• Please review our updated Terms and Rules here

Q/CD slot compatible cards

ccureau

Member
Joined
Sep 28, 2019
Messages
32
Location
Louisiana
I'm sure that this has been answered before, so my apologies...I couldn't get results by doing a search for "q/cd".

I've just gotten a PDP 11/23+ (M8189-BA) and a 512k memory module (M8067) that I'd like to use to build a small system. I've got both a small H9270 backplane (not yet wired for 22-bits) and a BA23 with a H9278A in it.

What I'd like to do is to use the 11/23 and memory in the BA23. From my research (DEC MicroNote 5) I *think* that I can put both of these into the Q/CD slots. The manual for the CPU definitely mentions that some jumpers need to be arranged for the Q/CD, but I haven't yet looked at the memory.

The real question - is it safe to use the Q/CD slots for these boards? If not, I'll wire-wrap the smaller backplane to 22-bits and use them there.
 
With a little more digging, I found this in some mailing lists from 2003. If anyone can confirm, I'd be most appreciative!

There's a rumour that the first 3/4 slots in a BA23/123 are special and
you can't use them for anything except memory or processor; that's
rubbish, they're normal Q22-CD slots and you can put anything that's
Q-CD compatible in them, so long as it's after the processor. What
*is* special is the range of PMI memory boards; if you put them in
*front* of a KDJ11-B they use the CD interconnect for the Private
Memory Interface
but if you put them *after* the CPU they use the Q-Bus like all other
memory boards.
 
Yes, that's true. The first 3 slots in a BA23(H9278A) or 4 slots in a BA123 are QQ-CD. There are a small number of quad cards that are safe in a QQ-CD slot as is and some require jumper changes. There are also some that will fry themselves if you do so you have to check. Offhand I don't remember which ones nor where to find out other than the user manual for the cards. You can, of course put a dual-wide QQ card in the left side and ignore the CD side as well. If you skip a slot then the bus grant card (M9047) goes in the QQ side and nothing in the CD side.

Your M8189 is safe in the QQ-CD slot as mentioned. I'm not sure about the memory. It's a quad wide and could possibly expect a QQ-QQ slot. In which case you would put it in slot 4 of the BA23 and use bus grant cards to fill in. You could also put the CPU in the last QQ-CD (slot 3) and then the memory in slot 4 and ignore slots 1 and 2. I think that would work.

In the MSV11-P User guide on page 11 it has info about jumpers to set for Q22/Q22 or Q22/CD operations. It's for grant continuity. In an earilier page it show pinous and the board uses little on the CD connections other than power and the grant chain pins.

 
The CPU should always be in the top slot UNLESS it is a 11/83 or similar and has PMI memory, then it goes immediately below the PMI Memory (check the manuals for the boards concerned!) - your two boards are best put in the top two slots, CPU in the top, then RAM, with the next boards on the left in the third and fourth slots, then following the serpentine rules.
 
The CPU should always be in the top slot UNLESS it is a 11/83 or similar and has PMI memory, then it goes immediately below the PMI Memory (check the manuals for the boards concerned!) - your two boards are best put in the top two slots, CPU in the top, then RAM, with the next boards on the left in the third and fourth slots, then following the serpentine rules.
Thanks for the clarification! What I'm still not clear on is whether the KDF11B and the 512k memory can survive being inserted into the top three slots of my H9278A backplane, which are Q/CD slots.

I've looked through the manuals for both cards, and all I can find is that jumpers W1 and W2 should *not* be installed. I *think* that is if it should go into the Q/CD slots rather than the Q/Q slots, but I really don't know.

According to DEC Micronote 5, both cards are listed as Q22 compatible, and I can see that the manuals confirm that.

Maybe what would help is an idea of how Q/CD is different from Q/Q...if I can guarantee that no CD pins are being used on the boards then I'm happy putting them into the first three slots.
 
Thanks for the clarification! What I'm still not clear on is whether the KDF11B and the 512k memory can survive being inserted into the top three slots of my H9278A backplane, which are Q/CD slots.

I've looked through the manuals for both cards, and all I can find is that jumpers W1 and W2 should *not* be installed. I *think* that is if it should go into the Q/CD slots rather than the Q/Q slots, but I really don't know.

According to DEC Micronote 5, both cards are listed as Q22 compatible, and I can see that the manuals confirm that.

Maybe what would help is an idea of how Q/CD is different from Q/Q...if I can guarantee that no CD pins are being used on the boards then I'm happy putting them into the first three slots.
Yes. like I said above - they will be totally fine and in spec... the C/D slots are used to interconnect a pair of boards, like CPU and PMI memory, or the two boards in an RLV11 disk controller - if you look at the C/D connectors on your boards, you will see that they basically only take power - not signals. The bus grant links should be in for Q/Q slots but won't do any harm if left in - many boards don't even provide jumpers for them and just link the pins
 
Side thought: When I pull out my 11/73 CPU I just stick the 11/23+ CPU in slot 2 or 3 and be done with it. Doesn't seem to cause problems in a BA23 and avoids the rest of the CD stuff.

The only cards that I know of which have CD issues are true CD cards (MSV11-J) or the 8013/8014. They will blow up if put in a Q/Q. No issues that I recall with other Q/CD cards.
 
Thanks to everyone for lending a hand here, and I apologize for still being a neophyte when it comes to the internals of a PDP-11.

I plugged in the CPU board and the memory last night in slots 1 and 2, and turned on the system...no loss of magic smoke! The LEDs are cycling, but no output on the console.

With all the jumpers to factory settings, I get a code of 012, or what the manual affectionately calls a ROM Bootstrap Error. The HALT button does not bring me to ODT. I hooked up a logic probe to the console port and saw nothing come out of it, except for some garbage voltages on pin 8. Moving on to the 6402, I see no serial data coming from the serial output pins. I haven't yet snooped the bytes coming in from the bus to see if I get real characters.

Looking again over the manual, I tried setting the console test jumper (J18 to J17 installed, J18 to J19 removed) and got a code 017 (system hung) for my efforts.

If this would be best in another thread, we can start there. I truly appreciate any and all help I can get to resurrect this system.
 
The manual I have also states that code 012 is not used on a KDF11-BA (if that is what your card is).

I see the default links select power-up mode 2. I would set the link(s) for mode 1 (to enter ODT) and see where that gets you.

However, I have just seen that you identify J17, J18 and J19 as being related to the console test jumper but in my manual these are the power-up mode setting.

Can you post me a link to the manual you are using please?

Dave
 
That was the manual I was using.

Question: What ROM firmware is installed in E126 and E127?

Have you tried setting power-up mode 1 (ODT) but with no memory board installed. I usually do this test first on an unknown machine - it takes a fault on the memory board upsetting the bus away from the equation.

You can still use ODT to examine and modify registers.

However, I suspect from your fault explanation, that this may not get us anywhere either.

Dave
 
Regarding the EPROM: The stickers have long since lost their writing, so I'm not sure what version I have. I can see that they are SCM90448C, so 8k each. I'm not opposed to pulling these and seeing if I can read them with my programmer.

I pulled the memory board and tested both in mode 1 and mode 2 - there is no change.
 
Regarding the EPROM: The stickers have long since lost their writing, so I'm not sure what version I have. I can see that they are SCM90448C, so 8k each. I'm not opposed to pulling these and seeing if I can read them with my programmer.

I pulled the memory board and tested both in mode 1 and mode 2 - there is no change.
If your programmer can't handle these slightly strange EPROMs, I have on that will. That is, if you don't mind mailing them back and forth... Cheers!
 
Back
Top