Jerome Fine replies: My best guess is that you have a BA23 box since you seem to be describing having two 5 1/4" drives included in your configuration. Since RT-11 is quite happy to boot with the drives in WRITE PROTECT, I suggest that you attempt to figure out the "sort of" front panel which allows that to be done for the RD54 (Maxtor XT-2190). If you have a floppy as well, place a WRITE PROTECT tab on it. If you can also provide the names of the boards, that would be more helpful. With a KDJ11-A (dual) CPU which does not have an boot ROMs, you may have a third party controller for the drives with a boot ROM. Otherwise, if you have an RQDX3 (M7555 dual board - required for an RD54) which does not have a boot ROM and there is no other board with a boot ROM, you will have to enter the boot program via ODT. Since this takes only about 5 minutes to do, it is not an impossible situation. Any other question? Jerome Fine

On 26/03/2009 05:34, John Floren wrote: M8192 is a dual-height KDJ-11A for an 11/73 and as Jerome pointed out that has no bootstrap. Possible reasons for being left at the ODT prompt include the HALT switch being on, or there being no bootstrap anywhere else. However, the next board, M7195, is a multifunction MXV-11B, which probably does have a bootstrap on it. Look for a pair of EPROMs on it. The manual is around the net, and I expect there's a copy on bitsavers. The KDJ-11 doesn't have all the ODT commands that some earlier QBus processors have. "G" is the command to "Go" to an address. "P" is the command to "Proceed" without initialising the bus. The only other commands on a KDJ-11 are slash ('/'), 'R', carriage-return and linefeed. I don't know exactly what the next few boards are, though I know Matrox made video boards for a variety of systems (DEC and S100 at least) and a QRGB is some sort of framestore and QVAF is a digitiser, I think. The Dataram board probably is memory, especially if densely packed with an array of chips. The chips numbers should tell us. M7941 is a DRV-11 parallel interface, which might have been used for a printer or simply for some other digital I/O to/from another device. M7555 is an RQDX3 MSCP disk controller, which interfaces via a breakout board to a dual floppy drive and/or one or more ST412-style hard drives. Are there disks and cables in the system? Thinking of that, what type of box is this in? What does the front panel look like? Jerome suggested it might be a BA23, but this mixture of cards suggests to me that it might just as well be a system that was upgraded from something older, and that it could be a BA11S or BA11N. You appear to have boards from at least three systems... This is a terminator board for a much older QBus system. Each is an RXV11 controller for a dual RX02 8" floppy drive - not useful unless you also have those specific drives. I don't know, but I'd guess some Texas Instruments GPIB/IEEE-488 instrument interface, like DEC's IBV11. Older QBus memory card from an 11/23 or possibly 11/03 system. There will be a suffix to the 8045 that will tell us the size of each. The have parity, so possibly from an 11/23. Dunno. DLV-11J quad serial line cards, mainly used with 11/03 and 11/23 systems. Later versions were renamed DLVJ1. They'd probably work (depending on the revision level) with your 11/73, but as you have two serial lines on the MXV11, probably not very useful. Another KDJ-11A CPU. Dunno. Another MXV-11 multifunction card. MSV-11L memory from an early microPDP-11, either 128KB or 256KB. This is an old dual-height LSI-11/2 CPU from an 11/03 system. Dunno. Terminator for that LSI-11/2 processor system. Does the number have a suffix? There were several version, some with bootstraps. Here you have the upper halves of four RLV11 disk controllers. These each need to be paired with an M8014 to do anything. They're from an 11/03 or early 11/23 system. A DL-11W serial line and line-time clock from a Unibus system. Don't plug this in to your Qbus! Matrox framestore, I believe. -- Pete Peter Turnbull Network Manager University of York

2009/3/26 John Floren <slawmaster at gmail.com>: Those are the MXV11-B2 ROMS and have a good few bootstraps on them. As long as you're not starting with the CPU halted, they can be run at power-up. There are however plenty of jumpers on the M7195, some of which you may need to fiddle with. If you have all drives disconnected, it should drop you to a BOOT> prompt eventually (this can take a while and hitting Ctrl-C can speed things up). With a bootable device connected, the ROM code will even detect and boot it for you. Cheers, -- Steve Maddison http://www.cosam.org/

On Thu, Mar 26, 2009 at 9:20 PM, John Floren <slawmaster at gmail.com> wrote: It's been a long time since I've used those ROMs, so I can't say for sure (though disks on the RQDX _will_ be DU disks), but since you've been pulling things out, you probably have to take a moment to ensure that you have a complete grant chain. As has been discussed on this list many times before, Qbus and Unibus are not like ISA or PCI busses that most folks are used to. You have to have "grant cards" to pass grant signals across empty bus slots, but since Qbus grant cards are somewhat rare (I only have 1 or 2 myself), they are typically used for making room around hot cards or for temporary changes to the bus. Normally, you move all the cards up so that the grant signals are contiguous. In a BA23 chassis, that means that you have to pay attention to which slots are Q22-only and which slots have a "CD" slot on the right. In the case of a BA23, the top three slots are CD slots and the bottom slots are Q22 in a serpentine fashion. I'm not going to draw the slot priority - it's documented well elsewhere including the list archives. Suffice it to say that if you don't know what that means, you have some reading to do. The consequence of not having continuous grant is that non-interrupt and non-DMA things like boot roms will work, but operating systems will not. You will be able to read the boot block off of the boot medium, but then the machine will mysteriously "freeze". It's not frozen, it's stuck in a loop, waiting for the boot device to interrupt, but since there's a gap in the bus, the interrupt will never come. There used to be configuration guides to show how to load up Qbus machines, but I wouldn't know where to point people now. Generally speaking, it's CPU up top, memory next, then fast devices that can't wait (ethernet and TK50-type tape controllers), then comms (because they are slow), and then disk (because they are buffered and can wait). Unlike a PC, it totally matters what order things are in because the interrupt and DMA priority isn't determined by jumper, it's determined by what cards are closer to the CPU (which is why the grant chain matters). So that's the long of it. The short of it is... don't expect a running Qbus machine to run after you pull anything out unless you reorganize it or at least stuff in grant cards into the blank spots. If you can find a suitable Minicomputer handbook that covers at least the 11/23 if not the 11/73, you should be in good shape - it's all covered in the front few chapters. The only difference from model to model is what particular memory, comms and disk cards are described in detail. The generalities about how the Qbus works didn't change from the mid-1970s when it originally came out - just specific things like grant-chain patterns with different backplanes, termination, number of address lines, etc. Once you get the right configuration (CPU, memory, disk, serial, boostrap...), I'm sure you'll have a fun robust system. There's just a little work to do on the front-end to get there. It's not as straightforward as firing up a modern machine, but it's fun once you get past that part. -ethan

The Data Systems Design A44324 board is to interface a DSD440 to a Qbus. ISTR the other DSD board is probably another DSD440 interface - they also made them for Unibus and Omnibus. -- Ian ________________________________________ From: cctalk-bounces at classiccmp.org [cctalk-bounces at classiccmp.org] On Behalf Of Pete Turnbull [pete at dunnington.plus.com] Sent: Thursday, March 26, 2009 1:32 AM To: On-Topic and Off-Topic Posts Subject: Re: PDP-11 bootup On 26/03/2009 05:34, John Floren wrote: M8192 is a dual-height KDJ-11A for an 11/73 and as Jerome pointed out that has no bootstrap. Possible reasons for being left at the ODT prompt include the HALT switch being on, or there being no bootstrap anywhere else. However, the next board, M7195, is a multifunction MXV-11B, which probably does have a bootstrap on it. Look for a pair of EPROMs on it. The manual is around the net, and I expect there's a copy on bitsavers. The KDJ-11 doesn't have all the ODT commands that some earlier QBus processors have. "G" is the command to "Go" to an address. "P" is the command to "Proceed" without initialising the bus. The only other commands on a KDJ-11 are slash ('/'), 'R', carriage-return and linefeed. I don't know exactly what the next few boards are, though I know Matrox made video boards for a variety of systems (DEC and S100 at least) and a QRGB is some sort of framestore and QVAF is a digitiser, I think. The Dataram board probably is memory, especially if densely packed with an array of chips. The chips numbers should tell us. M7941 is a DRV-11 parallel interface, which might have been used for a printer or simply for some other digital I/O to/from another device. M7555 is an RQDX3 MSCP disk controller, which interfaces via a breakout board to a dual floppy drive and/or one or more ST412-style hard drives. Are there disks and cables in the system? Thinking of that, what type of box is this in? What does the front panel look like? Jerome suggested it might be a BA23, but this mixture of cards suggests to me that it might just as well be a system that was upgraded from something older, and that it could be a BA11S or BA11N. You appear to have boards from at least three systems... This is a terminator board for a much older QBus system. Each is an RXV11 controller for a dual RX02 8" floppy drive - not useful unless you also have those specific drives. I don't know, but I'd guess some Texas Instruments GPIB/IEEE-488 instrument interface, like DEC's IBV11. Older QBus memory card from an 11/23 or possibly 11/03 system. There will be a suffix to the 8045 that will tell us the size of each. The have parity, so possibly from an 11/23. Dunno. DLV-11J quad serial line cards, mainly used with 11/03 and 11/23 systems. Later versions were renamed DLVJ1. They'd probably work (depending on the revision level) with your 11/73, but as you have two serial lines on the MXV11, probably not very useful. Another KDJ-11A CPU. Dunno. Another MXV-11 multifunction card. MSV-11L memory from an early microPDP-11, either 128KB or 256KB. This is an old dual-height LSI-11/2 CPU from an 11/03 system. Dunno. Terminator for that LSI-11/2 processor system. Does the number have a suffix? There were several version, some with bootstraps. Here you have the upper halves of four RLV11 disk controllers. These each need to be paired with an M8014 to do anything. They're from an 11/03 or early 11/23 system. A DL-11W serial line and line-time clock from a Unibus system. Don't plug this in to your Qbus! Matrox framestore, I believe. -- Pete Peter Turnbull Network Manager University of York

In article <7d3530220903252234i20bbc634gdf87ee1fa5e664ce at mail.gmail.com>, John Floren <slawmaster at gmail.com> writes: This is probably a Data Translation A/D or D/A card. Googling gets me: "DATA TRANSLATION DT2766 4 CHANNEL ANALOGUE MODULE,Q-BUS" -- "The Direct3D Graphics Pipeline" -- DirectX 9 draft available for download <http://www.xmission.com/~legalize/book/download/index.html> Legalize Adulthood! <http://blogs.xmission.com/legalize/>