Restoration of a PDP11/60

[klapperp]

We are close to reach a milestone:


We are only waiting for the ordered replacement for the missing board holder in our H7850:


However, we will replace the 200mA 5V indicator bulbs with LEDs and an appropriate resistor so that we don’t have a failure due to a defective bulb

Once that is done, we will put together the entire power supply along with the cable harnesses and measure all voltages at the connectors for the backplanes before we install everything in our 11/60 cabinet.

To be continued ...

 

[klapperp]

Does anyone know if on the 11/60 CPU Backplane (KD11-K) the NPG (DMA) jumpers are closed by default on SPC slots 12, 13, and 14 or not?


On our backplane, they are not:

Check	Result	Meaning
Slot 12 CA1 → Slot 12 CB1	open	normal: slot 12 is an active NPG position or G7273 required
Slot 12 CB1 → Slot 13 CA1	closed	backplane chains slot 12 output to slot 13 input
Slot 13 CA1 → Slot 13 CB1	open	normal: slot 13 is an active NPG position or G7273 required
Slot 13 CB1 → Slot 14 CA1	closed	backplane chains slot 13 output to slot 14 input
Slot 14 CA1 → Slot 14 CB1	closed	slot 14 is already bypassed/continued because DL11W needs no NPG support

According to photos, in slots 12 and 13, two LP11 interfaces (M7258) were installed and they do not use DMA and also do not have the corresponding jumper:


Either the original NPG jumpers are missing here, because originally G727 Grant Continuity cards were supposed to be installed, or the slots on the KD11-K backplane are wired differently.

Does anyone have information on what the actual situation is?

Thanks, Peter

 

[pbirkel]

Does anyone know if on the 11/60 CPU Backplane (KD11-K) the NPG (DMA) jumpers are closed by default on SPC slots 12, 13, and 14 or not?
...
Either the original NPG jumpers are missing here, because originally G727 Grant Continuity cards were supposed to be installed, or the slots on the KD11-K backplane are wired differently.

See the attached photo of a DD11-K, unmodified. All three NPG pin-pairs are strapped deep in the nested WW. Surprise!

 

[klapperp]

@pbirkel Thanks a lot for your prompt answer! We will instal G7273 cards to close the NPG chain

 

[klapperp]

The backplanes and cards being prepared for reinstallation:


The CPU from our 11/60 (also in comparison to a J11 CPU from an 11/73):


To be continued ...

 

[m_thompson]

It will be interesting to see how much is alive when you power it on.

 

[klapperp]

It will be interesting to see how much is alive when you power it on.

We are all also very excited to see what will work when we put our "old lady" under power for the first time.

In any case, we will report on it.

 

[Henk]

I see that you also have the WCS option board. A rare one!
Do you also have documentation and - - software - - for the WCS option, microcode assembler and linker ?
I think that developing new microcode instructions is not "simple".
It would be great if somebody could/would write microcode so that the 11/60 can run PDP8 instructions. It has been done, so it is definitely possible. Only too bad that microcode has been lost forever.

I am very interested in the allocation of the boards. That "two busses" is still a "strange" thing to me.
I can load code in memory with the keypad in my 11/60 and single step it successfully but when I hit "RUN", it crashes. Still have not looked any further than this observation.
Soon ...

 

[pbirkel]

I see that you also have the WCS option board. A rare one!

To my eyes, no. I see the standard six modules that comprise a KD11-K. From right-to-left:

M7872 U Word (Microword)
M7873 Decode
M7874 Data Path
M7875 KT/Cache
M7876 Timing
M7877 Status

An FP11-E would add these four modules:

M7878 FNUA (Floating-point Next micro Address)
M7879 Exponent
M7880 FFP MULNET
M7881 FALU (Floating point ALU)

A M7870 Writable Control Store (WCS) looks like this:


If you go back to https://forum.vcfed.org/index.php?threads/restoration-of-a-pdp11-60.1255964/post-1484116 and look at the installed complement of modules you can see that slot 1 is empty. Also slots 8-11 where the FP11-E would sit are empty. Slots 12-14 are the three SPC slots mentioned earlier.

Slot 15 would be a gap between the KD11-K CPU and DD11-K Memory system units (backplanes).

Slots 19-22 are presumably M7984-EC 16K x 39-bit (32K-word) or possibly M7984-FC 32K x 39-bit (64K-word), and Slot 23 is the associated M7983 Memory Control. Slots 16-18 and 23 are SPC in C-F with slots 17-18 Hex-height MUD while slots 16 & 19 are standard Unibus in A/B.

Well ... that's my interpretation of the module positions in that photo. I defer to @klapperp for Ground Truth data :-}.

 

[Henk]

Thanks for the correction Paul. I am getting old
I really hope to check out my 11/60 soon!

 

[Cristiano Sword]

Hi everyone, I'm new here on the forum, I'm passionate about the PDP-11, I've been following your post, it's really cool to see every detail!

 

[klapperp]

Well ... that's my interpretation of the module positions in that photo. I defer to @klapperp for Ground Truth data :-}.

This is correct, our 11/60 has only a standard CPU. We have neither an FPU nor a WCS.

 

[pbirkel]

This is correct, our 11/60 has only a standard CPU. We have neither an FPU nor a WCS.

Which modules do you have for memory?

 

[klapperp]

Which modules do you have for memory?

Our 11/60 is equipped with M7984 MOS memory:


In my opinion this is the most critical part because of the complex setup of the H7850 PSU together with the H775 BBU.
We ordered two 12V 5Ah batteries for the BBU to be able to test everything and make sure that all power supply works before we do the first power on of the system.

 

[pbirkel]

Thank you for the photos. As it happens there are multiple versions of M7984 MOS memory. The one in your photos appears to be the -Ex version (which should be specified on another of the metal handle-tabs) as it uses the MK4027N-3 4Kx1 200nS part, so a total of 16K x 39-bit (32-Kword) -- that's the MS11-KE according to my notes. Example tab markings:



There's also a higher-density -Fx version that uses a 16Kx1 200ns MOS RAM (not sure of the RAM part number), so a total of 32K x 39-bit (64-Kword) -- that's the MS11-KF according to my notes. Just two of those modules would max-out the capability of the M7983 memory controller (and the 11/60 address space). I put together the following M7984-XX / MS11-Kx cheat sheet at one point that might prove helpful in the future:

Board	Model	Size	Configuration
M7984-Ax	MS11-KA	64KB	156 x 4K chips
M7984-Bx	MS11-KB	32KB	78 x 4K chips
M7984-Cx	MS11-KC	256KB	156 x 16K chips
M7984-Dx	MS11-KD	64KB	39 x 16K chips
M7984-Ex	MS11-KE	64KB	156 x 4K chips
M7984-Fx	MS11-KF	128KB	78 x 16K chips

Interestingly, the M7984-xx (although perhaps limited to the higher-capacity variants) are reported as also working in the 11/70 MK11 memory box, although normally you'll find M8728-Ax 128-Kword or M8750-Cx 512-Kword modules there.

 

[klapperp]

The one in your photos appears to be the -Ex version (which should be specified on another of the metal handle-tabs)

You are right, the 2nd metal handle-tab on all 4 memory modules shows "EC":


This will give us 128K-Word of memory when everything goes well.

 

[Terri Kennedy]

Thank you for the photos. As it happens there are multiple versions of M7984 MOS memory. The one in your photos appears to be the -Ex version (which should be specified on another of the metal handle-tabs) as it uses the MK4027N-3 4Kx1 200nS part, so a total of 16K x 39-bit (32-Kword) -- that's the MS11-KE according to my notes. Example tab markings:

FWIW, the second letter of the suffix on DEC memory boards of this era usually indicates the manufacturer of the RAM chips on the board.

 

[pbirkel]

FWIW, the second letter of the suffix on DEC memory boards of this era usually indicates the manufacturer of the RAM chips on the board.

I found it interesting that the example module appears to mix multiple manufacturers without obvious evidence of having resulted from chip replacements. The MS11-K FMPS is MP-00325. Online references: https://vt100.net/manx/part/dec/mp-00325-00/

M7984-EB	21-13914-01	4K MOS RAM 16 PIN IC	DEC (200NS)
M7984-EC	21-13735-01	4K MOS RAM 16 PIN IC	MOSTEK (200NS)
M7984-ED	21-14114-01	4K MOS RAM 16 PIN IC	MOTOROLA (200NS)
M7984-EE	21-14475-01	4K MOS RAM 16 PIN IC	FIJITSU (200NS)

It actually states "NO MIXING OF MOS CHIPS VENDORS ALLOWED". Hah!

 

[klapperp]

I found it interesting that the example module appears to mix multiple manufacturers

I think it is more likely that some faulty RAM chips have been replaced during the systems first life.
@Konstantin told me that this was a common practice in the eighties and he made this as well in his job at that time.

 

[pbirkel]

I think it is more likely that some faulty RAM chips have been replaced during the systems first life.
@Konstantin told me that this was a common practice in the eighties and he made this as well in his job at that time.

I have no doubt, but the example photo seems to indicate quite a few replacements, e.g., none of those Motorola ICs belong there if that's truly an -EC board (could it be an indistinctly-marked -ED?), and yet they are the majority of the RAM ICs. Perhaps a photo of the back will make clear where any replacements took place?