• Please review our updated Terms and Rules here

BM08 or BA08 for PDP-8/L

Sure Tom. In another thread ("Extended Arithmetic Element (EAE) re-implementation using current TTL ICs?") I hypothesized the following: "A CPLD implementation of the extended memory controller plus 28K-word of static RAM could be fit into a couple of empty slots in an 8/L, although some careful OTT additional signal cabling may be required as I'm not sure that there's a suitably large cluster of available slots that access all of the required signals; tying together modules in B/C/D 34, 35 and 36 is the logical group to consider. Alternative to a CPLD would be a fast small form-factor MCU-based emulation, which would open other surrogate-peripheral possibilities in the same space." Disk emulation would be a desirable emulation once there's sufficient emulated memory.

Obviously this approach would commandeer all of the normal extension cable-slots for I/O and DMA and thus preclude the use of any standard peripherals; those would become emulation-only. However it wouldn't interfere with the TTY or high-speed reader/punch which have separate direct controllers and cable connectors so real HW there would be A-OK.
 
I'm thinking that it should be feasible to put the logic bit of a BM8L on an Omnibus form factor card. The card would have headers for cables to the 8/L expansion sliots (which would also allow you to cable in Posibus devices ahead of it). It would act as "bus master", creating the necessary Omnibus signals and driving them out on the edge connector. (It's a lot of cables, so a card stiffener may also be needed.)

Then, what you'd do is populate any small Omnibus backplane (one is discussed in another thread) with this thing and some Omnibus memory.

One could even create cabling to a front panel extension, if that was considered important. (A real BM8L provides the other two switches each for IF and DF, and the missing lights for displaying the top two bits of the memory address.)

Vince
 
Yes, the missing IF/DF switches are an issue in both cases; something external to the 8/L chassis would need to be engineered. Are you suggesting that the Omnibus backplane could then host arbitrary peripherals or simply be a easier means to support RAM vie the re-implemented memory controller?
 
Yes, the missing IF/DF switches are an issue in both cases; something external to the 8/L chassis would need to be engineered. Are you suggesting that the Omnibus backplane could then host arbitrary peripherals or simply be a easier means to support RAM vie the re-implemented memory controller?
I'm not sure in practice how much the lack of front panel switches would be a liability. I find myself using the higher fields from the front panel quite rarely. The lack of lamps might be confusing, though.

I didn't actually remember whether DW8E functionality was included in the BM8L. Looking at the schematic, it looks like it isn't. So DW8E functionality would be desirable, else one might as well just add the half dozen chips it would take to implement 32K of memory and skip the Omnibus backplane.

Vince
 
I used both lamps and switches when testing/debugging BITD. In the present world, probably not as useful given that most folks presumably just want to run existing system software and indications of which 4K is active isn't critical. The one reason for going internal is that one could energize the first bit display in the extended address (EA) and well as use the existing FP switch-pair.

My personal desire is to also achieve AX08 functionality with the 8/L.
 
Great thread! Those of us with multiple pdp8/L's would love the opportunity to do a bit more. I personally have enough flip chips and parts to populate an 8kw BA08 but would be equally happy having an Omnibus adapter that would have the five paddles to go to the 8/L and provide access to Omnibus memory and an RX8 controller. Something more complex like a RL8a would be nice, but optional at the moment.

Chris
 
I did a Verilog extraction from the schematic, but I've run accross this message:
Error (163011): Design requires 210 I/O pins, but the selected device can contain only 164 I/O pins

Basically that is all the Posibus pins, the special pins, all the Omnibus pins, and a handful of extra pins for the delay lines and such.

Clearly that's too many pins for any remotely solderable device, so I'll have to partition things somehow. I think there are piles of pins that just need to buffers, so perhaps it won't be too difficult.
 
I've been working on a stackable module idea to plug in the expansion slots of the 8/L. My idea was to have a second serial port for the disk server, memory expansion and a bootstrap loader. Then I changed to another job, and now I'm working on an university nearby. But everything is new and it takes a lot of time from me... Way more then I expected to be honest... But I'll do everything for our students... :)

As time went by I thought: what is the use of the 8/L expansion besides the fun of making it... I mean, I have several PDP8 machines running OS/8. Then Dough started to work on a OS/8 disk emulator which can use just one serial port on the PDP8. That would mean the expansion for the 8/L doesn't need the extra serial port... In the end I thought, maybe it is a good idea to use the 8/L just for a paper tape Fortran version... Use the small machine for something that was meant for a PDP8/L...

I was offered a BM8/L recently. The the deal was to swap it for my Mark 8, but that is not going to happen... :) I'm happy with the PDP8/L as it is. If there comes an easy way to do more with it, then I'm all in.

Regards, Roland
 

Attachments

  • 1.jpg
    1.jpg
    439.4 KB · Views: 15
  • 2.jpg
    2.jpg
    401.4 KB · Views: 13
  • 3.jpg
    3.jpg
    346.4 KB · Views: 12
  • 4.jpg
    4.jpg
    401.5 KB · Views: 14
  • 5.jpg
    5.jpg
    334.9 KB · Views: 15
Made a little progress:
Total macrocells 141 / 128 ( 110 % )
Total pins 121 / 100 ( 121 % )

Mostly I removed connections to the standard busses which were never used for input or output.
 
Back
Top