is a clone.

The 11/84 (says it's an 11/84-PC on the chassis) has an Emulex SMD controller, two MDB DR-11B DMA boards, two (clone) DL11-W SLU's, the M8191 UBA, two M7458-AC RAM cards (MSV-11-RA), the M8190 CPU (I didn't check the rev on this one, but will), and the M7677 MDM.

I have the key for the 11/60 but not for the 11/84; the 11/60 uses the XX2247 key, but not sure what the 11/84 would use.

So I'm looking for the collective experience in bringing these systems up; things like trouble spots to look for, etc, along with an idea of which system would be the more appropriate system to have up and running both as a living museum piece as well as a data-recovery piece. My gut feel is that the 11/84 may be the more 'useful' one to have running, with the 11/60 in display condition, but I'm certainly willing to be corrected on that.

The two units were used to control a historic (and gorgeous, in a technical sort of way) astronomical plate scanner, and so there is a rack full of custom interface electronics that interfaces to the scanner hardware as well. The bulk of the software is written in PolyFORTH, I think.

I can provide photos a bit later. Any advice is appreciated!

 

[Uniballer]

I don't think there were very many 11/60s sold, compared to, say, 11/70s. I only ever saw one. I remember students complaining that when the 11/60 at SU was running you couldn't pick up the campus FM radio station anywhere in the ECE building. I would rather use the /84 (faster and 22-bit with I/D space), but I guess the rarity of the /60 is kinda cool.

 

[m_thompson]

The 11/60 is rare, so cooler. You can send the less cool 11/84 to me, so I can give it a good home.

 

[gslick]

It must be significantly easier to get replacement parts for an 11/84, if necessary. As far as I know the only boards specific to that are the M7677 MDM monitor and distribution module and the M8181 KTJ11-B UBA module.

The M8190 KDJ11-B CPU modules are fairly common. One thing that doesn't seem all that common are the M7458 MSV11-R PMI parity memory modules. I have never seen any of those. The M8637 MSV11-J PMI ECC memory modules are fairly common though and would be drop in replacements if necessary.

 

[pbirkel@gmail.com]

The 11/60 is rare, so cooler. You can send the less cool 11/84 to me, so I can give it a good home.

I agree, however as noted by others the 11/84 will significantly outperform it and be easier to maintain.

But IMO it would be a crime not to get that 11/60 back online. It *does* have a physical console, if only 7-segment displays.

However, if you decide to go for the utility-approach (11/84), then I'd be happy to come pick up that re-racked 11/60 and ensure that it gets a lot of TLC and restoration ... and compensate you suitably :->.

Looking forward to the pics!

-----
paul

 

[Marty]

Hi All;
I can't help You in the decision that you need to make, The company I worked for in the 1980's had an 11/60, but I was not allowed to Touch it, as it was also used for Tape Backup of all of the files for the different boards.. And the 11/84 is too new for me to have any experience with, of course for that matter so is the 11/60..
But, as I see it, You have four choices, Keep Both of them, Sell Both of them and get something else, or sell one or the other..
I would say first take an assessment of what You have and If possible what works, Also, check the Power Supply Situation on both of them, before making any decisions.. When taking your assessment, also figure out what You would need that You don't have for each system, and its cost.. Then You should be in a better position to decide which way You want to go and how much You will have to spend to get there..

THANK YOU Marty

 

[billdeg]

answer = BOTH!
you'll get to a certain point with one, then move over to the other. I don't think rarity has anything to do with it really. You'll fix one faster if you work on both. You will learn quickly which of the two is the better candidate for restoration.
b

 

[lowen]

answer = BOTH!
you'll get to a certain point with one, then move over to the other. I don't think rarity has anything to do with it really. You'll fix one faster if you work on both. You will learn quickly which of the two is the better candidate for restoration.
b

Thanks to all for the advice thus far.

I believe I will need to do what Bill has suggested and see which one comes up the easiest. I was hoping someone with an actual 11/60 would chime in, but it appears no one here has one, so I'll forge out on my own. The bitsavers archive has the print set it appears, and there were some docs with the systems.

Anyone here brought up a CDC 9720 SMD drive? One thing I am lacking is a good overview of termination and such for SMD; one of the drives has what appears to be a set of terminators plugged in, but one of the SIP's is broken on the terminator board; the other two have what look like jumper boards, both of which are broken (I know, a picture is worth a thousand words, and I need to get some pics...... in a few days.....).

 

[Marty]

Hi All;
Iowen, "" Anyone here brought up a CDC 9720 SMD drive? ""
I may have, but that has been about 30 Years ago, and I have forgotten, more than I knew.. Since we used SMD Drives on our Spectra Boards for testing..
THANK YOU Marty

 

[toober]

I had a PDP-11/60 a couple of years ago that I pulled out of the Black Hole, but I sold it. Not my bag.

Apparently they were a bit of a troubled product from DEC, and their lunch was quickly eaten by the 11/44.

--
Will

 

[AK6DN]

I had a PDP-11/60 a couple of years ago that I pulled out of the Black Hole, but I sold it. Not my bag.

Apparently they were a bit of a troubled product from DEC, and their lunch was quickly eaten by the 11/44.

Will

The 11/60 is an interesting machine for its time, especially with the user programmable control store option (an add-in board). You could define your own microcode for some of the reserved PDP-11 opcodes and implement those in microcode. One of the more interesting microcode implementations was for a PDP-8 emulation, done by an internal DEC engineer (Richie Lary, long time PDP-8 guru). This made the 11/60 at the time the fastest PDP-8 available, and they used it for software development.

The biggest problem with the 11/60 was it was designed to be a 256K (18b) machine targeted for core memory use, just when MOS memory was becoming much cheaper. Had the 11/60 supported 4MB (22b) like the soon to be released 11/44 most likely there would have been no need for the 11/44 at all. The 11/60 also did have the highest performance hardware floating point option of any PDP-11 CPU (the FP11-E).

The other issue was the fact the 11/60 only came in the corporate cab option as a full cabinet. There was no option to put it in a BA11 10.5" rack mount box like the 11/04-34-35-44 could be placed in (important for OEMs). DEC's cabinet choice made the 11/60 unsuitable for OEMs, and really only suitable for an end-user configuration. This was a (failed) marketing choice. The 11/60 backplane (and console) could have been adapted to the classic BA11-A/K box packaging but never was.

All this resulted in the 11/60 being (I believe) the shortest lifetime (1978-81) and lowest volume PDP-11 of the entire family.

In terms of restoration I have not heard of any running 11/60 systems at this time. I know Henk in the Netherlands has one but I don't think it is operational. I know several people have hardware (I have an 11/60 backplane, CPu board set, FPU, DCS, and am going to attempt to repackage it into a spare BA11-K box I have) but no fully running system.

An 11/60 restoration would certainly bring high points for the 'uniqueness' factor.

Disclaimer: I was at DEC 1975-82 and wrote all the hardware diagnostics for the 11/60 FP11-E FPU option, as well as writing all the microcode for the 11/60 DCS diagnostic control store option (an add in board that run microdiagnostics aimed at chip-level fault isolation of failures).

Don

 

[Marty]

Hi All;
Don, WOW !!!!!!!!!!!
"" I was at DEC 1975-82 and wrote all the hardware diagnostics for the 11/60 FP11-E FPU option, as well as writing all the microcode for the 11/60 DCS diagnostic control store option (an add in board that run microdiagnostics aimed at chip-level fault isolation of failures). ""
I am Amazed, and AWE STRUCK !!!!!
"" An 11/60 restoration would certainly bring high points for the 'uniqueness' factor. "" !!!!!!!!!!!!!!!!!!!!!!!!!!!!!
Don, No wonder, You are so good with the TU58 Programs that You have put out for the Rest of Us..
Did You have any Experience with any of the Other PDP-11's , During that time ??

THANK YOU Marty

 

[AK6DN]

Hi All;
Don, WOW !!!!!!!!!!!
"" I was at DEC 1975-82 and wrote all the hardware diagnostics for the 11/60 FP11-E FPU option, as well as writing all the microcode for the 11/60 DCS diagnostic control store option (an add in board that run microdiagnostics aimed at chip-level fault isolation of failures). ""
I am Amazed, and AWE STRUCK !!!!!
"" An 11/60 restoration would certainly bring high points for the 'uniqueness' factor. "" !!!!!!!!!!!!!!!!!!!!!!!!!!!!!
Don, No wonder, You are so good with the TU58 Programs that You have put out for the Rest of Us..
Did You have any Experience with any of the Other PDP-11's , During that time ??

THANK YOU Marty

Well thanks but amazing and awe is probably a bit of an overstatement. But thanks anyway. I was just a newly minted EE out of college when I went to work for DEC in '75. It was a very good time to be there for a new engineer. Lots to do, lots to learn.

As to other PDP-11s, after the 11/60 project finished I went to work on microcode for the PDP-11/74 CIS commercial instruction set option for the 11/74 project. The 11/74 CIS was a three board IIRC addin into the 11/74 backplane that implemented the PDP-11 commercial instruction set at 'optimum' performance (ie, our goal was to be able to run all CIS opcodes at full PDP-11/74 memory bandwidth). We were able to do this, but shortly after we finished the implementation DEC decided to cancel the 11/74 upgrade that included provision for the CIS. They reused the 11/74 model name but it was basically a repackaged 11/70 with MP hooks for a one to four processor configuration.

Here is a 'real' 11/74 front panel: http://ak6dn.dyndns.org/stuff/1174.jpg that I saved after the project was cancelled. Notice the extra positions on the lower right datapath switch for 'CIS uADDRS', 'CIS DECL', and 'CIS DESCR'.

Here is what a standard 11/70 front panel looks like: http://upload.wikimedia.org/wikipedia/commons/f/fe/Pdp-11-70-panel.jpg

At the time I was working on the 11/74 CIS microcode another engineer nearby was working on the 11/44 CIS microcode. It was a much less ambitious effort, only targeting getting all the features implemented. Performance was not the ultimate goal.

In the lab at the time the 11/44 hardware was being brought up as well as the 11/730 (NEBULA) and the 11/24.

Don

 

[Marty]

Hi All;
Don, "" Here is a 'real' 11/74 front panel: http://ak6dn.dyndns.org/stuff/1174.jpg that I saved after the project was cancelled. ""
Do You have the Whole Front Panel (Console, Switches, electronics) or just the FacePlate ??
"" The 11/74 CIS was a three board IIRC addin into the 11/74 backplane that implemented the PDP-11 commercial instruction set at 'optimum' performance (ie, our goal was to be able to run all CIS opcodes at full PDP-11/74 memory bandwidth). ""
By chance is there Schematics for any of these Boards ?? (probably not..)
Still You had to get into the Micro-Code and understanding of the different Instructions and what they did, whether Regular Instructions or Commercial Instructions.. In other words You had to understand what microcodes, combined in what order made what Instruction..
That's NO Small Feat !!!
"" and the 11/24."" I have an 11/24 Board, I don't know whether it works or not, it came with my PDP 11/04 chassis, I think the previous owner tried to put the 11/24 into the 11/04 chassis, hoping it would work..

THANK YOU Marty

 

[AK6DN]

Hi All;
Don, "" Here is a 'real' 11/74 front panel: http://ak6dn.dyndns.org/stuff/1174.jpg that I saved after the project was cancelled. ""
Do You have the Whole Front Panel (Console, Switches, electronics) or just the FacePlate ??

Unfortunately just the plexiglass. The electronics console board itself is different than the standard 11/70 front panel console board.

The 11/74 CIS was a three board IIRC addin into the 11/74 backplane that implemented the PDP-11 commercial instruction set at 'optimum' performance (ie, our goal was to be able to run all CIS opcodes at full PDP-11/74 memory bandwidth). ""
By chance is there Schematics for any of these Boards ?? (probably not..)

Nope, don't have the schematics either. DEC never officially released them. I do remember that of the three hex boards one was a packed decimal processing datapath, one was an integer processing datapath, and the third was a microsequencer/control store. The microcode store for the CIS was 96 bits wide by 4096 words deep. There were three of us writing microcode.

Don

 

[Marty]

Hi All;
It makes me wonder what other projects got cancelled by them..
They cancelled the PDP-X, and this for the PDP 11/74, it would be interesting to find out what else they started and somewhere in the middle of it, they axed it..
"" and the third was a microsequencer/control store. The microcode store for the CIS was 96 bits wide by 4096 words deep. "" That had to be some Micro-sequencer..
THANK YOU Marty

 

[Marty]

Hi All;
Don, "" I have an 11/60 backplane, CPu board set, FPU, DCS, and am going to attempt to repackage it into a spare BA11-K box I have) but no fully running system. ""
When You say ""but no fully running system. "" Do you mean You don't have all of the Boards needed to make a fully Running System, OR
That You just don't have the Time or something like that to put the Backplane into the BA11-K, or is it more like You have the Boards and the Chassis, but it needs a working Power Supply or something of that nature ??
I am Just Wondering What is Holding You up ??
You Would have a very Unique System, and with All You know about an 11/60, You would be the Right Person to do the Job..

THANK YOU Marty

 

[lowen]

The 11/60 is an interesting machine for its time, ...

The user writable microprogram control store is possibly the reason this particular one was purchased for its particular use. Although there is no WCS card in this box, it could have been something that was being looked at; the system software itself was written in FORTH, and the WCS fits the FORTH paradigm perfectly.

In terms of restoration I have not heard of any running 11/60 systems at this time. I know Henk in the Netherlands has one but I don't think it is operational. I know several people have hardware (I have an 11/60 backplane, CPu board set, FPU, DCS, and am going to attempt to repackage it into a spare BA11-K box I have) but no fully running system.

An 11/60 restoration would certainly bring high points for the 'uniqueness' factor.

Full disclosure: This machine was recently donated to the 501(c)(3) that I work for, which has a science museum exhibit gallery that includes NASA STS artifacts, and having a unique and running 'visible PDP 11/60' might be an interesting thing, especially since various PDP-11's once were used here for classified operations. And when I say 'visible PDP 11/60' I mean visible, including the cards being visible (but behind acrylic) and operating. The 11/84 could then be used as a media exchange machine for any archives we might get that would require such (unless we were to get funding for something like a full commercial Ersatz-11 license with physical Unibus and Q-bus adapters).

Disclaimer: I was at DEC 1975-82 and wrote all the hardware diagnostics for the 11/60 FP11-E FPU option, as well as writing all the microcode for the 11/60 DCS diagnostic control store option (an add in board that run microdiagnostics aimed at chip-level fault isolation of failures).

Don

Don, thanks for contributing to the thread! That's fantastic.

Here are a few pics:

The 11/60 in the Landau 19-inch rack, from the back with card cage visible:


The console of the 11/60 with the 'Sun 4/360' badge on it:


Closer look at the cards in the card cage (sorry for the bit of blur; phones aren't the easiest things to hold steady!):


Card list by slot (I'll edit the post later with more and better descriptions, and forgive me for non-standard nomenclature in naming the edge connector sockets, as I'm a bit green on Unibus.....):
Backplane #1, 14 slots:
1: Empty (WCS would go here)
2: M7872
3: M7873
4: M7874
5: M7875
6: M7876
7: M7877
8-11: Empty (FP11-E goes here, as I recall)
12: Emulex SC3110201 four drive SMD controller
13: M7856 in lower quad; M9312 in upper dual
14: Four-port non-DEC serial card in lower quad; right half of M920 in upper dual
Backplane #2, 9 (?) slots,
16: Bus req/grant continuity card in lower quad; left half of M920 in upper dual
17: 256KB parity RAM card, made by 'Standard Memories'
18-23: Req-grant continuity cards
24: Req-grant continuity cards lower quad; upper dual right half of custom backplane bus jumper (photo below)
Backplane #3, 4 slots
26: Upper card cage interface card (non-DEC) in lower quad; upper dual left half custom bus jumper
27 (unlabeled on cage): MDB-DR11B
28: Two short cards with handles (lower one probably standard req/grant continuity card; upper one unknown)
29: non-DEC interface board, unknown type, probably a parallel interface in lower quad; M9302 upper dual.

Photo of M920 and other custom bus jumper/extender:

 

[lowen]

More pics:
Backplanes from the rear:


(spare) Standard Memories 256KB RAM (note: carpet in photo is anti-static raised floor tile):


Emulex SMD controller (on top of the 11/84); close-up of labeled chip from one in 11/60:


VME DR11W interface in Sun 4/360:

 

[lowen]

Pics continue:

System diagram drawing. Has some interesting numbers on it:


Image Processor is the 4/360; uP is a pair of custom 6809 single-board computers. The 11/23 was not with the system we received; likely the 11/84 took its place and appears to be a warm spare for the 11/60, as there is a complete duplicate set of boards in the 11/84. The printer card is in the 11/84, for instance.

This system was used to drive a gorgeous astronomical plate scanner.