MicroPDP 11/73

[RSX11M+]

Yes, it's identical.

Of course, the first Cabinet Kit belongs to your MicroVAX II CPU [the one with the dial], so keep it and the cable with that CPU. if you should ever trade or sell it, the value will be enhanced by including it.

In fact, it's a good idea to do this with all cable bearing modules.

Be sure to get the cabinet kit and cable for / with the KDJ11-B when you locate it. You should also know that there are several differences between KDJ11-Bs. Primarily these have to do with operability of the FPJ11 socket, presence of the FPJ11 chip, 15Mhz vs 18Mhz CPU, and the Gate Array function with PMI memory. I am told the Gate array function is always "workable", so the most important to you will be the other features. The 11/83 employed the 18Mhz CPUs, 11/73s are the 15Mhz versions. Either one could have issues with a working socket for the FPJ11. That's pretty much what I'm aware of.

A note: These BA23 rear panels are originally attached to the BA23 box via two short wire tethers. These take the weight of the assembly when it's not screwed in, so that the electronic cables are not strained. It's not thought out too well for the pedistal stand however, and gets in the way a lot, which is why they're often removed. [Not so in a Rackmount situation - where they are quite nice to have] I tell you this because this feature of originality may matter to you if you want to do a "Berserk" faithful restoration. If you want to compromise in a way that won't deduct from originality, detach the cables at one end and tuck them into the unit on the other.

 

[leaknoil]

Yes, it's identical.

Of course, the first Cabinet Kit belongs to your MicroVAX II CPU [the one with the dial], so keep it and the cable with that CPU. if you should ever trade or sell it, the value will be enhanced by including it.

I'm a little confused, which isn't uncommon when talking about DEC stuff. I won't need the dial/switch/DB9 for a PDP cpu ? What does a 11/73 have stock ?

 

[RSX11M+]

Same purpose, different part.

I couldn't find a decent pic, so here's a cheap and dirty one... I don't think it's the same as the MicroVax, is it?

 

[leaknoil]

Same purpose, different part.

I couldn't find a decent pic, so here's a cheap and dirty one... I don't think it's the same as the MicroVax, is it?

Ok got it. That's not the same at all. All the 11/73-83 came with that as opposed to the mVAX-II versions that came with the one I have. Right ? I've never seen a microPDP before this one but, lots of mVAX so, you'll have to bear with me a bit. I've never seen the PDP panel before.

 

[RSX11M+]

No problemo.

This cabinet Kit was only used with SLU KDJ11s, so it wouldn't even be MicroPDP11 generic. [I.E. - Won't work in a PDP11/23+]

Look on the bright site - If someone didn't make it useless, you might not have gotten it?

You may as well get used to reading PDP11 documentation if you're going to really restore this thing. Go to bitsavers and download the KDJ11 manuals. I've no time at the moment, or I'd do it myself and provide you links.

CUL8R

 

[RSX11M+]

I've located my January 1985 S&O Catalog.

It was filed with the VAX S&Os. In January of that year, the PDP-11 & Micro-VAX S&Os were a combined issue. [Although there was a VAX only issue too] After that, VAX and PDP-11 were issued separately.

Anyway, I see no additional information that pertains to your system.

 

[leaknoil]

Thanks for checking anyway. I'm still looking for all the pieces to make this happen. Case is nice and clean now anyway.

 

[Doug G]

Just so you know you're not alone, I have a model 173QA-B2 chassis without the pedestal, incomplete, but in mine is a KDJ11/A (m8192), no memory, an Emulex appears to be a TC05 emulation of some type, two Sigma 8port multiplexers and a 4line console board. It's been in my garage for close to 25 years, my recollection it was a modified configuration from the DEC original.

I found this thread from a google search when I was trying to identify the 173QA model number, I'd just gotten done telling my son it was a BA-123 chassis only to find out it wasn't.

 

[RSX11M+]

Just so you know you're not alone, I have a model 173QA-B2 chassis without the pedestal, incomplete, but in mine is a KDJ11/A (m8192), no memory, an Emulex appears to be a TC05 emulation of some type, two Sigma 8port multiplexers and a 4line console board. It's been in my garage for close to 25 years, my recollection it was a modified configuration from the DEC original.

I found this thread from a google search when I was trying to identify the 173QA model number, I'd just gotten done telling my son it was a BA-123 chassis only to find out it wasn't.

May we know the serial number of that system?


It might help us understand the manufacturing source and nomenclature. Please also look for dated stickers, particularly on the internal cables and parts. I obtained additional S&O and price lists since this thread, but still have not found this number. Could this have been an OEM sourced item, manufactured specifically for a vendor or re-seller?

Anything you can tell us about it's origin would be interesting.

 

[Doug G]

Anything you can tell us about it's origin would be interesting.

I'll get some more info off the chassis tomorrow, I have it out on a bench now.

To the best of my recollection I inherited this system when a small company I worked for dissolved in the late 80's. I believe the unit had been purchased from the whaddya call it "gray market" resale market as a possible alternative to the non-DEC OEM cabinets we'd been using for building small q-bus systems. Anyway, I ended up with this when the company went under and it's been sitting around ever since.

I think we put our own boards in it, so any dates or info off the pcb's may be misleading.

There is another sticker by the chassis model/serial sticker that says
<quote>
Product Variation
Model Changed
FROM: 173QA-B2
TO: 11/73-BC
</quote>
I think those numbers indicate the original processor was a KDJ11-B. We built a ton of 11/73 systems but they all were based on the 1/2 height KDJ11-A processors. I don't recall ever seeing a KDJ11-B board in person.

There's also a weird date code on the Astec power supply that looks to be TB715

 

[leaknoil]

So... RX50, Pedestal, probably RQDX1 which would limit it to RD51 or RD52 hard drives. I still maintain the most likely CPU is the KDJ11-B and it included DHV11 and MSV11-P memories.

How does that sound?

I am just getting back around to this (among so many other things) but, I opened up my other mV2 and it has a M7555 drive controller. A double wide qbus card. That's a RQDX3 right ? Will this work with the drive panel in a BA23 ? The one in the BA123 mV2 is different enough I thought I should ask first.

My plan right now is to get the system running as a MV2 at first and then switchout the ram and cpu and get it running as an 11/23+. It's the only cpu card I have so far. Still looking for a 73 or later card. Obviously not running off the same os but, I think I have everything I need to install an OS if I get that far ok. I just have no idea if the 11/23 and ram card I have work. I have tested the vax cpu and ram already in the other box. Seems like this way I can say everything else works ok then deal with the unknowns.

 

[RSX11M+]

I am just getting back around to this (among so many other things) but, I opened up my other mV2 and it has a M7555 drive controller. A double wide qbus card. That's a RQDX3 right ?

Yes.

Will this work with the drive panel in a BA23 ?

Yes.

The one in the BA123 mV2 is different enough I thought I should ask first.

My plan right now is to get the system running as a MV2 at first and then switchout the ram and cpu and get it running as an 11/23+. It's the only cpu card I have so far.

If you have the cabinet kit for the 11/23+

Still looking for a 73 or later card. Obviously not running off the same os but, I think I have everything I need to install an OS if I get that far ok. I just have no idea if the 11/23 and ram card I have work. I have tested the vax cpu and ram already in the other box. Seems like this way I can say everything else works ok then deal with the unknowns.

Take things at your own speed. Let us know how it goes.

 

[leaknoil]

If you have the cabinet kit for the 11/23+

The console panel for the 73 is the same as for the 23+ right ? If so, yes. If not, no. Looks right though. I have the mV2 panel off a spare mV2 to use for the vax cpu and the KDJ11 panel I bought from the place you recommended for the 23+ cpu. Already tested he vax cpu console with no drive controller.

Out of curiosity (and being lazy about digging through a pdf) the BA123 has the first four slot labeled "Q/CD" and the rest labeled "Q/Q" on the top. What does that mean ?

Here's a question I couldn't easily find the answer to and I actually looked. Whats the difference between RSX-11m and RSX-11m Plus ? I have a bunch of different disks images of things and I'm not really sure which to use. As I understand it you need at least 4.6 for y2k . Booting simh seems to confirm this. However is the version of Plus and the non-plus version the same ? I have something that calls itself rsx11m-4_8-bl80 then things like rsk11mplus-4_6-bl87 among many other variants.

 

[RSX11M+]

The console panel for the 73 is the same as for the 23+ right ?

Um... no. The 23+ one should have two serial port connectors.

...Out of curiosity (and being lazy about digging through a pdf) the BA123 has the first four slot labeled "Q/CD" and the rest labeled "Q/Q" on the top. What does that mean ?

This is a pretty big question, and requires a lot of explanation. Suffice it to say - if you don't know how to answer this one yet, you're endangering your hardware [flame is involved] to attempt to plug boards into a Q-BUS backplane (VAX or PDP-11) and power them. So please, set lazy aside and read up on it. All slots are definitely NOT the same, and neither are all BOARDS able to survive in just any slot.

Here's a question I couldn't easily find the answer to and I actually looked. Whats the difference between RSX-11m and RSX-11m Plus ? I have a bunch of different disks images of things and I'm not really sure which to use. As I understand it you need at least 4.6 for y2k . Booting simh seems to confirm this. However is the version of Plus and the non-plus version the same ? I have something that calls itself rsx11m-4_8-bl80 then things like rsk11mplus-4_6-bl87 among many other variants.

This one's a little easier. RSX11M+ is distinguished by it's ability to use the Instruction and Dataspace features of the 11/7X series CPUs. There are other less major differences, but that's "the" big one. Now to be clear - you didn't need to use this, but only M+ had it. Some designers used RSX11M+ on non-7X CPUs for it's other additional features - Disk cache, Overlapped I/O Seeks, etc.

This document on Bitsavers gives a lot of insight and details on the genesis and origins of RSX11M+, which was based on RSX11M v3.1 (the "GOLD standard" RSX for may years). RSX11M v3.2 was developed and released concurrently with RSX11M+ v1.0. The "GOLD Standard M+" revision would turn out to be v2.1.

I recall placing the purchase / upgrade request to take us to RSX11M+ (we received M+ v2.1 in response) as part of work I was doing in 1983 to re-write our Macro-Assembler / Linker to use separate I&D space.
​

Another viewpoint is that RSX11M will run on all PDP-11 machines up to that time (UNIBUS, QBUS, Mapped and Unmapped), whereas RSX11M+ was intended for "22-bit" PDP-11s and this means PDP-11/23 and beyond, Mapped only.

A minimum of RSX11M v4 or RSX11M+ v2 are required for MSCP (RQDX/UDA/KDA) series disk support.

Those system images you mentioned are probably "Virgin" SYS-GEN Kits, not yet "Built" for a specific target system or CPU. If so, they're usable "As Is" but may not contain a device list suitable to your needs.

Were the versions "the same" numerically?

No - not at all. As I recall, RSX11M 4.0 was out the same time as RSX11M+ 2.1.

I decided to research this a little, for the record: (from the actual Software Product Description(s) (SPD))

• RSX11M v3.1 September 1978 SPD 14.35.10 AE-3407K-TC
• RSX11M v3.2 December 1979 SPD 14.35.13 AE-3407N-TC
• RSX11M v3.2 April 1981 SPD 14.35.15 AE-3407P-TC
• RSX11M+ v2.1 December 1984 SPD 14.70.11 AE-H862L-TC **
• RSX11M v4.1 June 1984 SPD 14.35.22 AE-3407W-TC ***
• RSX11M+ v3.0 October 1985 SPD 14.70.12 AE-H862M-TX ****
• RSX11M v4.2 October 1985 SPD 14.35.24 AE-3407Y-TC ***

Notes:

** other docs place RSX11M+ v2.1 as early as April, 1983, but these may be "pre-release"
** RSX11M+ v2.0 to 2.1 transition doc: SPD 20.99.22 AE-M779W-TC is dated December 1984
**** RSX11M+ v2.1 to 3.0 transition doc: SPD 20.99.26 AE-GH14A-TC is dated August 1985
*** RSX11M v4.0 to 4.1 transition doc: SPD 20.98.09 AE-M778J-TC is dated June 1983
​

The last version of RSX11M+ I find any reference to is v4.6 dated January 1999. (This would be a Mentec product) I confess I lost track in the late 90s' after the Mentec purchase of "all things PDP-11". It may have turned out that M and M+ both ended on a similar revison number, but it's likely because they stopped M development and continued M+ longer, so it "caught up" numerically. As for Y2K compatibility, I recall a Y2K advisory and certification from Mentec, but which versions it applied to I couldn't say. I would guess that M v4.8 and M+ v4.6 were both Y2K compliant, being the final (?) RSX releases.


Ok, so now that part of the RSX time line is on the web. I'll try to do a more thorough one and post it in an article or blog at some point. I think I saved the PDP-11 Mentec web pages somewhere, although they were on the Waybackmachine last time I looked. I have RSX docs here that go back as far as RSX11M v3.0 (circa 1978?), about when I started in PDP-11's. (PDP-10's before that)

 

[leaknoil]

Um... no. The 23+ one should have two serial port connectors.

I guess I was hoping I could just set the baud rate on the cpu card and use just the serial line portion of the KDJ11 panel as the console. This is all just for testing really anyway. Until I get and actual 73 cpu. I was assuming, and probably shouldn't, that all these 10pin console ports are wired the same on the various qbus cpus. They all are even keyed the same. I could just wire something together if not.

The KDF11-b manual describes a 20 pin ribbon cable going from the panel to J2 of the backplane assembly. What is this backplane assembly of which they speak ? Did the 11/23 have a qbus backplane with a connector somewhere just for this ? Cpu card only has the two seperate 10pin slu ports. Do ba23 have this J2 available or they talking about a different backplane altogether.

 

[gslick]

The KDF11-b manual describes a 20 pin ribbon cable going from the panel to J2 of the backplane assembly. What is this backplane assembly of which they speak ? Did the 11/23 have a qbus backplane with a connector somewhere just for this ? Cpu card only has the two seperate 10pin slu ports. Do ba23 have this J2 available or they talking about a different backplane altogether.

The KDF11-B console connector bulkhead looks like this:

 

[RSX11M+]

Thanks Glen, I don't have a 23+ to take pics of.

1. The CPU plugs into the backplane
2. The Cabinet Kit "cable" plugs into the CPU
3. The Cabinet Kit "module" mounts to the "Back Panel" - not the "BackPlane". [if the manual says "BackPlane" it's a mistake]

The "physical arrangement" is the same as for your 11/73B Cabinet Kit, but it's not the "same one" as for the 11/23+.

Got it?

 

[leaknoil]

[if the manual says "BackPlane" it's a mistake]

"A ribbon cable installed in J2 of the backplane assembly carries the CPU signals to the 20-pin connector of the control panel." is what the KDF11-B manual has. Its EK-245AA-MG-001 I downloaded from somewhere or other. It seemed odd to me too. As the KDJ11 panel has a 10 pin and a 20 pin cable I was hoping the 10 pin was just serial data with the 20 pin for baud rate selection and lcd.

What I couldn't find in the manual, unless I missed it, is the pinout for the A0 and A1 slu ports on the KDF11-B. That would have answered the question. Does anyone have it ? They show the DB25 outs of the SLU but, not the cpu card. I can just ring out the lines if I need to. The baud rate can be set on the cpu card directly.

 

[gslick]

An interesting and non-obvious thing about the KDF11-B console bulkhead connector panel is that it is an active device in that it contains a dual baud rate generator. See section 2.2.7 Internal/External SLU Clock Jumpers in this manual:

http://www.bitsavers.org/pdf/dec/pdp11/1123/KDF11BA_UsersManual.pdf

If you don't have a KDF11-B console bulkhead connector panel you can just build a serial cable the same as you would for a DLV11-J and use the jumpers on board the KDF11-B to set the baud rate.

See pages 285 and 286 of this manual for DLV11-J cable wiring examples:

http://www.bitsavers.org/pdf/dec/qbus/EB-23144-18_QbusIntrfs_1983.pdf

Also see page 386 of that manual for a diagram of the C-D Bus Interconnection Scheme as implemented in the H9276. Note that is NOT the same backplane used in a BA23 chassis, in which only the first 3 slots are C-D slots.

The C-D interconnect is an application specific connection between each two adjacent slots. There are no signals which are bused across all slots in the CD half, except for power and ground. In the MicroVAX application it is required to be used as the private bus between the CPU and memory boards. In the 11/73,83 application it can also be used as a private bus between the CPU and memory boards (PMI), which is not compatible with the MicroVAX private memory bus.

You should assume that plugging any dual wide Q-bus card into the CD half of a backplane will damage cards in the backplane.
You should assume that plugging a quad wide Q-bus card that uses the CD interconnect into a non-CD slot will damage cards in the backplane.
A quad wide Q-bus card that only has power connections on half of the bus connectors can usually be plugged into either a Q-Q slot or a Q-CD slot, but still you should really know what you are doing and read the manuals to be sure. For example you may need to remove bus continuity jumpers if you plug a quad wide card into a Q-CD slot.

 

[leaknoil]

See pages 285 and 286 of this manual for DLV11-J cable wiring examples:

http://www.bitsavers.org/pdf/dec/qbus/EB-23144-18_QbusIntrfs_1983.pdf

That's exactly what I needed. Would have been easier if these panels were a little more interchangeable but, I can wire that up easy enough.

Also see page 386 of that manual for a diagram of the C-D Bus Interconnection Scheme as implemented in the H9276. Note that is NOT the same backplane used in a BA23 chassis, in which only the first 3 slots are C-D slots.

The C-D interconnect is an application specific connection between each two adjacent slots. There are no signals which are bused across all slots in the CD half, except for power and ground. In the MicroVAX application it is required to be used as the private bus between the CPU and memory boards. In the 11/73,83 application it can also be used as a private bus between the CPU and memory boards (PMI), which is not compatible with the MicroVAX private memory bus.

You should assume that plugging any dual wide Q-bus card into the CD half of a backplane will damage cards in the backplane.
You should assume that plugging a quad wide Q-bus card that uses the CD interconnect into a non-CD slot will damage cards in the backplane.
A quad wide Q-bus card that only has power connections on half of the bus connectors can usually be plugged into either a Q-Q slot or a Q-CD slot, but still you should really know what you are doing and read the manuals to be sure. For example you may need to remove bus continuity jumpers if you plug a quad wide card into a Q-CD slot.

I read a little about all this and think I got the jist now. BA23 had first three slots (Why do they number right to left ? Drives me nuts.) Q/CD rest Q/Q. I can plug in a dual card into 3Q though correct ? Just not 3CD without risking it. Any quad cards have to start in 4.