ETH Lilith DPU-board remake

[jdreesen]

I intend to remake the ETH Lilith DPU board. This was an optional board with a MC6802 subsystem that could take over a Lilith by overruling the bitslice microcode- and addressbus. This was used during development of the Lilith and later to maintain the systems. I intend to use it in bringing my machine back to life.
Using GALs and modern(-ish) RAM & ROM allows for some free board space, this will be filled with either a 2.5"" IDE harddisk or a CF-card.
This should take care of the seriously problematic Honywel-Bull cartridge disk. This works is simulation, but has yet to run on real hardware.

Who needs / wants a copy of such a DPU board ?

Jos

 

[bear]

I'm in.

 

[jdreesen]

Slowly slowly, but it is coming along...

 

[jdreesen]

This board has now proven to be functional, and I already found a first fault in my Lilith using it. Alas it is not the only fault...

 

[Ruud]

Jos, jij bent goed bezig! (Dutch praise)

 

[jdreesen]

Just in case any potential Lilith owner needs this ( fat chance...) : here are the recreated asssembler sources for the 6802 code in the DPU unit.
U24 & U25 are the contents of the DPU eprom, "alu" checks major parts of the Lilith's ALU anc MCU boxes, as well as the main memory. "ifu" checks the instruction fetch unit as well as keybopard, mouse and RS232. "ex" allows to feed arbitrary microcode to the Lilith.
"alu", "ifu" and "ex" need to be loaded, in Motorola S9 format, into the DPU's RAM. These contain selfmodyfing code......

hpterm.c is a specific terminal program, running on a modern PC, needed to communicate with the DPU.

3 RAM errors, a defective keyboard serial-to-par unit, and a fault memory priority register found so far....
Still hope to have a functional Lilith to show at the Munich VCF !