• Please review our updated Terms and Rules here

Data General Nova 2/10

T

thunter0512

Veteran Member
Joined
Sep 27, 2020
Messages
855
Location
Perth in Western Australia
Recently I have been debugging and fixing a Nova 2/10 I found in a barn.

It has the following boards:
  • CPU
  • 16 K core
  • 8 K core
  • Cassette I/O
Here is a short log of what I have done so far.
  • Cleaned out several decades worth of accumulated dust, spider webs, dead insects and other debris.
  • Fixed the +15V rail in the power supply.
  • Fixed the leaking filter capacitor issue which kept tripping the RCD.
  • Replaced the broken AC3 examine/deposit switch.
  • Replaced two of the broken console (front-panel) lights. I could only find 2 of the miniature lights locally but have more on order from both the US and China.
  • Debugged an apparent problem on the console board and proved to myself that it is not the culprit. The problem is with the AC0-AC3 deposit switches where switch settings don’t take (i.e. a deposit behaves like an examine).
  • After studying the CPU board schematics I gained a reasonable understanding of the processing for the ACx deposit machinery and finally traced the problem to a faulty "Console ROM 0" which does not assert the /1WEN when needed. I am not sure what the exact failure mode of this bipolar PROM is. It may be that the fuse "healed" or that the associated open-collector driver transistor has died.
A tiny 2 word program incrementing AC0 runs successfully.

Best regards
Tom Hunter
 
I have since replaced the faulty "Console ROM 0" so ACx deposits work perfectly.

The TTY subsystem on the "Cassette I/O" board works perfectly after replacing a faulty N8H90 hex inverter shown as U113 in the schematics which is U109 on the actual PCB. I found it by tracing the SET_TTI_DONE signal from pin 19 of U40 (the LSI UART) following the tracks to locate the hex inverter. I replaced the faulty part with a socketed SN74F04 which is pin and speed compatible with the Signetics N8H90.

I have added the "Program Load" option to the CPU board so that the "Program Load" switch loads and runs a bootstrap program (found in the schematics from Bitsavers).

The 8 K core has a dead short between the +15V (VINHR) and GND and also between +5v and GND. There are a bunch of electrolytic and tantalum caps which might be causing this. To find the shorting compoent I am thinking of precision Ohm-meter or maybe applying current limited +15 V and then watch it with a thermal camera. Of course I have neither of these gadgets, but maybe there is a much simpler method I have not thought of.

Any thoughts on how to find the short(s) short of unsoldering components starting with capacitors? :)

Thanks
Tom
 
Last edited:
I forgot to mention in my earlier message that I have replaced all the front panel lights with "#2176 Miniature Bulb Wire Terminal Base" I bought from "Bulb Town". The originals were failing and also the old wire terminals have become very fragile. Some broke just looking at it. :)

In the past two days I have replaced a faulty address bus driver (SN75453) so that /MAB3 is now fully functional and there is no longer the nasty wrap-around of addresses 010000 (octal) and above.

I replaced a faulty Signetic N8H90 with a functionally identical SN7404 hex inverted where one of the inverters was bad causing the TTY Input done flag to not be set when a character was received from the TTY.

I also replaced a faulty SN7474 which impmented the busy/done flags for TTY Output.

Finally now all the diagnostics I have access to (and know how to run) work and pleasingly even BASIC runs perfectly.

The last job will be the nasty dead short on the 8 K core board and maybe a cleanup of the "Cassette I/O" board badly butchered by a previous owner.

Best regards
Tom
 
Last edited:
All problems with the 8 k core are now fixed and the system passes all diagnostics with the full complement of 24 k core. The final problem was a sense amplifier which developed a weird fault. Luckily I found a compatible replacement (SN75234) and it arrived today via UPS from the US.

Now what can I do with this machine. Without mass storage (disk and/or tape) you can't do too much with it.
 
Shouldn't be too difficult to rig up some kind of serial loader with which one could read in dumps of some interesting programs, I wouldn't think...
 
Assuming the Program Load switch is working properly, you can connect a laptop to the serial port and boot it from there. Set the front panel switches to 10 (octal) and press Program Load. See Section VI of How to Use the Nova Computers for the format. Use that to load the binary loader code from document 093-000003-06 (User's Manual: Binary Loader). If your machine doesn't have the Program Load option installed (or it's not working), then you have to toggle in the boot loader. Or just toggle in the binary loader directly. You shouldn't have to do this every time, as it lives at the high end of your (non-volatile!) core memory and is unlikely to be overwritten in normal operation.
 
The "Program Load" switch/feature makes it easy to get the "Binary Loader" loaded. The "Binary Loader" can then load diagnostics, BASIC, etc.
All loading by the "Binary Loader" is done via the TTY interface but also all other serial communications is done via that one serial port after the corresponding program has been loaded.

This means unless one is fortunate enough to have two "Cassette I/O" boards you won't have an alternate serial port to implement something like SerialDisk in the PDP-8 universe.

Someone with the right electronics skills could create a DG Nova equivalent of Joerg Hoppe's Unibone used in Unibus PDP-11 machines to provide emulated disk and/or tape drives.
 
There's still a lot you can do without a hard drive if you consider it a target system rather than a self-contained development machine. I would focus on what you CAN do with it as it exists today rather than fixate on what you CAN'T do because you don't have a particular bit of additional kit. And use any limitations as inspiration to come up with solutions, be they innovative or kludgey.

I doubt that "someone" will create a plug-and-play Nova disk emulator - there don't seem to be that many Novas in hobbyist hands, especially compared to the DEC machines...
 
Forgive me my intrusion.
@thunder0512
@all
does anybody know, are backplanes of various Nova models/generations are identical?
I'm looking for signal pinout for Nova 2 backplane (or for the backplane itself, either buy or change for extra boards).
I have a bunch of boards: CPUs, core memory 16K, various interfaces seems mostly from Nova 2. Some of the interface boards are marked 1969. It seems that they are from different Nova generations but still were used in a single system. Its history is unknown but I've bought them as a single bunch. No backplane neither chassis.
Still I don't give up and going to make the backplane myself.
Backplane signal pinout is available for SuperNova (bitsavers/archive.org) but is incomplete for Nova 2.

So, could anyone give a clue for a further info? I know about Bruce Ray but he seems not an easy person to reach.
 
For the Nova 2 backplane pinout you could check out page 12 of:

http://www.bitsavers.org/pdf/dg/Nova...chems_1973.pdf

I don't know what differences there are between the various backplanes. I suspect there was a level of interchangeability of Nova boards except the CPU boards. I/O boards and maybe even core may have been interchangeable.

There is currently an apparently working Nova 2/10 on Ebay but the seller want some serious money for it:

https://www.ebay.com/itm/194533442931
 
You must log in or register to reply here.
Back
Top